Ten Critical Topics inReproductive Medicine

Produced by the ScienceAAASCustom Publishing Office

Sponsored by A Sponsored Supplement to Science

Ten Critical Topics inReproductive Medicine

Produced by the ScienceAAASCustom Publishing Office

Sponsored by A Sponsored Supplement to Science

1

CONTENTS

This booklet was produced by the ScienceAAAS Custom Publishing Office and sponsored by the Serono Symposia International Foundation Materials that appear in this booklet were commissioned edited and published by the ScienceAAAS Custom Publishing Office and were not reviewed or assessed by the Science Editorial staff Articles in this booklet can be cited using the following format [AUTHOR NAME(S)] [ARTICLE TITLE] in Ten Critical Topics in Reproductive Medicine S Sanders Ed (ScienceAAAS Washington DC 2013) pp [xx-xx] DOI 101126science34261641393-c

The Serono Symposia International Foundation is a member of the European lsquoGood CME Practice Grouprsquo proactively working with other medical education providers to enhance the standards of medical education provision

Editor Sean Sanders PhDProofreaderCopyeditor Yuse Lajiminmuhip Designer Amy Hardcastle

copy 2013 by The American Association for the Advancement of Science All rights reserved 13 December 2013

Produced by the ScienceAAAS Custom Publishing Office

2 The Science of ART Sean Sanders

3 From the Chief Executive of the Serono Symposia International Foundation Rachel Clark

4 ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

8 HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the Nobel JLHEversandAntonioPellicer

10 Germline Stem Cells in Adult Mammalian Ovaries DoriCWoodsandJonathanLTilly

13 Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte Maturation JiaPengJohnJEppigandMartinMMatzuk

15 An Introduction to Human Embryo Development ReneeAReijo-Pera

18 Environmentally Induced Epigenetic Transgenerational Inheritance of Disease MichaelKSkinner

21 Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic Pregnancy XiaofeiSunandSudhansuKDey

23 Single Embryo Transfer in IVF Live Birth Rates and Obstetrical Outcomes ChristinaBergh

25 Oocyte Vitrification A Major Milestone in Assisted Reproduction AnaCobo

27 Accurate Single Cell 24 Chromosome Aneuploidy Screening RichardTScottJrandNathanRTreff

29 What Is a ldquoNormalrdquo Semen Analysis DavidSGuzick

32 Circulating Angiogenic Factors and the Risk of Preeclampsia SAnanthKarumanchiandRaviThadhani

35 The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights RichardSLegroandMartinMMatzuk

40 Infertility The Male Factor DoloresJLambandLarryILipshultz

44 Molecular Interplay in Successful Implantation JeeyeonChaFelipeVilellaSudhansuDeyandCarlosSimoacuten

To learn more visit aaasorgplusyouproject2061

AAAS is here ndash promoting universal science literacy

In 1985 AAAS founded Project 2061 with the goal of helping all Americans become literate in science mathematics and technology With its landmark publications Science for All Americans and Benchmarks for Science Literacy Project 2061 set out recommendations for what all students should know and be able to do in science mathematics and technology by the time they graduate from high school Today many of the state standards in the United States have drawn their content from Project 2061

Every day Project 2061 staff use their expertise as teachers researchers and scientists to evaluate textbooks and assessments create conceptual strand maps for educators produce groundbreaking research and innovative books CD-ROMs and profes-sional development workshops for educators all in the service of achieving our goal of universal science literacy

As a AAAS member your dues help support Project 2061 as it works to improve science education If you are not yet a AAAS member join us Together we can make a difference

Cover design by the Serono Symposia International FoundationAll Images copy istockphotocom

2 3

ity andSterility respectively discuss the exciting directions in which they believe reproductive medicine and biology are heading There also are reviews of key research areas in re-productive medicine It all makes for a stimulating read which we hope you will enjoy We sincerely appreciate the contributions of all of the authors featured in this booklet It would not have been possible without their concerted writing efforts or indeed their dedication and passion for their work In particular Professor Carlos Simoacuten who was instrumental in the conception and execution of the conference and this booklet

SSIF is committed to sharing medi-cal knowledge and debate beyond our educational events With our Top Ten event we wanted to reach out to key opinion formers who might not be aware of our work and that is why we are pleased to be able to provide financial support for the publication of this important educational book-let with the assistance of ScienceAAAS

Rachel ClarkChief Executive OfficerSerono Symposia International Foundation

From the Chief Executive of the Serono Symposia International Foundation

Produced by the ScienceAAAS Custom Publishing Office

The Science of ART

In 1978 Robert Edwards and Patrick Steptoe achieved what many likely thought impossible the birth of a baby conceived outside of a womanrsquos body commonly referred to as the first test tube baby This was not only a breakthrough in biological research and a significant advance in the fieldrsquos understanding of human repro-duction but also a cultural and soci-etal leap that released woman (and men) from being at the mercy of the reproductive vagaries of their bodies A solution for many forms of infertility was now available affording couples previously unable to bear biological children that opportunity Although controversial at the time assisted re-productive technology (ART) is now seen as a standard instrument in the toolbox at a doctorrsquos disposal to as-sist infertile patients

However ART has not been with-out its challenges difficulties and failings An honest and open discus-sion of these hurdles as well as the implications of new research and advances is this area was the mo-tivation for a conference that took place in September 2013 in Florence Italy and ultimately this publication Ten previously published journal articles five on basic research and five on clinical work were chosen by a panel of experts to be pre-sented and challenged by peers in the field at this unique two-day meeting

Topics covered included the high incidence of multiple births follow-ing ART and how this might be ad-dressed through new techniques

At the Serono Symposia International Foundation (SSIF) we be-lieve that it is more important than ever to ensure that the very best independent continuing medical education (CME) is avail-able to clinicians scientists and researchers Every day across the world new research findings and improvements in diagno-sis and treatment mean new opportunities to improve outcomes for patients

This booklet focuses on one very important strand of our work in CME reproductive medicine It clearly exemplifies the ap-proach that SSIF takes to all of the areas we cover and the many events we hold each year sharing the knowledge and insights of internationally renowned experts with audiences worldwide through live educational activities and increasingly onlineTheTopTeninReproductiveMedicine project has been both

exciting and challenging With so much groundbreaking research worldwide how could we identify just 10 papers During the summer of 2012 we brought together an expert panel to select the best papers from a shortlist of 50 produced by a special-ist independent bibliographer The shortlist featured 25 papers related to basic research and 25 that were clinical and field-based The papers were chosen by the number of downloads they had received and the number of citations in the past 10 years Our expert panel of eight leading figures in reproductive medicine then spent many hours debating which would form the

final Top Ten asking which papers had really challenged tra-ditional views or which had actually changed diagnosis

and treatmentOur aim with the Top Ten project is to high-light research that has taken reproductive

medicine forward in benefiting patients Following a year of preparation in Sep-

tember 2013 we welcomed an inter-national audience to Florence Italy to hear from the top 10 authors them-selves and challenge their findings Out of that two day event grew this booklet featuring synopses of the top 10 papers and the subse-quent debates as well as contain-ing so much more Hans Evers and

Antonio Pellicer the editors in chief of Human Reproduction and Fertil-

This publication

provides a brief

overview of the

conference a

first of its kind

as far as format

and intent is

concerned

Improved genetic screening of embryos was also on the agenda including the challenges regarding testing for diseases with par-tial penetrance and the fear expressed by some of the specter of so-called designer babies This publication provides a brief overview of the conference a first of its kind as far as format and intent is concerned We also present a review article from each of the authors of the 10 original manuscripts providing context to their work in the current research landscape The 10 articles are prefaced by a short perspective from JLH Evers and An-tonio Pellicer both editors in chief of prestigious reproductive medicine journals and experts in this field who provide their as-sessment of ART following Edwards and Steptoersquos long-awaited Nobel prize in Medicine in 2010 We close out the booklet with three excellent reviews from leaders in reproductive medicine Legro and Matzuk who discuss disorders that affect ovarian function and related treatments Lamb and Lipshultz whose re-search has focused on the male role in infertility and Cha Vilel-la Dey and Simoacuten who together review the complex topic of the molecular mechanisms involved in embryo implantation in the uterus

It is our hope that this booklet made possible by the sponsor-ship and support of the Serono Symposia International Founda-tion will provide a broad review of the important and impact-ful subject of reproductive medicine summarizing advances challenging dogmas and advancing new paradigms as seen from the perspective of conference attendees and presenting authors The intention is not to stop here but to use this conference as a model to continue the discussion in this field as well as many others that impact our lives and well-being

Sean Sanders PhDEditor Custom Publishing OfficeScienceAAAS

Our aim with the

Top Ten project

is above all

to highlight

research that

has taken

reproductive

medicine

forward to

the benefit of

patients

4 5

In September 2013 scientists and cli-nicians in the field of human reproduc-tion participated in a groundbreaking conference held in Florence Italy Over the course of two days authors of the 10 most cited viewed or downloaded research articles in the field presented their papers Each author was confront-ed by a ldquochallengerrdquo after which par-ticipants discussed the paperrsquos findings and significance The goal to advance rigorous science in a field dominated by small-scale studies and trial and error ldquoEveryone is aware of the clini-cal relevance of assisted reproduction technology [ART]rdquo said Carlos Simoacuten professor of obstetrics and gynecology at the University of Valencia Spain and a conference scientific organizer ldquoWe need to increase our scientific rel-evancerdquo

In 2009 the World Health Organiza-tion officially recognized infertility as a disease Not all countries however have followed suit Consequently large-scale funding of fertility treatments has been scarce Regulation differs among countries making adoption of standard practices difficult And on the basic re-search front strict embryo protection laws in some countries restrict studies on human embryos ldquoIf you look at all the things we do in the clinic I would bet half of them have never been prov-en efficient in multicenter trialsrdquo said Hans Evers a professor in the depart-ment of obstetrics and gynecology at Maastricht University Medical Centre in Maastricht The Netherlands Intracyto-plasmic Sperm Injection (ICSI) whereby sperm is injected directly into an egg

for example is now being used to treat unexplained infertility in about half of all treatment cycles said Evers ldquoBut no one has ever shown that ICSI is necessary in unexplained infertilityrdquo

ICSI was an advance initially developed to help men with poor sperm quality become fathers when traditional in vitro fertilization failed The past 10 years have seen additional advances in technol-ogy aimed at helping greater numbers of people become parents too Several of the most promising of these were discussed at the conference Embryo cryopreservation for example may make in vi-tro fertilization (IVF) safer and less stressful by enabling a woman to undergo ovarian stimulation only once to retrieve eggs (See Cobo page 25) And embryo selection technology that can distinguish healthy embryos from unhealthy ones promises to raise success rates Further in the future it will be possible to diagnose any number of gene mutations in an embryo and harvest eggs from ovarian stem cells so that older couples can have biological children (see Legro page 35 and Lamb page 40)

However the best application of these technologies should be de-bated before they come into widespread use said Evers Moreover some technologies pose ethical questions that require societal input For example while diagnostic tests can detect the presence of a gene the tests may be meaningless since genes donrsquot predict with 100 certainty whether a person will actually develop a disease Also if women beyond normal reproductive age can indeed one day have biological children then how old is too old ldquoYou have to think about these things before the possibility arisesrdquo said Evers

To start addressing these issues The ldquoTop Ten in Reproductive Medicinerdquo conference highlighted the best papers as an example of how research into human reproduction and infertility should be done The format inspired by Simoacuten and developed by the Serono Symposia International Foundation (SSIF) was the first of its kind in the field By discussing each papermdashfive clinical and five basic researchmdashat length conference organizers hope to promote good clinical practice that comes from adopting technologies and prac-tices that are backed by strong scientific evidence and that have been validated by ethical debate

ASSiSTeD RePRODuCTiOn TeChnOlOgieS We CAn DO BeTTeR IVF has changed little since Robert Edwards and Patrick Steptoe introduced the technique over 35 years ago While there have been

advances in the field such as the ability to freeze sperm or eggs and test embryos for specific ge-netic abnormalities the process remains much the same Fertil-ize egg with sperm transfer the resulting embryo into a womanrsquos uterus and hope for a full-term pregnancy The technique has brought five million babies into the world and fulfilled the longing of parenthood for couples struggling with infertility

But IVF has a downside The drugs given to stimulate the ovaries to produce multiple eggs can overstimulate the ovaries and land wom-en in the hospital when the clinical condition is particularly risky The procedure is stressful primarily because itrsquos hard to predict whether it will work Depending on the age of the eggs the clinic and the procedure used any embryo transfer cycle has between a 25 and 40 chance of resulting in a baby With the odds in favor of failure doctors have typically transferred more than one embryo to boost the chance of success While the rate of triplets and higher order births has come down drastically the number of twins born from reproduc-tive technologies remains at about 30 of all ART births in the US A twin or multiple pregnancy has long been known to raise the risk of harm to both mother and babies Maternal complications include increased rates of pre-eclampsia gestational diabetes and preterm labor Risks for babies include low birth weight and prematurity as well as a higher risk of long-term disability and death as compared with singleton pregnancies

WiTh neW TeChnOlOgieS SuCCeSS iS MORe likelyAt the conference Christina Bergh from the Sahlgrenska University Hospital in Gothenburg Sweden presented her 2004 paper that for the first time provided evidence that transferring a single embryo yields almost the same birth rate as transferring two in women un-der age 36 (see page 23) Berghrsquos paper was the first clinical paper presented and set the backdrop for the presentations that followed It began with the question How can we treat infertility and inflict the least harm The paper was among the most highly viewed over the last 10 years primarily because it was published when rates of mul-tiple births using ART were much higher than they are today Since the paper was published single embryo transfer (SET) has been

widely adopted particularly in Northern Europe In Sweden doctors use SET in 75 to 80 of all IVF cycles in women of any age said Bergh But there re-main several countries in which it hasnrsquot caught on she added and this is likely because of dif-ferences in reimbursement prac-

tices for reproductive procedures Bergh presented additional data showing that SET works well in

women up to age 42 Additionally she studied children born from IVF procedures and showed that at the same time the rate of complica-tions had also decreased drastically ldquoHer work has really pushed the SET idea worldwiderdquo said her challenger Robert Fischer medical director of Fertility Center in Hamburg Germany and SSIF Scien-tific Committee Member However Swedish statistics may not hold for other populations said a clinician from the audience Patients in Scandinavia opt for IVF treatments much more quickly than patients in southern Europe he added Consequently IVF patients in south-ern Europe tend to be older and have poorer outcomes In such a setting SET may not be as successful Another participant pointed out that for SET to really take off success rates would need to be much higher

eMBRyO SeleCTiOn AnD FReezingOne way to raise success rates would be to distinguish good em-bryos from bad ones Enter Richard Scottrsquos 2010 paper on detect-ing aneuploidy in an embryo using whole genome amplification and single nucleotide polymorphism arrays (see page 27) Scott showed that the technology can distinguish abnormal from normal embryos with a 4 error rate Since publication additional data shows that the error rate is actually only around 1 said Scott director of repro-ductive science at Robert Wood Johnson Medical School in Basking Ridge New Jersey The paper was the first of a series of four that together show that the technology gives clinically meaningful results Scott said

For example in one study Scott and his colleagues biopsied em-bryos and transferred them back into women as part of their infertility care before they knew the biopsy results Once they determined the success or failure of the procedure they then compared the biopsy

The ldquoTop Ten in

Reproductive

Medicinerdquo

conference

highlighted the

best papers as

an example of

how research

on human

reproduction

and infertility

should be done

ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

Produced by the ScienceAAAS Custom Publishing Office

76

analysis data with birth outcomes and showed that the DNA array technology could detect with 99 accuracy an abnormal embryo that was unlikely to develop into a baby Another study showed that selecting embryos using this technology does indeed yield higher birth rates The predictive power of the technology is so great that the Reproductive Medicine Associates of New Jersey fertility clinic of which Scott is the director now performs SET using the diagnostic technology in 70 of all IVF cycles he said The clinic also offers the diagnostic service to other clinics

The technology isnrsquot easy to implement however and is expen-sive compared with other technologies on the market challenger Carlos Simoacuten pointed out and no one has yet demonstrated that the array technology yields better clinical outcomes as compared with those alternatives that are cheaper and easier to use Discus-sion also raised the point that array technologies may be rendered obsolete in the near future by next generation technologies that can sequence genes faster and cheaper In fact there are already com-panies selling tests that purport to predict the chance of a child in-heriting certain traits based on sequencing specific parental genes Adapting such tests to screen embryos for specific traits would be a simple taskmdashthe prospect of which has raised more than a few eyebrows Nevertheless technologies to parse normal embryos from abnormal ones are a boon to the field because they raise suc-cess rates Scottrsquos clinic is seeing pregnancy rates of greater than 50 he said and preliminary data suggests that embryo selec-tion is bringing preterm births rates down to a level on par with the general population

Microarrays arenrsquot the only promising embryo selection technol-ogy available During the meeting Renee Reijo Pera a professor in the department of obstetrics and gynecology at Stanford Univer-sity School of Medicine presented her 2010 paper on noninvasive imaging of human embryos (see page 15) The paper showed that time-lapse image analysis of two-day old zygotes together with gene expression profiling could with 93 accuracy predict which embryos

would develop to the blastocyst stage Three param-eters captured by imaging appeared to be the

most predictive duration of the first cell division the time interval between the

end of first mitosis and initiation of the second (mitosis is the replica-

tion of chromosomes that pre-cedes cell division) and the

time interval between the second and third mitoses ldquoThis is an example of out of the box thinkingrdquo said challenger Simoacuten However he pointed out there is a lack of randomized clinical trial data showing that embryo selection using

this technology results in higher birth rates Discus-

sion also raised the issue of a lack of standards with

groups trying to replicate the results using different measurement techniques

In yet another advance embryo freezing would make IVF safer and finally make it possible for young women suffering from ovarian cancer and premature ovarian failure to have families later in life Ana Cobo an embryologist from the Infertility Institute in Valencia presented her 2008 paper comparing fresh eggs to those cryopre-served by Cryotopmdasha commercially available method to vitrify hu-man tissues (see page 25) Vitrification is the conversion of a water containing solution to a glass-like solid that is free from the ice crys-tals normally formed during freezing She and her colleagues found that cryopreserved eggs were just as readily fertilized and able to develop into blastocysts as fresh ones Cobo also presented recent data showing that the pregnancy rate from cryopreserved embryos is similar to that of fresh embryos

ldquoThe paper got a lot of interest because oocytes have been ex-tremely difficult to freezerdquo said challenger Evers ldquoThe Cobo study showed for the first time in a large number of patients that you can do it and have good outcomesrdquo Moreover embryo cryopreserva-tion if broadly implemented might reduce the need to put back more than one embryo Evers added because you can freeze several and put them back in subsequent cycles if a woman doesnrsquot be-come pregnant It may even enhance pregnancy rates because the endometrial environment appears to be ldquomore friendlyrdquo during nor-mal cycles as compared to drug stimulated cycles he added

exPAnDing The BOunDARieS OF TReATMenTIn the future there may be no limit to the age at which a woman is able to bear children according to Jonathan Tilly professor and chair of biology at Northeastern University His 2012 paper provided evidence that human ovarian tissue contains pools of ovarian stem cells that develop into human oocytesmdashconfirming what researchers have found in other animals and overturning the long-held dogma positing that women are born with a fixed number of eggs that are never replenished (see page 10) Tilly first shocked the field in 2004 when he and his colleagues extracted ovarian stem cells from fe-male mice for the first time

In the recent paper Tilly and his colleagues developed a more sen-sitive method to identify and collect mouse ovarian stem cells Once the team confirmed that it had isolated the mouse ovarian stem cells by this new method it repeated the technique with frozen ovaries do-nated from sex-reassignment patients When cultured the retrieved oogonial stem cells (OSCs) turned into immature oocytes The team then tagged the cells with green fluorescent protein to trace them and injected them into pieces of human ovarian tissue which were then transplanted into mice After about two weeks the OSCs had differentiated into green-glowing cells that by all measures appeared to be human oocytes

ldquoThis was one of the most astonishing papers published in 2012rdquo said challenger Felipe Vilella from the Valencia Institute of Infertility in Spain However no one has yet shown that the human-derived OSCs can be fertilized to form an embryo Vilella said and no one knows whether any offspring would be healthy Although studies on animal derived OSCs have produced offspring no data on the health of the offspring has been provided said Vilella Nevertheless re-search on monkeys and ultimately humans is in the planning stages

Tilly said If the research pans out it would give women with ovarian cancer the chance to have children later in life It would also open the possibility that women well into their 50rsquos and even older could bear biological childrenmdashthe ethics of which ldquoare not for me to deciderdquo concluded Tilly

Vilellarsquos questions about the health of offspring raised another im-portant point of debate whether any of the procedures used in IVF currently or in the future might unwittingly be harming the result-ing offspring At the meeting Michael Skinner director of the center for reproductive biology at Washington State University in Pullman Washington presented his serendipitous finding revealed in his 2005 paper on the epigenetic transgenerational effects of endocrine disruptor chemicals on rodents (see page 18) If a pregnant female rodent was exposed to endocrine disrupting chemicals during a window when the fetusrsquos sex cells were developing the chemical exposure caused epigenetic alterationsmdashchanges in methylation patterns in the genomemdashin sperm These alterations were passed down through at least four generations and reduced male fertility Skinner has since tested several additional chemicals and found the same or similar effects on sperm and that these epigenetic changes can cause disease in subsequent generations In one study [ReprodToxicol34708 (2012)] for example Skinner showed that pregnant female rats exposed to the pesticide mixture DEETmdashcommonly found in insect repellantsmdashgave birth to pups with higher rates of disease that included pubertal abnormalities testis disease and ovarian disease Disease susceptibility was predominantly passed via males and persisted for three generations

A similar issue has been raised before in regard to IVF tech-niques Could the egg or culture media be exerting an effect on the epigenetics of the developing embryo To date rigorous studies

have not yet been done to answer this question But the debate certainly gen-erated enough interest and ideas for future studies

WRAP uPAfter the meeting participants lauded the challenge and debate format ldquoFrom my perspective this type of format should be broadly appliedrdquo said Tilly ldquoItrsquos an awesome ideardquo Scott added ldquoand Irsquom not prone to hyperbolerdquo Some however expressed a desire for more probing questions ldquoWe were all being a little too nicerdquo Reijo Pera said But everyone agreed that in contrast to large conferences where there is little time for discussion and questions this one enabled much more opportunity to learn Such feedback will be impor-tant in organizing the second Top Ten in reproductive medicine conference which is already in the works Accord-ing to Simoacuten the next one will likely take place in four yearsmdashenough time to publish new papers and to measure whether the first conference had an im-pact on spreading ideas and improving the way research in reproductive medi-cine is done

everyone

agreed that in

contrast to large

conferences

where there is

little time for

discussion and

questions this

one enabled

much more

opportunity to

learn

Produced by the ScienceAAAS Custom Publishing Office

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

1

CONTENTS

This booklet was produced by the ScienceAAAS Custom Publishing Office and sponsored by the Serono Symposia International Foundation Materials that appear in this booklet were commissioned edited and published by the ScienceAAAS Custom Publishing Office and were not reviewed or assessed by the Science Editorial staff Articles in this booklet can be cited using the following format [AUTHOR NAME(S)] [ARTICLE TITLE] in Ten Critical Topics in Reproductive Medicine S Sanders Ed (ScienceAAAS Washington DC 2013) pp [xx-xx] DOI 101126science34261641393-c

The Serono Symposia International Foundation is a member of the European lsquoGood CME Practice Grouprsquo proactively working with other medical education providers to enhance the standards of medical education provision

Editor Sean Sanders PhDProofreaderCopyeditor Yuse Lajiminmuhip Designer Amy Hardcastle

copy 2013 by The American Association for the Advancement of Science All rights reserved 13 December 2013

Produced by the ScienceAAAS Custom Publishing Office

2 The Science of ART Sean Sanders

3 From the Chief Executive of the Serono Symposia International Foundation Rachel Clark

4 ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

8 HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the Nobel JLHEversandAntonioPellicer

10 Germline Stem Cells in Adult Mammalian Ovaries DoriCWoodsandJonathanLTilly

13 Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte Maturation JiaPengJohnJEppigandMartinMMatzuk

15 An Introduction to Human Embryo Development ReneeAReijo-Pera

18 Environmentally Induced Epigenetic Transgenerational Inheritance of Disease MichaelKSkinner

21 Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic Pregnancy XiaofeiSunandSudhansuKDey

23 Single Embryo Transfer in IVF Live Birth Rates and Obstetrical Outcomes ChristinaBergh

25 Oocyte Vitrification A Major Milestone in Assisted Reproduction AnaCobo

27 Accurate Single Cell 24 Chromosome Aneuploidy Screening RichardTScottJrandNathanRTreff

29 What Is a ldquoNormalrdquo Semen Analysis DavidSGuzick

32 Circulating Angiogenic Factors and the Risk of Preeclampsia SAnanthKarumanchiandRaviThadhani

35 The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights RichardSLegroandMartinMMatzuk

40 Infertility The Male Factor DoloresJLambandLarryILipshultz

44 Molecular Interplay in Successful Implantation JeeyeonChaFelipeVilellaSudhansuDeyandCarlosSimoacuten

To learn more visit aaasorgplusyouproject2061

AAAS is here ndash promoting universal science literacy

In 1985 AAAS founded Project 2061 with the goal of helping all Americans become literate in science mathematics and technology With its landmark publications Science for All Americans and Benchmarks for Science Literacy Project 2061 set out recommendations for what all students should know and be able to do in science mathematics and technology by the time they graduate from high school Today many of the state standards in the United States have drawn their content from Project 2061

Every day Project 2061 staff use their expertise as teachers researchers and scientists to evaluate textbooks and assessments create conceptual strand maps for educators produce groundbreaking research and innovative books CD-ROMs and profes-sional development workshops for educators all in the service of achieving our goal of universal science literacy

As a AAAS member your dues help support Project 2061 as it works to improve science education If you are not yet a AAAS member join us Together we can make a difference

Cover design by the Serono Symposia International FoundationAll Images copy istockphotocom

2 3

ity andSterility respectively discuss the exciting directions in which they believe reproductive medicine and biology are heading There also are reviews of key research areas in re-productive medicine It all makes for a stimulating read which we hope you will enjoy We sincerely appreciate the contributions of all of the authors featured in this booklet It would not have been possible without their concerted writing efforts or indeed their dedication and passion for their work In particular Professor Carlos Simoacuten who was instrumental in the conception and execution of the conference and this booklet

SSIF is committed to sharing medi-cal knowledge and debate beyond our educational events With our Top Ten event we wanted to reach out to key opinion formers who might not be aware of our work and that is why we are pleased to be able to provide financial support for the publication of this important educational book-let with the assistance of ScienceAAAS

Rachel ClarkChief Executive OfficerSerono Symposia International Foundation

From the Chief Executive of the Serono Symposia International Foundation

Produced by the ScienceAAAS Custom Publishing Office

The Science of ART

In 1978 Robert Edwards and Patrick Steptoe achieved what many likely thought impossible the birth of a baby conceived outside of a womanrsquos body commonly referred to as the first test tube baby This was not only a breakthrough in biological research and a significant advance in the fieldrsquos understanding of human repro-duction but also a cultural and soci-etal leap that released woman (and men) from being at the mercy of the reproductive vagaries of their bodies A solution for many forms of infertility was now available affording couples previously unable to bear biological children that opportunity Although controversial at the time assisted re-productive technology (ART) is now seen as a standard instrument in the toolbox at a doctorrsquos disposal to as-sist infertile patients

However ART has not been with-out its challenges difficulties and failings An honest and open discus-sion of these hurdles as well as the implications of new research and advances is this area was the mo-tivation for a conference that took place in September 2013 in Florence Italy and ultimately this publication Ten previously published journal articles five on basic research and five on clinical work were chosen by a panel of experts to be pre-sented and challenged by peers in the field at this unique two-day meeting

Topics covered included the high incidence of multiple births follow-ing ART and how this might be ad-dressed through new techniques

At the Serono Symposia International Foundation (SSIF) we be-lieve that it is more important than ever to ensure that the very best independent continuing medical education (CME) is avail-able to clinicians scientists and researchers Every day across the world new research findings and improvements in diagno-sis and treatment mean new opportunities to improve outcomes for patients

This booklet focuses on one very important strand of our work in CME reproductive medicine It clearly exemplifies the ap-proach that SSIF takes to all of the areas we cover and the many events we hold each year sharing the knowledge and insights of internationally renowned experts with audiences worldwide through live educational activities and increasingly onlineTheTopTeninReproductiveMedicine project has been both

exciting and challenging With so much groundbreaking research worldwide how could we identify just 10 papers During the summer of 2012 we brought together an expert panel to select the best papers from a shortlist of 50 produced by a special-ist independent bibliographer The shortlist featured 25 papers related to basic research and 25 that were clinical and field-based The papers were chosen by the number of downloads they had received and the number of citations in the past 10 years Our expert panel of eight leading figures in reproductive medicine then spent many hours debating which would form the

final Top Ten asking which papers had really challenged tra-ditional views or which had actually changed diagnosis

and treatmentOur aim with the Top Ten project is to high-light research that has taken reproductive

medicine forward in benefiting patients Following a year of preparation in Sep-

tember 2013 we welcomed an inter-national audience to Florence Italy to hear from the top 10 authors them-selves and challenge their findings Out of that two day event grew this booklet featuring synopses of the top 10 papers and the subse-quent debates as well as contain-ing so much more Hans Evers and

Antonio Pellicer the editors in chief of Human Reproduction and Fertil-

This publication

provides a brief

overview of the

conference a

first of its kind

as far as format

and intent is

concerned

Improved genetic screening of embryos was also on the agenda including the challenges regarding testing for diseases with par-tial penetrance and the fear expressed by some of the specter of so-called designer babies This publication provides a brief overview of the conference a first of its kind as far as format and intent is concerned We also present a review article from each of the authors of the 10 original manuscripts providing context to their work in the current research landscape The 10 articles are prefaced by a short perspective from JLH Evers and An-tonio Pellicer both editors in chief of prestigious reproductive medicine journals and experts in this field who provide their as-sessment of ART following Edwards and Steptoersquos long-awaited Nobel prize in Medicine in 2010 We close out the booklet with three excellent reviews from leaders in reproductive medicine Legro and Matzuk who discuss disorders that affect ovarian function and related treatments Lamb and Lipshultz whose re-search has focused on the male role in infertility and Cha Vilel-la Dey and Simoacuten who together review the complex topic of the molecular mechanisms involved in embryo implantation in the uterus

It is our hope that this booklet made possible by the sponsor-ship and support of the Serono Symposia International Founda-tion will provide a broad review of the important and impact-ful subject of reproductive medicine summarizing advances challenging dogmas and advancing new paradigms as seen from the perspective of conference attendees and presenting authors The intention is not to stop here but to use this conference as a model to continue the discussion in this field as well as many others that impact our lives and well-being

Sean Sanders PhDEditor Custom Publishing OfficeScienceAAAS

Our aim with the

Top Ten project

is above all

to highlight

research that

has taken

reproductive

medicine

forward to

the benefit of

patients

4 5

In September 2013 scientists and cli-nicians in the field of human reproduc-tion participated in a groundbreaking conference held in Florence Italy Over the course of two days authors of the 10 most cited viewed or downloaded research articles in the field presented their papers Each author was confront-ed by a ldquochallengerrdquo after which par-ticipants discussed the paperrsquos findings and significance The goal to advance rigorous science in a field dominated by small-scale studies and trial and error ldquoEveryone is aware of the clini-cal relevance of assisted reproduction technology [ART]rdquo said Carlos Simoacuten professor of obstetrics and gynecology at the University of Valencia Spain and a conference scientific organizer ldquoWe need to increase our scientific rel-evancerdquo

In 2009 the World Health Organiza-tion officially recognized infertility as a disease Not all countries however have followed suit Consequently large-scale funding of fertility treatments has been scarce Regulation differs among countries making adoption of standard practices difficult And on the basic re-search front strict embryo protection laws in some countries restrict studies on human embryos ldquoIf you look at all the things we do in the clinic I would bet half of them have never been prov-en efficient in multicenter trialsrdquo said Hans Evers a professor in the depart-ment of obstetrics and gynecology at Maastricht University Medical Centre in Maastricht The Netherlands Intracyto-plasmic Sperm Injection (ICSI) whereby sperm is injected directly into an egg

for example is now being used to treat unexplained infertility in about half of all treatment cycles said Evers ldquoBut no one has ever shown that ICSI is necessary in unexplained infertilityrdquo

ICSI was an advance initially developed to help men with poor sperm quality become fathers when traditional in vitro fertilization failed The past 10 years have seen additional advances in technol-ogy aimed at helping greater numbers of people become parents too Several of the most promising of these were discussed at the conference Embryo cryopreservation for example may make in vi-tro fertilization (IVF) safer and less stressful by enabling a woman to undergo ovarian stimulation only once to retrieve eggs (See Cobo page 25) And embryo selection technology that can distinguish healthy embryos from unhealthy ones promises to raise success rates Further in the future it will be possible to diagnose any number of gene mutations in an embryo and harvest eggs from ovarian stem cells so that older couples can have biological children (see Legro page 35 and Lamb page 40)

However the best application of these technologies should be de-bated before they come into widespread use said Evers Moreover some technologies pose ethical questions that require societal input For example while diagnostic tests can detect the presence of a gene the tests may be meaningless since genes donrsquot predict with 100 certainty whether a person will actually develop a disease Also if women beyond normal reproductive age can indeed one day have biological children then how old is too old ldquoYou have to think about these things before the possibility arisesrdquo said Evers

To start addressing these issues The ldquoTop Ten in Reproductive Medicinerdquo conference highlighted the best papers as an example of how research into human reproduction and infertility should be done The format inspired by Simoacuten and developed by the Serono Symposia International Foundation (SSIF) was the first of its kind in the field By discussing each papermdashfive clinical and five basic researchmdashat length conference organizers hope to promote good clinical practice that comes from adopting technologies and prac-tices that are backed by strong scientific evidence and that have been validated by ethical debate

ASSiSTeD RePRODuCTiOn TeChnOlOgieS We CAn DO BeTTeR IVF has changed little since Robert Edwards and Patrick Steptoe introduced the technique over 35 years ago While there have been

advances in the field such as the ability to freeze sperm or eggs and test embryos for specific ge-netic abnormalities the process remains much the same Fertil-ize egg with sperm transfer the resulting embryo into a womanrsquos uterus and hope for a full-term pregnancy The technique has brought five million babies into the world and fulfilled the longing of parenthood for couples struggling with infertility

But IVF has a downside The drugs given to stimulate the ovaries to produce multiple eggs can overstimulate the ovaries and land wom-en in the hospital when the clinical condition is particularly risky The procedure is stressful primarily because itrsquos hard to predict whether it will work Depending on the age of the eggs the clinic and the procedure used any embryo transfer cycle has between a 25 and 40 chance of resulting in a baby With the odds in favor of failure doctors have typically transferred more than one embryo to boost the chance of success While the rate of triplets and higher order births has come down drastically the number of twins born from reproduc-tive technologies remains at about 30 of all ART births in the US A twin or multiple pregnancy has long been known to raise the risk of harm to both mother and babies Maternal complications include increased rates of pre-eclampsia gestational diabetes and preterm labor Risks for babies include low birth weight and prematurity as well as a higher risk of long-term disability and death as compared with singleton pregnancies

WiTh neW TeChnOlOgieS SuCCeSS iS MORe likelyAt the conference Christina Bergh from the Sahlgrenska University Hospital in Gothenburg Sweden presented her 2004 paper that for the first time provided evidence that transferring a single embryo yields almost the same birth rate as transferring two in women un-der age 36 (see page 23) Berghrsquos paper was the first clinical paper presented and set the backdrop for the presentations that followed It began with the question How can we treat infertility and inflict the least harm The paper was among the most highly viewed over the last 10 years primarily because it was published when rates of mul-tiple births using ART were much higher than they are today Since the paper was published single embryo transfer (SET) has been

widely adopted particularly in Northern Europe In Sweden doctors use SET in 75 to 80 of all IVF cycles in women of any age said Bergh But there re-main several countries in which it hasnrsquot caught on she added and this is likely because of dif-ferences in reimbursement prac-

tices for reproductive procedures Bergh presented additional data showing that SET works well in

women up to age 42 Additionally she studied children born from IVF procedures and showed that at the same time the rate of complica-tions had also decreased drastically ldquoHer work has really pushed the SET idea worldwiderdquo said her challenger Robert Fischer medical director of Fertility Center in Hamburg Germany and SSIF Scien-tific Committee Member However Swedish statistics may not hold for other populations said a clinician from the audience Patients in Scandinavia opt for IVF treatments much more quickly than patients in southern Europe he added Consequently IVF patients in south-ern Europe tend to be older and have poorer outcomes In such a setting SET may not be as successful Another participant pointed out that for SET to really take off success rates would need to be much higher

eMBRyO SeleCTiOn AnD FReezingOne way to raise success rates would be to distinguish good em-bryos from bad ones Enter Richard Scottrsquos 2010 paper on detect-ing aneuploidy in an embryo using whole genome amplification and single nucleotide polymorphism arrays (see page 27) Scott showed that the technology can distinguish abnormal from normal embryos with a 4 error rate Since publication additional data shows that the error rate is actually only around 1 said Scott director of repro-ductive science at Robert Wood Johnson Medical School in Basking Ridge New Jersey The paper was the first of a series of four that together show that the technology gives clinically meaningful results Scott said

For example in one study Scott and his colleagues biopsied em-bryos and transferred them back into women as part of their infertility care before they knew the biopsy results Once they determined the success or failure of the procedure they then compared the biopsy

The ldquoTop Ten in

Reproductive

Medicinerdquo

conference

highlighted the

best papers as

an example of

how research

on human

reproduction

and infertility

should be done

ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

Produced by the ScienceAAAS Custom Publishing Office

76

analysis data with birth outcomes and showed that the DNA array technology could detect with 99 accuracy an abnormal embryo that was unlikely to develop into a baby Another study showed that selecting embryos using this technology does indeed yield higher birth rates The predictive power of the technology is so great that the Reproductive Medicine Associates of New Jersey fertility clinic of which Scott is the director now performs SET using the diagnostic technology in 70 of all IVF cycles he said The clinic also offers the diagnostic service to other clinics

The technology isnrsquot easy to implement however and is expen-sive compared with other technologies on the market challenger Carlos Simoacuten pointed out and no one has yet demonstrated that the array technology yields better clinical outcomes as compared with those alternatives that are cheaper and easier to use Discus-sion also raised the point that array technologies may be rendered obsolete in the near future by next generation technologies that can sequence genes faster and cheaper In fact there are already com-panies selling tests that purport to predict the chance of a child in-heriting certain traits based on sequencing specific parental genes Adapting such tests to screen embryos for specific traits would be a simple taskmdashthe prospect of which has raised more than a few eyebrows Nevertheless technologies to parse normal embryos from abnormal ones are a boon to the field because they raise suc-cess rates Scottrsquos clinic is seeing pregnancy rates of greater than 50 he said and preliminary data suggests that embryo selec-tion is bringing preterm births rates down to a level on par with the general population

Microarrays arenrsquot the only promising embryo selection technol-ogy available During the meeting Renee Reijo Pera a professor in the department of obstetrics and gynecology at Stanford Univer-sity School of Medicine presented her 2010 paper on noninvasive imaging of human embryos (see page 15) The paper showed that time-lapse image analysis of two-day old zygotes together with gene expression profiling could with 93 accuracy predict which embryos

would develop to the blastocyst stage Three param-eters captured by imaging appeared to be the

most predictive duration of the first cell division the time interval between the

end of first mitosis and initiation of the second (mitosis is the replica-

tion of chromosomes that pre-cedes cell division) and the

time interval between the second and third mitoses ldquoThis is an example of out of the box thinkingrdquo said challenger Simoacuten However he pointed out there is a lack of randomized clinical trial data showing that embryo selection using

this technology results in higher birth rates Discus-

sion also raised the issue of a lack of standards with

groups trying to replicate the results using different measurement techniques

In yet another advance embryo freezing would make IVF safer and finally make it possible for young women suffering from ovarian cancer and premature ovarian failure to have families later in life Ana Cobo an embryologist from the Infertility Institute in Valencia presented her 2008 paper comparing fresh eggs to those cryopre-served by Cryotopmdasha commercially available method to vitrify hu-man tissues (see page 25) Vitrification is the conversion of a water containing solution to a glass-like solid that is free from the ice crys-tals normally formed during freezing She and her colleagues found that cryopreserved eggs were just as readily fertilized and able to develop into blastocysts as fresh ones Cobo also presented recent data showing that the pregnancy rate from cryopreserved embryos is similar to that of fresh embryos

ldquoThe paper got a lot of interest because oocytes have been ex-tremely difficult to freezerdquo said challenger Evers ldquoThe Cobo study showed for the first time in a large number of patients that you can do it and have good outcomesrdquo Moreover embryo cryopreserva-tion if broadly implemented might reduce the need to put back more than one embryo Evers added because you can freeze several and put them back in subsequent cycles if a woman doesnrsquot be-come pregnant It may even enhance pregnancy rates because the endometrial environment appears to be ldquomore friendlyrdquo during nor-mal cycles as compared to drug stimulated cycles he added

exPAnDing The BOunDARieS OF TReATMenTIn the future there may be no limit to the age at which a woman is able to bear children according to Jonathan Tilly professor and chair of biology at Northeastern University His 2012 paper provided evidence that human ovarian tissue contains pools of ovarian stem cells that develop into human oocytesmdashconfirming what researchers have found in other animals and overturning the long-held dogma positing that women are born with a fixed number of eggs that are never replenished (see page 10) Tilly first shocked the field in 2004 when he and his colleagues extracted ovarian stem cells from fe-male mice for the first time

In the recent paper Tilly and his colleagues developed a more sen-sitive method to identify and collect mouse ovarian stem cells Once the team confirmed that it had isolated the mouse ovarian stem cells by this new method it repeated the technique with frozen ovaries do-nated from sex-reassignment patients When cultured the retrieved oogonial stem cells (OSCs) turned into immature oocytes The team then tagged the cells with green fluorescent protein to trace them and injected them into pieces of human ovarian tissue which were then transplanted into mice After about two weeks the OSCs had differentiated into green-glowing cells that by all measures appeared to be human oocytes

ldquoThis was one of the most astonishing papers published in 2012rdquo said challenger Felipe Vilella from the Valencia Institute of Infertility in Spain However no one has yet shown that the human-derived OSCs can be fertilized to form an embryo Vilella said and no one knows whether any offspring would be healthy Although studies on animal derived OSCs have produced offspring no data on the health of the offspring has been provided said Vilella Nevertheless re-search on monkeys and ultimately humans is in the planning stages

Tilly said If the research pans out it would give women with ovarian cancer the chance to have children later in life It would also open the possibility that women well into their 50rsquos and even older could bear biological childrenmdashthe ethics of which ldquoare not for me to deciderdquo concluded Tilly

Vilellarsquos questions about the health of offspring raised another im-portant point of debate whether any of the procedures used in IVF currently or in the future might unwittingly be harming the result-ing offspring At the meeting Michael Skinner director of the center for reproductive biology at Washington State University in Pullman Washington presented his serendipitous finding revealed in his 2005 paper on the epigenetic transgenerational effects of endocrine disruptor chemicals on rodents (see page 18) If a pregnant female rodent was exposed to endocrine disrupting chemicals during a window when the fetusrsquos sex cells were developing the chemical exposure caused epigenetic alterationsmdashchanges in methylation patterns in the genomemdashin sperm These alterations were passed down through at least four generations and reduced male fertility Skinner has since tested several additional chemicals and found the same or similar effects on sperm and that these epigenetic changes can cause disease in subsequent generations In one study [ReprodToxicol34708 (2012)] for example Skinner showed that pregnant female rats exposed to the pesticide mixture DEETmdashcommonly found in insect repellantsmdashgave birth to pups with higher rates of disease that included pubertal abnormalities testis disease and ovarian disease Disease susceptibility was predominantly passed via males and persisted for three generations

A similar issue has been raised before in regard to IVF tech-niques Could the egg or culture media be exerting an effect on the epigenetics of the developing embryo To date rigorous studies

have not yet been done to answer this question But the debate certainly gen-erated enough interest and ideas for future studies

WRAP uPAfter the meeting participants lauded the challenge and debate format ldquoFrom my perspective this type of format should be broadly appliedrdquo said Tilly ldquoItrsquos an awesome ideardquo Scott added ldquoand Irsquom not prone to hyperbolerdquo Some however expressed a desire for more probing questions ldquoWe were all being a little too nicerdquo Reijo Pera said But everyone agreed that in contrast to large conferences where there is little time for discussion and questions this one enabled much more opportunity to learn Such feedback will be impor-tant in organizing the second Top Ten in reproductive medicine conference which is already in the works Accord-ing to Simoacuten the next one will likely take place in four yearsmdashenough time to publish new papers and to measure whether the first conference had an im-pact on spreading ideas and improving the way research in reproductive medi-cine is done

everyone

agreed that in

contrast to large

conferences

where there is

little time for

discussion and

questions this

one enabled

much more

opportunity to

learn

Produced by the ScienceAAAS Custom Publishing Office

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

2 3

ity andSterility respectively discuss the exciting directions in which they believe reproductive medicine and biology are heading There also are reviews of key research areas in re-productive medicine It all makes for a stimulating read which we hope you will enjoy We sincerely appreciate the contributions of all of the authors featured in this booklet It would not have been possible without their concerted writing efforts or indeed their dedication and passion for their work In particular Professor Carlos Simoacuten who was instrumental in the conception and execution of the conference and this booklet

SSIF is committed to sharing medi-cal knowledge and debate beyond our educational events With our Top Ten event we wanted to reach out to key opinion formers who might not be aware of our work and that is why we are pleased to be able to provide financial support for the publication of this important educational book-let with the assistance of ScienceAAAS

Rachel ClarkChief Executive OfficerSerono Symposia International Foundation

From the Chief Executive of the Serono Symposia International Foundation

Produced by the ScienceAAAS Custom Publishing Office

The Science of ART

In 1978 Robert Edwards and Patrick Steptoe achieved what many likely thought impossible the birth of a baby conceived outside of a womanrsquos body commonly referred to as the first test tube baby This was not only a breakthrough in biological research and a significant advance in the fieldrsquos understanding of human repro-duction but also a cultural and soci-etal leap that released woman (and men) from being at the mercy of the reproductive vagaries of their bodies A solution for many forms of infertility was now available affording couples previously unable to bear biological children that opportunity Although controversial at the time assisted re-productive technology (ART) is now seen as a standard instrument in the toolbox at a doctorrsquos disposal to as-sist infertile patients

However ART has not been with-out its challenges difficulties and failings An honest and open discus-sion of these hurdles as well as the implications of new research and advances is this area was the mo-tivation for a conference that took place in September 2013 in Florence Italy and ultimately this publication Ten previously published journal articles five on basic research and five on clinical work were chosen by a panel of experts to be pre-sented and challenged by peers in the field at this unique two-day meeting

Topics covered included the high incidence of multiple births follow-ing ART and how this might be ad-dressed through new techniques

At the Serono Symposia International Foundation (SSIF) we be-lieve that it is more important than ever to ensure that the very best independent continuing medical education (CME) is avail-able to clinicians scientists and researchers Every day across the world new research findings and improvements in diagno-sis and treatment mean new opportunities to improve outcomes for patients

This booklet focuses on one very important strand of our work in CME reproductive medicine It clearly exemplifies the ap-proach that SSIF takes to all of the areas we cover and the many events we hold each year sharing the knowledge and insights of internationally renowned experts with audiences worldwide through live educational activities and increasingly onlineTheTopTeninReproductiveMedicine project has been both

exciting and challenging With so much groundbreaking research worldwide how could we identify just 10 papers During the summer of 2012 we brought together an expert panel to select the best papers from a shortlist of 50 produced by a special-ist independent bibliographer The shortlist featured 25 papers related to basic research and 25 that were clinical and field-based The papers were chosen by the number of downloads they had received and the number of citations in the past 10 years Our expert panel of eight leading figures in reproductive medicine then spent many hours debating which would form the

final Top Ten asking which papers had really challenged tra-ditional views or which had actually changed diagnosis

and treatmentOur aim with the Top Ten project is to high-light research that has taken reproductive

medicine forward in benefiting patients Following a year of preparation in Sep-

tember 2013 we welcomed an inter-national audience to Florence Italy to hear from the top 10 authors them-selves and challenge their findings Out of that two day event grew this booklet featuring synopses of the top 10 papers and the subse-quent debates as well as contain-ing so much more Hans Evers and

Antonio Pellicer the editors in chief of Human Reproduction and Fertil-

This publication

provides a brief

overview of the

conference a

first of its kind

as far as format

and intent is

concerned

Improved genetic screening of embryos was also on the agenda including the challenges regarding testing for diseases with par-tial penetrance and the fear expressed by some of the specter of so-called designer babies This publication provides a brief overview of the conference a first of its kind as far as format and intent is concerned We also present a review article from each of the authors of the 10 original manuscripts providing context to their work in the current research landscape The 10 articles are prefaced by a short perspective from JLH Evers and An-tonio Pellicer both editors in chief of prestigious reproductive medicine journals and experts in this field who provide their as-sessment of ART following Edwards and Steptoersquos long-awaited Nobel prize in Medicine in 2010 We close out the booklet with three excellent reviews from leaders in reproductive medicine Legro and Matzuk who discuss disorders that affect ovarian function and related treatments Lamb and Lipshultz whose re-search has focused on the male role in infertility and Cha Vilel-la Dey and Simoacuten who together review the complex topic of the molecular mechanisms involved in embryo implantation in the uterus

It is our hope that this booklet made possible by the sponsor-ship and support of the Serono Symposia International Founda-tion will provide a broad review of the important and impact-ful subject of reproductive medicine summarizing advances challenging dogmas and advancing new paradigms as seen from the perspective of conference attendees and presenting authors The intention is not to stop here but to use this conference as a model to continue the discussion in this field as well as many others that impact our lives and well-being

Sean Sanders PhDEditor Custom Publishing OfficeScienceAAAS

Our aim with the

Top Ten project

is above all

to highlight

research that

has taken

reproductive

medicine

forward to

the benefit of

patients

4 5

In September 2013 scientists and cli-nicians in the field of human reproduc-tion participated in a groundbreaking conference held in Florence Italy Over the course of two days authors of the 10 most cited viewed or downloaded research articles in the field presented their papers Each author was confront-ed by a ldquochallengerrdquo after which par-ticipants discussed the paperrsquos findings and significance The goal to advance rigorous science in a field dominated by small-scale studies and trial and error ldquoEveryone is aware of the clini-cal relevance of assisted reproduction technology [ART]rdquo said Carlos Simoacuten professor of obstetrics and gynecology at the University of Valencia Spain and a conference scientific organizer ldquoWe need to increase our scientific rel-evancerdquo

In 2009 the World Health Organiza-tion officially recognized infertility as a disease Not all countries however have followed suit Consequently large-scale funding of fertility treatments has been scarce Regulation differs among countries making adoption of standard practices difficult And on the basic re-search front strict embryo protection laws in some countries restrict studies on human embryos ldquoIf you look at all the things we do in the clinic I would bet half of them have never been prov-en efficient in multicenter trialsrdquo said Hans Evers a professor in the depart-ment of obstetrics and gynecology at Maastricht University Medical Centre in Maastricht The Netherlands Intracyto-plasmic Sperm Injection (ICSI) whereby sperm is injected directly into an egg

for example is now being used to treat unexplained infertility in about half of all treatment cycles said Evers ldquoBut no one has ever shown that ICSI is necessary in unexplained infertilityrdquo

ICSI was an advance initially developed to help men with poor sperm quality become fathers when traditional in vitro fertilization failed The past 10 years have seen additional advances in technol-ogy aimed at helping greater numbers of people become parents too Several of the most promising of these were discussed at the conference Embryo cryopreservation for example may make in vi-tro fertilization (IVF) safer and less stressful by enabling a woman to undergo ovarian stimulation only once to retrieve eggs (See Cobo page 25) And embryo selection technology that can distinguish healthy embryos from unhealthy ones promises to raise success rates Further in the future it will be possible to diagnose any number of gene mutations in an embryo and harvest eggs from ovarian stem cells so that older couples can have biological children (see Legro page 35 and Lamb page 40)

However the best application of these technologies should be de-bated before they come into widespread use said Evers Moreover some technologies pose ethical questions that require societal input For example while diagnostic tests can detect the presence of a gene the tests may be meaningless since genes donrsquot predict with 100 certainty whether a person will actually develop a disease Also if women beyond normal reproductive age can indeed one day have biological children then how old is too old ldquoYou have to think about these things before the possibility arisesrdquo said Evers

To start addressing these issues The ldquoTop Ten in Reproductive Medicinerdquo conference highlighted the best papers as an example of how research into human reproduction and infertility should be done The format inspired by Simoacuten and developed by the Serono Symposia International Foundation (SSIF) was the first of its kind in the field By discussing each papermdashfive clinical and five basic researchmdashat length conference organizers hope to promote good clinical practice that comes from adopting technologies and prac-tices that are backed by strong scientific evidence and that have been validated by ethical debate

ASSiSTeD RePRODuCTiOn TeChnOlOgieS We CAn DO BeTTeR IVF has changed little since Robert Edwards and Patrick Steptoe introduced the technique over 35 years ago While there have been

advances in the field such as the ability to freeze sperm or eggs and test embryos for specific ge-netic abnormalities the process remains much the same Fertil-ize egg with sperm transfer the resulting embryo into a womanrsquos uterus and hope for a full-term pregnancy The technique has brought five million babies into the world and fulfilled the longing of parenthood for couples struggling with infertility

But IVF has a downside The drugs given to stimulate the ovaries to produce multiple eggs can overstimulate the ovaries and land wom-en in the hospital when the clinical condition is particularly risky The procedure is stressful primarily because itrsquos hard to predict whether it will work Depending on the age of the eggs the clinic and the procedure used any embryo transfer cycle has between a 25 and 40 chance of resulting in a baby With the odds in favor of failure doctors have typically transferred more than one embryo to boost the chance of success While the rate of triplets and higher order births has come down drastically the number of twins born from reproduc-tive technologies remains at about 30 of all ART births in the US A twin or multiple pregnancy has long been known to raise the risk of harm to both mother and babies Maternal complications include increased rates of pre-eclampsia gestational diabetes and preterm labor Risks for babies include low birth weight and prematurity as well as a higher risk of long-term disability and death as compared with singleton pregnancies

WiTh neW TeChnOlOgieS SuCCeSS iS MORe likelyAt the conference Christina Bergh from the Sahlgrenska University Hospital in Gothenburg Sweden presented her 2004 paper that for the first time provided evidence that transferring a single embryo yields almost the same birth rate as transferring two in women un-der age 36 (see page 23) Berghrsquos paper was the first clinical paper presented and set the backdrop for the presentations that followed It began with the question How can we treat infertility and inflict the least harm The paper was among the most highly viewed over the last 10 years primarily because it was published when rates of mul-tiple births using ART were much higher than they are today Since the paper was published single embryo transfer (SET) has been

widely adopted particularly in Northern Europe In Sweden doctors use SET in 75 to 80 of all IVF cycles in women of any age said Bergh But there re-main several countries in which it hasnrsquot caught on she added and this is likely because of dif-ferences in reimbursement prac-

tices for reproductive procedures Bergh presented additional data showing that SET works well in

women up to age 42 Additionally she studied children born from IVF procedures and showed that at the same time the rate of complica-tions had also decreased drastically ldquoHer work has really pushed the SET idea worldwiderdquo said her challenger Robert Fischer medical director of Fertility Center in Hamburg Germany and SSIF Scien-tific Committee Member However Swedish statistics may not hold for other populations said a clinician from the audience Patients in Scandinavia opt for IVF treatments much more quickly than patients in southern Europe he added Consequently IVF patients in south-ern Europe tend to be older and have poorer outcomes In such a setting SET may not be as successful Another participant pointed out that for SET to really take off success rates would need to be much higher

eMBRyO SeleCTiOn AnD FReezingOne way to raise success rates would be to distinguish good em-bryos from bad ones Enter Richard Scottrsquos 2010 paper on detect-ing aneuploidy in an embryo using whole genome amplification and single nucleotide polymorphism arrays (see page 27) Scott showed that the technology can distinguish abnormal from normal embryos with a 4 error rate Since publication additional data shows that the error rate is actually only around 1 said Scott director of repro-ductive science at Robert Wood Johnson Medical School in Basking Ridge New Jersey The paper was the first of a series of four that together show that the technology gives clinically meaningful results Scott said

For example in one study Scott and his colleagues biopsied em-bryos and transferred them back into women as part of their infertility care before they knew the biopsy results Once they determined the success or failure of the procedure they then compared the biopsy

The ldquoTop Ten in

Reproductive

Medicinerdquo

conference

highlighted the

best papers as

an example of

how research

on human

reproduction

and infertility

should be done

ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

Produced by the ScienceAAAS Custom Publishing Office

76

analysis data with birth outcomes and showed that the DNA array technology could detect with 99 accuracy an abnormal embryo that was unlikely to develop into a baby Another study showed that selecting embryos using this technology does indeed yield higher birth rates The predictive power of the technology is so great that the Reproductive Medicine Associates of New Jersey fertility clinic of which Scott is the director now performs SET using the diagnostic technology in 70 of all IVF cycles he said The clinic also offers the diagnostic service to other clinics

The technology isnrsquot easy to implement however and is expen-sive compared with other technologies on the market challenger Carlos Simoacuten pointed out and no one has yet demonstrated that the array technology yields better clinical outcomes as compared with those alternatives that are cheaper and easier to use Discus-sion also raised the point that array technologies may be rendered obsolete in the near future by next generation technologies that can sequence genes faster and cheaper In fact there are already com-panies selling tests that purport to predict the chance of a child in-heriting certain traits based on sequencing specific parental genes Adapting such tests to screen embryos for specific traits would be a simple taskmdashthe prospect of which has raised more than a few eyebrows Nevertheless technologies to parse normal embryos from abnormal ones are a boon to the field because they raise suc-cess rates Scottrsquos clinic is seeing pregnancy rates of greater than 50 he said and preliminary data suggests that embryo selec-tion is bringing preterm births rates down to a level on par with the general population

Microarrays arenrsquot the only promising embryo selection technol-ogy available During the meeting Renee Reijo Pera a professor in the department of obstetrics and gynecology at Stanford Univer-sity School of Medicine presented her 2010 paper on noninvasive imaging of human embryos (see page 15) The paper showed that time-lapse image analysis of two-day old zygotes together with gene expression profiling could with 93 accuracy predict which embryos

would develop to the blastocyst stage Three param-eters captured by imaging appeared to be the

most predictive duration of the first cell division the time interval between the

end of first mitosis and initiation of the second (mitosis is the replica-

tion of chromosomes that pre-cedes cell division) and the

time interval between the second and third mitoses ldquoThis is an example of out of the box thinkingrdquo said challenger Simoacuten However he pointed out there is a lack of randomized clinical trial data showing that embryo selection using

this technology results in higher birth rates Discus-

sion also raised the issue of a lack of standards with

groups trying to replicate the results using different measurement techniques

In yet another advance embryo freezing would make IVF safer and finally make it possible for young women suffering from ovarian cancer and premature ovarian failure to have families later in life Ana Cobo an embryologist from the Infertility Institute in Valencia presented her 2008 paper comparing fresh eggs to those cryopre-served by Cryotopmdasha commercially available method to vitrify hu-man tissues (see page 25) Vitrification is the conversion of a water containing solution to a glass-like solid that is free from the ice crys-tals normally formed during freezing She and her colleagues found that cryopreserved eggs were just as readily fertilized and able to develop into blastocysts as fresh ones Cobo also presented recent data showing that the pregnancy rate from cryopreserved embryos is similar to that of fresh embryos

ldquoThe paper got a lot of interest because oocytes have been ex-tremely difficult to freezerdquo said challenger Evers ldquoThe Cobo study showed for the first time in a large number of patients that you can do it and have good outcomesrdquo Moreover embryo cryopreserva-tion if broadly implemented might reduce the need to put back more than one embryo Evers added because you can freeze several and put them back in subsequent cycles if a woman doesnrsquot be-come pregnant It may even enhance pregnancy rates because the endometrial environment appears to be ldquomore friendlyrdquo during nor-mal cycles as compared to drug stimulated cycles he added

exPAnDing The BOunDARieS OF TReATMenTIn the future there may be no limit to the age at which a woman is able to bear children according to Jonathan Tilly professor and chair of biology at Northeastern University His 2012 paper provided evidence that human ovarian tissue contains pools of ovarian stem cells that develop into human oocytesmdashconfirming what researchers have found in other animals and overturning the long-held dogma positing that women are born with a fixed number of eggs that are never replenished (see page 10) Tilly first shocked the field in 2004 when he and his colleagues extracted ovarian stem cells from fe-male mice for the first time

In the recent paper Tilly and his colleagues developed a more sen-sitive method to identify and collect mouse ovarian stem cells Once the team confirmed that it had isolated the mouse ovarian stem cells by this new method it repeated the technique with frozen ovaries do-nated from sex-reassignment patients When cultured the retrieved oogonial stem cells (OSCs) turned into immature oocytes The team then tagged the cells with green fluorescent protein to trace them and injected them into pieces of human ovarian tissue which were then transplanted into mice After about two weeks the OSCs had differentiated into green-glowing cells that by all measures appeared to be human oocytes

ldquoThis was one of the most astonishing papers published in 2012rdquo said challenger Felipe Vilella from the Valencia Institute of Infertility in Spain However no one has yet shown that the human-derived OSCs can be fertilized to form an embryo Vilella said and no one knows whether any offspring would be healthy Although studies on animal derived OSCs have produced offspring no data on the health of the offspring has been provided said Vilella Nevertheless re-search on monkeys and ultimately humans is in the planning stages

Tilly said If the research pans out it would give women with ovarian cancer the chance to have children later in life It would also open the possibility that women well into their 50rsquos and even older could bear biological childrenmdashthe ethics of which ldquoare not for me to deciderdquo concluded Tilly

Vilellarsquos questions about the health of offspring raised another im-portant point of debate whether any of the procedures used in IVF currently or in the future might unwittingly be harming the result-ing offspring At the meeting Michael Skinner director of the center for reproductive biology at Washington State University in Pullman Washington presented his serendipitous finding revealed in his 2005 paper on the epigenetic transgenerational effects of endocrine disruptor chemicals on rodents (see page 18) If a pregnant female rodent was exposed to endocrine disrupting chemicals during a window when the fetusrsquos sex cells were developing the chemical exposure caused epigenetic alterationsmdashchanges in methylation patterns in the genomemdashin sperm These alterations were passed down through at least four generations and reduced male fertility Skinner has since tested several additional chemicals and found the same or similar effects on sperm and that these epigenetic changes can cause disease in subsequent generations In one study [ReprodToxicol34708 (2012)] for example Skinner showed that pregnant female rats exposed to the pesticide mixture DEETmdashcommonly found in insect repellantsmdashgave birth to pups with higher rates of disease that included pubertal abnormalities testis disease and ovarian disease Disease susceptibility was predominantly passed via males and persisted for three generations

A similar issue has been raised before in regard to IVF tech-niques Could the egg or culture media be exerting an effect on the epigenetics of the developing embryo To date rigorous studies

have not yet been done to answer this question But the debate certainly gen-erated enough interest and ideas for future studies

WRAP uPAfter the meeting participants lauded the challenge and debate format ldquoFrom my perspective this type of format should be broadly appliedrdquo said Tilly ldquoItrsquos an awesome ideardquo Scott added ldquoand Irsquom not prone to hyperbolerdquo Some however expressed a desire for more probing questions ldquoWe were all being a little too nicerdquo Reijo Pera said But everyone agreed that in contrast to large conferences where there is little time for discussion and questions this one enabled much more opportunity to learn Such feedback will be impor-tant in organizing the second Top Ten in reproductive medicine conference which is already in the works Accord-ing to Simoacuten the next one will likely take place in four yearsmdashenough time to publish new papers and to measure whether the first conference had an im-pact on spreading ideas and improving the way research in reproductive medi-cine is done

everyone

agreed that in

contrast to large

conferences

where there is

little time for

discussion and

questions this

one enabled

much more

opportunity to

learn

Produced by the ScienceAAAS Custom Publishing Office

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

4 5

In September 2013 scientists and cli-nicians in the field of human reproduc-tion participated in a groundbreaking conference held in Florence Italy Over the course of two days authors of the 10 most cited viewed or downloaded research articles in the field presented their papers Each author was confront-ed by a ldquochallengerrdquo after which par-ticipants discussed the paperrsquos findings and significance The goal to advance rigorous science in a field dominated by small-scale studies and trial and error ldquoEveryone is aware of the clini-cal relevance of assisted reproduction technology [ART]rdquo said Carlos Simoacuten professor of obstetrics and gynecology at the University of Valencia Spain and a conference scientific organizer ldquoWe need to increase our scientific rel-evancerdquo

In 2009 the World Health Organiza-tion officially recognized infertility as a disease Not all countries however have followed suit Consequently large-scale funding of fertility treatments has been scarce Regulation differs among countries making adoption of standard practices difficult And on the basic re-search front strict embryo protection laws in some countries restrict studies on human embryos ldquoIf you look at all the things we do in the clinic I would bet half of them have never been prov-en efficient in multicenter trialsrdquo said Hans Evers a professor in the depart-ment of obstetrics and gynecology at Maastricht University Medical Centre in Maastricht The Netherlands Intracyto-plasmic Sperm Injection (ICSI) whereby sperm is injected directly into an egg

for example is now being used to treat unexplained infertility in about half of all treatment cycles said Evers ldquoBut no one has ever shown that ICSI is necessary in unexplained infertilityrdquo

ICSI was an advance initially developed to help men with poor sperm quality become fathers when traditional in vitro fertilization failed The past 10 years have seen additional advances in technol-ogy aimed at helping greater numbers of people become parents too Several of the most promising of these were discussed at the conference Embryo cryopreservation for example may make in vi-tro fertilization (IVF) safer and less stressful by enabling a woman to undergo ovarian stimulation only once to retrieve eggs (See Cobo page 25) And embryo selection technology that can distinguish healthy embryos from unhealthy ones promises to raise success rates Further in the future it will be possible to diagnose any number of gene mutations in an embryo and harvest eggs from ovarian stem cells so that older couples can have biological children (see Legro page 35 and Lamb page 40)

However the best application of these technologies should be de-bated before they come into widespread use said Evers Moreover some technologies pose ethical questions that require societal input For example while diagnostic tests can detect the presence of a gene the tests may be meaningless since genes donrsquot predict with 100 certainty whether a person will actually develop a disease Also if women beyond normal reproductive age can indeed one day have biological children then how old is too old ldquoYou have to think about these things before the possibility arisesrdquo said Evers

To start addressing these issues The ldquoTop Ten in Reproductive Medicinerdquo conference highlighted the best papers as an example of how research into human reproduction and infertility should be done The format inspired by Simoacuten and developed by the Serono Symposia International Foundation (SSIF) was the first of its kind in the field By discussing each papermdashfive clinical and five basic researchmdashat length conference organizers hope to promote good clinical practice that comes from adopting technologies and prac-tices that are backed by strong scientific evidence and that have been validated by ethical debate

ASSiSTeD RePRODuCTiOn TeChnOlOgieS We CAn DO BeTTeR IVF has changed little since Robert Edwards and Patrick Steptoe introduced the technique over 35 years ago While there have been

advances in the field such as the ability to freeze sperm or eggs and test embryos for specific ge-netic abnormalities the process remains much the same Fertil-ize egg with sperm transfer the resulting embryo into a womanrsquos uterus and hope for a full-term pregnancy The technique has brought five million babies into the world and fulfilled the longing of parenthood for couples struggling with infertility

But IVF has a downside The drugs given to stimulate the ovaries to produce multiple eggs can overstimulate the ovaries and land wom-en in the hospital when the clinical condition is particularly risky The procedure is stressful primarily because itrsquos hard to predict whether it will work Depending on the age of the eggs the clinic and the procedure used any embryo transfer cycle has between a 25 and 40 chance of resulting in a baby With the odds in favor of failure doctors have typically transferred more than one embryo to boost the chance of success While the rate of triplets and higher order births has come down drastically the number of twins born from reproduc-tive technologies remains at about 30 of all ART births in the US A twin or multiple pregnancy has long been known to raise the risk of harm to both mother and babies Maternal complications include increased rates of pre-eclampsia gestational diabetes and preterm labor Risks for babies include low birth weight and prematurity as well as a higher risk of long-term disability and death as compared with singleton pregnancies

WiTh neW TeChnOlOgieS SuCCeSS iS MORe likelyAt the conference Christina Bergh from the Sahlgrenska University Hospital in Gothenburg Sweden presented her 2004 paper that for the first time provided evidence that transferring a single embryo yields almost the same birth rate as transferring two in women un-der age 36 (see page 23) Berghrsquos paper was the first clinical paper presented and set the backdrop for the presentations that followed It began with the question How can we treat infertility and inflict the least harm The paper was among the most highly viewed over the last 10 years primarily because it was published when rates of mul-tiple births using ART were much higher than they are today Since the paper was published single embryo transfer (SET) has been

widely adopted particularly in Northern Europe In Sweden doctors use SET in 75 to 80 of all IVF cycles in women of any age said Bergh But there re-main several countries in which it hasnrsquot caught on she added and this is likely because of dif-ferences in reimbursement prac-

tices for reproductive procedures Bergh presented additional data showing that SET works well in

women up to age 42 Additionally she studied children born from IVF procedures and showed that at the same time the rate of complica-tions had also decreased drastically ldquoHer work has really pushed the SET idea worldwiderdquo said her challenger Robert Fischer medical director of Fertility Center in Hamburg Germany and SSIF Scien-tific Committee Member However Swedish statistics may not hold for other populations said a clinician from the audience Patients in Scandinavia opt for IVF treatments much more quickly than patients in southern Europe he added Consequently IVF patients in south-ern Europe tend to be older and have poorer outcomes In such a setting SET may not be as successful Another participant pointed out that for SET to really take off success rates would need to be much higher

eMBRyO SeleCTiOn AnD FReezingOne way to raise success rates would be to distinguish good em-bryos from bad ones Enter Richard Scottrsquos 2010 paper on detect-ing aneuploidy in an embryo using whole genome amplification and single nucleotide polymorphism arrays (see page 27) Scott showed that the technology can distinguish abnormal from normal embryos with a 4 error rate Since publication additional data shows that the error rate is actually only around 1 said Scott director of repro-ductive science at Robert Wood Johnson Medical School in Basking Ridge New Jersey The paper was the first of a series of four that together show that the technology gives clinically meaningful results Scott said

For example in one study Scott and his colleagues biopsied em-bryos and transferred them back into women as part of their infertility care before they knew the biopsy results Once they determined the success or failure of the procedure they then compared the biopsy

The ldquoTop Ten in

Reproductive

Medicinerdquo

conference

highlighted the

best papers as

an example of

how research

on human

reproduction

and infertility

should be done

ldquoTop Tenrdquo Conference Promotes Good Science in Human Reproduction

Produced by the ScienceAAAS Custom Publishing Office

76

analysis data with birth outcomes and showed that the DNA array technology could detect with 99 accuracy an abnormal embryo that was unlikely to develop into a baby Another study showed that selecting embryos using this technology does indeed yield higher birth rates The predictive power of the technology is so great that the Reproductive Medicine Associates of New Jersey fertility clinic of which Scott is the director now performs SET using the diagnostic technology in 70 of all IVF cycles he said The clinic also offers the diagnostic service to other clinics

The technology isnrsquot easy to implement however and is expen-sive compared with other technologies on the market challenger Carlos Simoacuten pointed out and no one has yet demonstrated that the array technology yields better clinical outcomes as compared with those alternatives that are cheaper and easier to use Discus-sion also raised the point that array technologies may be rendered obsolete in the near future by next generation technologies that can sequence genes faster and cheaper In fact there are already com-panies selling tests that purport to predict the chance of a child in-heriting certain traits based on sequencing specific parental genes Adapting such tests to screen embryos for specific traits would be a simple taskmdashthe prospect of which has raised more than a few eyebrows Nevertheless technologies to parse normal embryos from abnormal ones are a boon to the field because they raise suc-cess rates Scottrsquos clinic is seeing pregnancy rates of greater than 50 he said and preliminary data suggests that embryo selec-tion is bringing preterm births rates down to a level on par with the general population

Microarrays arenrsquot the only promising embryo selection technol-ogy available During the meeting Renee Reijo Pera a professor in the department of obstetrics and gynecology at Stanford Univer-sity School of Medicine presented her 2010 paper on noninvasive imaging of human embryos (see page 15) The paper showed that time-lapse image analysis of two-day old zygotes together with gene expression profiling could with 93 accuracy predict which embryos

would develop to the blastocyst stage Three param-eters captured by imaging appeared to be the

most predictive duration of the first cell division the time interval between the

end of first mitosis and initiation of the second (mitosis is the replica-

tion of chromosomes that pre-cedes cell division) and the

time interval between the second and third mitoses ldquoThis is an example of out of the box thinkingrdquo said challenger Simoacuten However he pointed out there is a lack of randomized clinical trial data showing that embryo selection using

this technology results in higher birth rates Discus-

sion also raised the issue of a lack of standards with

groups trying to replicate the results using different measurement techniques

In yet another advance embryo freezing would make IVF safer and finally make it possible for young women suffering from ovarian cancer and premature ovarian failure to have families later in life Ana Cobo an embryologist from the Infertility Institute in Valencia presented her 2008 paper comparing fresh eggs to those cryopre-served by Cryotopmdasha commercially available method to vitrify hu-man tissues (see page 25) Vitrification is the conversion of a water containing solution to a glass-like solid that is free from the ice crys-tals normally formed during freezing She and her colleagues found that cryopreserved eggs were just as readily fertilized and able to develop into blastocysts as fresh ones Cobo also presented recent data showing that the pregnancy rate from cryopreserved embryos is similar to that of fresh embryos

ldquoThe paper got a lot of interest because oocytes have been ex-tremely difficult to freezerdquo said challenger Evers ldquoThe Cobo study showed for the first time in a large number of patients that you can do it and have good outcomesrdquo Moreover embryo cryopreserva-tion if broadly implemented might reduce the need to put back more than one embryo Evers added because you can freeze several and put them back in subsequent cycles if a woman doesnrsquot be-come pregnant It may even enhance pregnancy rates because the endometrial environment appears to be ldquomore friendlyrdquo during nor-mal cycles as compared to drug stimulated cycles he added

exPAnDing The BOunDARieS OF TReATMenTIn the future there may be no limit to the age at which a woman is able to bear children according to Jonathan Tilly professor and chair of biology at Northeastern University His 2012 paper provided evidence that human ovarian tissue contains pools of ovarian stem cells that develop into human oocytesmdashconfirming what researchers have found in other animals and overturning the long-held dogma positing that women are born with a fixed number of eggs that are never replenished (see page 10) Tilly first shocked the field in 2004 when he and his colleagues extracted ovarian stem cells from fe-male mice for the first time

In the recent paper Tilly and his colleagues developed a more sen-sitive method to identify and collect mouse ovarian stem cells Once the team confirmed that it had isolated the mouse ovarian stem cells by this new method it repeated the technique with frozen ovaries do-nated from sex-reassignment patients When cultured the retrieved oogonial stem cells (OSCs) turned into immature oocytes The team then tagged the cells with green fluorescent protein to trace them and injected them into pieces of human ovarian tissue which were then transplanted into mice After about two weeks the OSCs had differentiated into green-glowing cells that by all measures appeared to be human oocytes

ldquoThis was one of the most astonishing papers published in 2012rdquo said challenger Felipe Vilella from the Valencia Institute of Infertility in Spain However no one has yet shown that the human-derived OSCs can be fertilized to form an embryo Vilella said and no one knows whether any offspring would be healthy Although studies on animal derived OSCs have produced offspring no data on the health of the offspring has been provided said Vilella Nevertheless re-search on monkeys and ultimately humans is in the planning stages

Tilly said If the research pans out it would give women with ovarian cancer the chance to have children later in life It would also open the possibility that women well into their 50rsquos and even older could bear biological childrenmdashthe ethics of which ldquoare not for me to deciderdquo concluded Tilly

Vilellarsquos questions about the health of offspring raised another im-portant point of debate whether any of the procedures used in IVF currently or in the future might unwittingly be harming the result-ing offspring At the meeting Michael Skinner director of the center for reproductive biology at Washington State University in Pullman Washington presented his serendipitous finding revealed in his 2005 paper on the epigenetic transgenerational effects of endocrine disruptor chemicals on rodents (see page 18) If a pregnant female rodent was exposed to endocrine disrupting chemicals during a window when the fetusrsquos sex cells were developing the chemical exposure caused epigenetic alterationsmdashchanges in methylation patterns in the genomemdashin sperm These alterations were passed down through at least four generations and reduced male fertility Skinner has since tested several additional chemicals and found the same or similar effects on sperm and that these epigenetic changes can cause disease in subsequent generations In one study [ReprodToxicol34708 (2012)] for example Skinner showed that pregnant female rats exposed to the pesticide mixture DEETmdashcommonly found in insect repellantsmdashgave birth to pups with higher rates of disease that included pubertal abnormalities testis disease and ovarian disease Disease susceptibility was predominantly passed via males and persisted for three generations

A similar issue has been raised before in regard to IVF tech-niques Could the egg or culture media be exerting an effect on the epigenetics of the developing embryo To date rigorous studies

have not yet been done to answer this question But the debate certainly gen-erated enough interest and ideas for future studies

WRAP uPAfter the meeting participants lauded the challenge and debate format ldquoFrom my perspective this type of format should be broadly appliedrdquo said Tilly ldquoItrsquos an awesome ideardquo Scott added ldquoand Irsquom not prone to hyperbolerdquo Some however expressed a desire for more probing questions ldquoWe were all being a little too nicerdquo Reijo Pera said But everyone agreed that in contrast to large conferences where there is little time for discussion and questions this one enabled much more opportunity to learn Such feedback will be impor-tant in organizing the second Top Ten in reproductive medicine conference which is already in the works Accord-ing to Simoacuten the next one will likely take place in four yearsmdashenough time to publish new papers and to measure whether the first conference had an im-pact on spreading ideas and improving the way research in reproductive medi-cine is done

everyone

agreed that in

contrast to large

conferences

where there is

little time for

discussion and

questions this

one enabled

much more

opportunity to

learn

Produced by the ScienceAAAS Custom Publishing Office

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

76

analysis data with birth outcomes and showed that the DNA array technology could detect with 99 accuracy an abnormal embryo that was unlikely to develop into a baby Another study showed that selecting embryos using this technology does indeed yield higher birth rates The predictive power of the technology is so great that the Reproductive Medicine Associates of New Jersey fertility clinic of which Scott is the director now performs SET using the diagnostic technology in 70 of all IVF cycles he said The clinic also offers the diagnostic service to other clinics

The technology isnrsquot easy to implement however and is expen-sive compared with other technologies on the market challenger Carlos Simoacuten pointed out and no one has yet demonstrated that the array technology yields better clinical outcomes as compared with those alternatives that are cheaper and easier to use Discus-sion also raised the point that array technologies may be rendered obsolete in the near future by next generation technologies that can sequence genes faster and cheaper In fact there are already com-panies selling tests that purport to predict the chance of a child in-heriting certain traits based on sequencing specific parental genes Adapting such tests to screen embryos for specific traits would be a simple taskmdashthe prospect of which has raised more than a few eyebrows Nevertheless technologies to parse normal embryos from abnormal ones are a boon to the field because they raise suc-cess rates Scottrsquos clinic is seeing pregnancy rates of greater than 50 he said and preliminary data suggests that embryo selec-tion is bringing preterm births rates down to a level on par with the general population

Microarrays arenrsquot the only promising embryo selection technol-ogy available During the meeting Renee Reijo Pera a professor in the department of obstetrics and gynecology at Stanford Univer-sity School of Medicine presented her 2010 paper on noninvasive imaging of human embryos (see page 15) The paper showed that time-lapse image analysis of two-day old zygotes together with gene expression profiling could with 93 accuracy predict which embryos

would develop to the blastocyst stage Three param-eters captured by imaging appeared to be the

most predictive duration of the first cell division the time interval between the

end of first mitosis and initiation of the second (mitosis is the replica-

tion of chromosomes that pre-cedes cell division) and the

time interval between the second and third mitoses ldquoThis is an example of out of the box thinkingrdquo said challenger Simoacuten However he pointed out there is a lack of randomized clinical trial data showing that embryo selection using

this technology results in higher birth rates Discus-

sion also raised the issue of a lack of standards with

groups trying to replicate the results using different measurement techniques

In yet another advance embryo freezing would make IVF safer and finally make it possible for young women suffering from ovarian cancer and premature ovarian failure to have families later in life Ana Cobo an embryologist from the Infertility Institute in Valencia presented her 2008 paper comparing fresh eggs to those cryopre-served by Cryotopmdasha commercially available method to vitrify hu-man tissues (see page 25) Vitrification is the conversion of a water containing solution to a glass-like solid that is free from the ice crys-tals normally formed during freezing She and her colleagues found that cryopreserved eggs were just as readily fertilized and able to develop into blastocysts as fresh ones Cobo also presented recent data showing that the pregnancy rate from cryopreserved embryos is similar to that of fresh embryos

ldquoThe paper got a lot of interest because oocytes have been ex-tremely difficult to freezerdquo said challenger Evers ldquoThe Cobo study showed for the first time in a large number of patients that you can do it and have good outcomesrdquo Moreover embryo cryopreserva-tion if broadly implemented might reduce the need to put back more than one embryo Evers added because you can freeze several and put them back in subsequent cycles if a woman doesnrsquot be-come pregnant It may even enhance pregnancy rates because the endometrial environment appears to be ldquomore friendlyrdquo during nor-mal cycles as compared to drug stimulated cycles he added

exPAnDing The BOunDARieS OF TReATMenTIn the future there may be no limit to the age at which a woman is able to bear children according to Jonathan Tilly professor and chair of biology at Northeastern University His 2012 paper provided evidence that human ovarian tissue contains pools of ovarian stem cells that develop into human oocytesmdashconfirming what researchers have found in other animals and overturning the long-held dogma positing that women are born with a fixed number of eggs that are never replenished (see page 10) Tilly first shocked the field in 2004 when he and his colleagues extracted ovarian stem cells from fe-male mice for the first time

In the recent paper Tilly and his colleagues developed a more sen-sitive method to identify and collect mouse ovarian stem cells Once the team confirmed that it had isolated the mouse ovarian stem cells by this new method it repeated the technique with frozen ovaries do-nated from sex-reassignment patients When cultured the retrieved oogonial stem cells (OSCs) turned into immature oocytes The team then tagged the cells with green fluorescent protein to trace them and injected them into pieces of human ovarian tissue which were then transplanted into mice After about two weeks the OSCs had differentiated into green-glowing cells that by all measures appeared to be human oocytes

ldquoThis was one of the most astonishing papers published in 2012rdquo said challenger Felipe Vilella from the Valencia Institute of Infertility in Spain However no one has yet shown that the human-derived OSCs can be fertilized to form an embryo Vilella said and no one knows whether any offspring would be healthy Although studies on animal derived OSCs have produced offspring no data on the health of the offspring has been provided said Vilella Nevertheless re-search on monkeys and ultimately humans is in the planning stages

Tilly said If the research pans out it would give women with ovarian cancer the chance to have children later in life It would also open the possibility that women well into their 50rsquos and even older could bear biological childrenmdashthe ethics of which ldquoare not for me to deciderdquo concluded Tilly

Vilellarsquos questions about the health of offspring raised another im-portant point of debate whether any of the procedures used in IVF currently or in the future might unwittingly be harming the result-ing offspring At the meeting Michael Skinner director of the center for reproductive biology at Washington State University in Pullman Washington presented his serendipitous finding revealed in his 2005 paper on the epigenetic transgenerational effects of endocrine disruptor chemicals on rodents (see page 18) If a pregnant female rodent was exposed to endocrine disrupting chemicals during a window when the fetusrsquos sex cells were developing the chemical exposure caused epigenetic alterationsmdashchanges in methylation patterns in the genomemdashin sperm These alterations were passed down through at least four generations and reduced male fertility Skinner has since tested several additional chemicals and found the same or similar effects on sperm and that these epigenetic changes can cause disease in subsequent generations In one study [ReprodToxicol34708 (2012)] for example Skinner showed that pregnant female rats exposed to the pesticide mixture DEETmdashcommonly found in insect repellantsmdashgave birth to pups with higher rates of disease that included pubertal abnormalities testis disease and ovarian disease Disease susceptibility was predominantly passed via males and persisted for three generations

A similar issue has been raised before in regard to IVF tech-niques Could the egg or culture media be exerting an effect on the epigenetics of the developing embryo To date rigorous studies

have not yet been done to answer this question But the debate certainly gen-erated enough interest and ideas for future studies

WRAP uPAfter the meeting participants lauded the challenge and debate format ldquoFrom my perspective this type of format should be broadly appliedrdquo said Tilly ldquoItrsquos an awesome ideardquo Scott added ldquoand Irsquom not prone to hyperbolerdquo Some however expressed a desire for more probing questions ldquoWe were all being a little too nicerdquo Reijo Pera said But everyone agreed that in contrast to large conferences where there is little time for discussion and questions this one enabled much more opportunity to learn Such feedback will be impor-tant in organizing the second Top Ten in reproductive medicine conference which is already in the works Accord-ing to Simoacuten the next one will likely take place in four yearsmdashenough time to publish new papers and to measure whether the first conference had an im-pact on spreading ideas and improving the way research in reproductive medi-cine is done

everyone

agreed that in

contrast to large

conferences

where there is

little time for

discussion and

questions this

one enabled

much more

opportunity to

learn

Produced by the ScienceAAAS Custom Publishing Office

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

8 9

The 1960rsquos through the 1970rsquos was a period of great change in reproductive medicine Knowledge about reproduc-tion increased dramatically Before in vitro fertilization (IVF) gynecology had little more to offer infertile couples than making a diagnosis and offering tender loving care (1) When the efforts of the two pioneers Robert G Edwards and Patrick C Steptoe resulted in the birth of the first IVF baby in 1978 the world slowly started to understand the insur-gency that these two icons had caused in the medical field Not without open antagonism from some Edwards be-came acclaimed as the father of IVF and his discoveries were belatedly rec-ognized by the awarding of the Nobel Prize in Physiology or Medicine in 2010 By then several millions of babies had been conceived through IVF and the term assisted reproductive technology (ART) was coined to include other pro-cedures that improved infertility treat-ment Edwardsrsquo work laid the founda-tion for further exciting developments in reproductive medicine such as preim-plantation genetic diagnosis (PGD)

ART is focused on bringing better gametes into close proximity in order to generate a viable embryo which can then be placed in a receptive endometri-um to generate a healthy full-term preg-nancy The greatest progress has been made in improving embryo viability and increasing implantation rates which un-intentionally has led to the main draw-backmdashthe unacceptably high incidence of multiple births The rate of multiples has been substantially reduced by the introduction of elective single embryo

HISTORY AND PERSPECTIVE Reproductive Medicine Before and After the NobelJLH Evers and Antonio Pellicer

transfer (eSET) one of the few ART developments that has been introduced based on robust clinical data (2) Regulatory initiatives for eSET approval have followed around the world However there is still much to be done before eSET is universally accepted

An ART process begins with controlled ovarian stimulation (COS) the manipulation of the reproductive physiology to generate sev-eral mature oocytes simultaneously in a single cycle In the early days of ART this was a necessary functional step because the techniques for fertilization and culture were poor so several eggs needed to be fertilized to obtain just a few viable embryos Today the procedures used in ART labs are much more sophisticated so there is a diminished need for aggressive COS The field is now moving towards more gentle and patient-friendly stimulation proto-cols which will produce only an adequate number of embryos and no more (3)

in ViTRO SCReeningSeveral methods for performing genetic screens on embryos both invasive and noninvasive are in the research pipeline right now Af-ter some initial controversy regarding the biopsy and screening of embryos for a limited number of chromosomes the introduction of more accurate technologies that provide more detailed information on all chromosomesmdashsuch as single nucleotide polymorphism ar-raysmdashis looking promising especially for use in older patients (4) In vitro embryo observation utilizing time-lapse technology (5) and the analysis of proteins and metabolites consumed and secreted by the developing embryo in culture are other noninvasive methods that hold promise

The same molecular techniques are applied during PGD in order to detect those embryos carrying particular disease-associated muta-tions The list of diseases is steadily increasing and in the future it will become possible to diagnose multiple disorders simultaneously in a single embryo shifting the frontiers in human health care These advances will however also raise important ethical questions that will need to be addressed (6) such as how to contend with the partial expression or limited penetrance of certain diseases that might only manifest in later life

CRyOPReSeRVATiOnAlthough the failure of embryo implantation was a concern in the early days of IVF and a valid justification for the use of multiple

embryos the problem was compounded by the poor performance of embryo cryopreservation techniques However vitrificationmdashintroduced as a method of cryopreservation in the 1980rsquosmdashcompletely changed the field for both embryo and oocyte freezing (7) Today survival rates after freezingthawing of human oocytes and embryos are greater than 90 bringing about new possibilities for ART while also reducing the need for ethically questionable selective abortions

Vitrification has also been applied as a means to preserve fertility in young women with cancer allowing for the freezing of oocytes prior to chemotherapy or radiation treatment Meanwhile fertility preservation has more recently been introduced to avoid the delete-rious effects of aging on the oocytes (lsquosocial freezingrsquo) Since age correlates positively with aneuploidies an increasing number of women are requesting oocyte freezing before age 38 in an attempt to abrogate this effect (8)

Two other complications of COS deserve our attention an altered endometrial receptivity that decreases the chance of implantation and the development of ovarian hyperstimulation syndrome (OHSS) a rare but potentially life-threatening condition A new two-step ART approach that involves the implantation of previ-ously frozen embryos in subsequent natural cycles will potentially lead to a safer and more patient-friendly ART approach with fewer complications (910)

The enDOMeTRiAl enViROnMenTMost attention in ART is focused on the embryo but endometrial receptivity is an issue that is often neglected predominantly because insufficient methods exist to ascertain the functional status of the uterine lining In todayrsquos ˈomics era however new tools are coming to the fore that will allow the best embryo(s) to be placed in a fully receptive endometrium (11)

Despite numerous advances in reproductive technologies we still have very limited possibilities for addressing the main cause of infertility namely the womanrsquos age The next two de-cades will see research in developing methods to rejuvenate oo-

cytes by exploring the cellular and molecular hallmarks of aging and attempting to at least partially reverse them (12) Similarly the creation of gametes by cell reprogramming could be another source of healthy oocytes (and spermatozoa) an advance that will certainly change our perspective on infertility in the coming years (1314)

Most of these potential advancements in ART are still based on small observational clinical ldquoproof-of-conceptrdquo studies They de-mand to be followed up with more comprehensive multicenter randomized clinical trials Subfertility couples belong to a vulner-able group and should not be exploited (15) We should provide them with dedicated care and evidence-based treatment During the next decade the medical establishment needs to provide more robust evidence for the value and integrity of the treatments cur-rently offered to patients As Robert Edwards our very own No-bel Laureate stated ldquothe legacy to future generations is a matter of liferdquo (16)

Several methods

for performing

genetic screens

on embryos both

invasive and

non-invasive are

in the research

pipeline right

now

Robert Edwards holds Louise Brown the first baby conceived through in vivo fertilization as Patrick Steptoe looks on July 25 1978

REFERENCES 1 J L H Evers Lancet 360151 (2002) 2 A Thurin et al N Engl J Med 351 2392 (2004) 3 R G Edwards R Lobo P Bouchard Hum Reprod 11 91

(1996) 4 N R Treff et al Fertil Steril 94 2017 (2010) 5 C C Wong et al Nat Biotechnol 28 1115 (2010) 6 JC Harper et al Hum Reprod Update 18 234 (2012) 7 A Cobo et al Fertil Steril 89 1657 (2008) 8 JA Garcia-Velasco et al Fertil Steril 99 1467 (2013)

9 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)10 P Devroey NP Polyzos C Blockeel Hum Reprod 26 2593 (2011)11 P Diacuteaz-Gimeno et al Fertil Steril 95 50 (2011)12 C Loacutepez-Otiacuten M A Blasco L Partridge M Serrano G Kroemer Cell 153 1194 (2013)13 YA White et al Nat Med 18 413 (2012)14 K Hayashi et al Science 338 971 (2012)15 A W Nap J L H Evers Hum Reprod 22 3262 (2007)16 R G Edwards P C Steptoe A Matter of Life The Story of IVF - a Medical Breakthrough (Hutchinson amp Co London 1980)

Produced by the ScienceAAAS Custom Publishing Office

JLH Evers

Antonio Pellicer

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

10 11

Germline Stem Cells in Adult Mammalian OvariesDori C Woods and Jonathan L Tilly

inTRODuCTiOnUntil recently a decades-old belief in the field of mammalian repro-ductive biology was that females are born with a finite endowment of oocytes termed the lsquoovarian reserversquo which is progressively deplet-ed throughout life by either ovulation or atresia (1) Once exhausted the ovaries cease to function In women this event heralds the onset of menopause (2) Although the traditional explanation for the ces-sation of ovarian functionmdashnamely that a woman simply runs out of oocytes fits the model of the ovarian reserve being set forth as a nonrenewable resource at birth a number of recent studies indicate that this paradigm is probably incorrect In its place has emerged a model that aligns gametogenesis in mammals more closely with that observed in non-mammalian species in that adult ovaries actively support formation of new oocytes and follicles (3ndash14) The underly-ing foundation of this major paradigm shift is rooted in the develop-ment of detailed methods for the isolation and characterization of a rare population of mitotically active germ cells from adult ovaries termed female germline stem cells (fGSCs) or oogonial stem cells (OSCs) (46812) Once isolated these cells can be stably propa-gated in vitro for months without loss of their ability to differentiate into oocytes following either defined culture in vitro or transplantation into ovarian tissue in vivo (457814) In mice oocytes formed from transplanted OSCs complete maturation to the metaphase-II stage of development and these eggs can be fertilized to produce viable embryos and offspring (47812)

While intraovarian transplantation models clearly show that mam-malian OSCs are capable of generating fully functional eggs the role of these cells if any in adult ovaries under normal physiological con-ditions has not yet been reported In non-mammalian vertebrates such as the teleost Medaka (Oryziaslatipes) genetic-lineagendashtrac-ing approaches have been successfully employed to show that fG-SCs actively contribute to the adult oocyte pool used for reproduction (15) Development and analysis of similar genetic models in mice which are currently under investigation in the Tilly laboratory should provide a means to map the rates of new oocyte formation during postnatal life and the contribution of these oocytes to offspring pro-duction In addition the ability to track a lsquomarkedrsquo pool of oocytes formed at a specific point in life will facilitate studies of in vivo oocyte lifespan to determine how long a given oocyte once generated per-sists in adult ovaries This information will help elucidate if the quality of eggs deteriorates in women with age simply because all of the oo-cytes have been sitting around for decades accumulating damage or if the decline in egg quality is instead tied to a progressive impair-ment in the ability of OSCs to generate competent oocytes with age as is the case in the aging Drosophila ovary (16)

ChARACTeRizATiOn OF MOuSe AnD huMAn OSCSAlthough the beginnings of contemporary evidence for the existence of OSCs and postnatal oogenesis date back nearly a decade (3) the recent development of protocols that allow for the enrichment or purification of OSCs from adult ovaries has greatly accelerated research in this area (46812) Building on earlier observations that mouse OSCs expose the C-terminus of the germ cell-specific protein Ddx4 [DEAD box polypeptide 4 also commonly referred to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell sur-face (4) we developed and validated an antibody-based fluores-cence-activated cell sorting (FACS) approach that purifies OSCs based on detection of this extracellular epitope of Ddx4 (extracel-lular Ddx4- or ecDdx4-positive cells) (8 12 13) The reason why Ddx4 is exposed on the surface of OSCs but is completely inter-nalized in oocytes is unknown but may be related to movement of the protein from a potentially sequestered transmembrane location to a more functionally active position in the cytoplasm as primitive germ cells undergo meiotic differentiation (13) Of note similar OSC isolation strategies have been reported using antibodies against the externalized domain of the primitive germ cell marker Ifitm3 (interferon-induced transmembrane protein 3 also referred to as Fragilis) (612)

Following isolation and expansion of OSCs in vitro the cells can be genetically modified to express a traceable gene (such as green fluorescent protein GFP) and returned to the ovaries of recipient wild type mice where they actively undergo differentiation into oo-cytes that mature ovulate and fertilize to produce viable embryos and offspring (47812) Although this type of cell fate-tracking ex-periment is not feasible in humans we have successfully employed a mouse xenograft model to establish the oocyte-forming capabili-ties of human OSCs expressing GFP in adult human ovarian tis-sue in an in vivo environment (812) The ensuing observations that human OSCs form what appear to be meiotically arrested oocytes [positive for expression of Y-Box protein 2 (YBX2) which is spe-cifically expressed in germ cells at the diplotene stage of meiosis (17 18)] contained in follicles within one to two weeks of grafting not only provides definitive evidence that adult human ovaries are fully capable of supporting oogenesis and folliculogenesis but also that oocyte and follicle formation from purified OSCs returned to their natural environment are events that can occur in a fairly rapid time- frame (812)

COnTROVeRSiAl neW FinDingSAlthough independent verification of the existence of OSCs in adult mouse ovaries is now available from several labs (4ndash812ndash14) two new studiesmdashboth based on circumstantial negative findings with micemdashclaim otherwise despite the fact that neither study actually attempted to reproduce what we and others have done or to iso-late OSCs from mouse ovaries using published protocols employed successfully by us and others The first of these from Liu and col-

leagues relied entirely on a transgenic reporter mouse generated by crossing Ddx4-Cre mice with Rosa26rbw+ mice on the assump-tion that OSCs could be specifically tracked due to Ddx4-driven Cre activation in these cells (as would be the case with all germ cells irrespective of their differentiation status) (19) Unfortunately an ab-sence of many key control experiments discussed in detail recently (12) precludes any clear interpretation of the findings In fact in as-yet unpublished studies we have since evaluated ovaries from the same genetic mouse line and combined this approach with our FACS-based protocol for OSC isolation to highlight the erroneous nature of their primary conclusion that in contrast to substantial evi-dence from us and others (3ndash1420) mitotically active germ cells do not exist in postnatal mouse ovaries

The second paper relies heavily on two key assumptions (21) The first is that oogenesis in embryonic and adult ovaries is ac-complished through identical processes In fetal ovaries primordial germ cells proliferate and form cyst-like clusters prior to meiosis (22) Upon meiotic commitment the cyst-like clusters associate with so-matic cells forming the ovigerous cords that will break down shortly after birth to produce the initial primordial follicle pool (2324) In this new study Lei and Spradling propose a model with no supporting

data that relies on formation of germ cell cysts as an indispensable step in postnatal oogenesis as well When they failed to observe evidence of cyst-like germ cell clusters in adult mouse ovaries the authors concluded that oogenesis is not occurring (21) Another as-sumption relates to the ldquolineage tracing modelrdquo used in which both Cre-recombinase and estrogen receptor are expressed in essentially all cells of the mice under study Short-term induction with Tamoxifen excises a stop-cassette located within a Rosa26-yellowfluorescentprotein (YFP) transgene yielding indiscriminate labeling in which all cell types are lsquomarkedrsquo with YFP at very low frequency (1ndash 2) Since all germ cells including OSCs and oocytes express YFP at equivalent frequencies it is impossible to discern if any labeled oocytes originated from labeled OSCs or if any unlabeled oocytes originated from unlabeled OSCs In fact the authors ac-knowledge identification of ldquoa few germ cells at a prefollicular state of developmentrdquo but do not discuss the meaning of such findings (21) Since oocytes are incapable of surviving in ovaries unless surrounded by granulosa cells as follicles the rare ldquoprefollicularrdquo germ cells identified by Lei and Spradling are likely OSCs or their immediate progeny

Whatever the case assuming 2 of OSCs are YFP labeled in their

ThisarticlebasedontheoriginalpublicationY A R White et al Nature Med 18 413 (2012)

Department of Biology Northeastern University Boston MA Corresponding Authors dwoodsneuedu or jtillyneuedu

Figure 1 Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs) Ovarian cortical tissue from women is dissociated into single cell suspension and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8 12) Purified OSCs are established in culture and expanded ex vivo to assess egg maturation potential or the rate of oogenesis To evaluate egg formation OSCs are transformed to express a reporter (ie GFP) and directly injected into human ovarian cortical strips which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles Growing follicles with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28ndash30) To evaluate oogenesis rates human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene which is activated during meiotic commitment in germ cells The cells are then screened for factors that activate DsRed expression Positive hits representing candidate meiotic (oogenic) activators are then assessed with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12 14) If a given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro) it is then tested for its ability to increase oocyte-containing primordial follicle numbers in adult mouse ovaries in vivo

Produced by the ScienceAAAS Custom Publishing Office

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

12 13

model one would expect the ratio of labeled and unlabeled oocytes comprising the primordial follicle pool to remain constant over time since a fixed proportion of labeled and unlabeled oocytes would be constantly entering the pool through de novo oogenesis from a chi-meric population of OSCs and subsequently exiting the pool through follicle growth activation This is exactly what Lei and Spradling ob-serve (21) It bears mention that a toxicity model was also employed to assess if OSC activation and follicle renewal in postnatal ova-ries occurs in response to damage When such evidence was not produced these negative findings were offered as further proof that OSCs do not exist Surprisingly however the cytotoxic drug used was busulfan which is a widely known GSC toxicant (325ndash27) Ac-cordingly it is unclear why one would expect to find damage-induced activation of a population of cells that are killed by the agent used to lsquoactivatersquo them

nexT STePS AnD huMAn heAlTh iMPliCATiOnSAs work with laboratory animal models continues to fill in key gaps in our understanding of mammalian OSCs the discovery of oocyte progenitor cells in human ovaries opens up many opportunities for their utilization some from the perspective of the clinical manage-ment of infertility and others from that of basic biomedical discovery (89) Given the remarkable similarities between mouse and human OSCs and in particular the ability of OSCs from both species to participate in new oocyte and follicle formation when delivered back into adult ovarian tissue (8 12) it would stand to reason that as observed in mice (47812) human OSCs also have the potential to generate developmentally competent eggs However because such functional testing of human OSCs in vivo is not possible strate-gies to mature human eggs derived from OSCs entirely ex vivo will be needed In this regard Telfer and McLaughlin have reported a multistep culture system in which microthin human ovarian cortical strips are maintained under defined serum-free conditions to initi-ate primordial follicle growth activation (2829) Once the preantral stage of development is reached the follicles are dissected out and subsequently cultured individually until the enclosed oocytes can be aspirated for in vitro maturation to metaphase-II eggs Should this methodology prove successful (30) it may provide an opportunity to test the developmental potential of human OSCs to form functional eggs (Fig 1) given that human OSCs have already been shown to generate primordial oocytes contained in follicles in adult human ovarian cortical strips (812)

In addition to the ability to test the competency of oocytes derived from OSCs recombined with somatic cells and other support neces-sary for ensuring ex vivo development the innate oocyte-forming capability of these cells is valuable for other reasons During serial passage of pure OSC cultures (ie with no somatic or feeder cells) under defined conditions a small subset of OSCs spontaneously ini-tiate a differentiation process that results in the formation of what appear to oocytes (812) Although these oocytes are not function-ally competent (having never been associated with nor instructed by granulosa cells) they can be used as a powerful bioassay to rapidly decipher the factors and signaling pathways that guide oogenesis (Fig 1) This can be achieved by assessing the rate of formation of in vitro-derived oocytes from a fixed number of OSCs exposed in different wells to various agents (1214) In addition to the scientific

value of such knowledge large-scale screening of cultured OSCs may facilitate identification and development of therapeutics to sus-tain or restore the ovarian reserve in women of advancing maternal age or after exposure to noxious insults such as chemotherapy Al-though more work is clearly needed to test these ideas at this stage we believe it is appropriate to at minimum end further debate over whether mammalian OSCs exist and instead constructively focus on what additional experiments are needed to more clearly define the regulation and function of these newly discovered cells in female reproductive biology

REFERENCES 1 S Zuckerman Rec Prog Horm Res 6 63 (1951) 2 H Buckler J Br Menopause Soc 11 61 (2005) 3 J Johnson et al Nature 428 145 (2004) 4 K Zou et al Nat Cell Biol 11 631 (2009) 5 J Pacchiarotti et al Differentiation 79 159 (2010) 6 K Zou et al Stem Cells Dev 20 2197 (2011) 7 Y Zhang et al J Mol Cell Biol 3 132 (2011) 8 Y A R White et al Nat Med 18 413 (2012) 9 D C Woods J L Tilly Fertil Steril 98 3 (2012)10 D C Woods Y A R White J L Tilly Reprod Sci 20 7 (2013)11 D C Woods J L Tilly Semin Reprod Med 31 24 (2013)12 D C Woods J L Tilly Nat Protoc 8 966 (2013)13 A N Imudia et al Fertil Steril 100 1451 (2013)14 E S Park D C Woods J L Tilly Fertil Steril 100 1468 (2013)15 S Nakamura et al Science 328 1561 (2010)16 R Zhao et al Aging Cell 7 344 (2008)17 W Gu et al Biol Reprod 59 1266 (1998)18 J Yang et al Proc Natl Acad Sci USA 102 5755 (2005)19 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)20 Y Hu et al Cell Prolif 45 287 (2012)21 L Lei A C Spradling Proc Natl Acad Sci USA 110 8585 (2013)22 M Pepling A Spradling Development 125 3323 (1998) 23 M Pepling A Spradling Dev Biol 234 339 (2001)24 C Nicholas K Haston R Reijo-Pera BMC Dev Biol 10 2 (2010)25 L R Bucci M L Meistrich Mutat Res 176 259 (1987)26 M C Pelloux et al Acta Endocrinol 118 218 (1988)27 C J Brinster et al Biol Reprod 69 412 (2003)28 E E Telfer et al Hum Reprod 23 1151 (2008)29 E E Telfer M McLaughlin Semin Reprod Med 29 15 (2011)30 C E Dunlop E E Telfer R A Anderson Maturitas doi101016j maturitas201304017 (2013)

Acknowledgments We thank Cooper Graphics (wwwcooper247com) for preparation of the final figure Work conducted by the authors discussed herein was supported by grants from the United States National Institutes of Health (R37-AG012279 R21-HD072280 F32-AG034809) the Henry and Vivian Rosenberg Philanthropic Fund and the Glenn Foun-dation for Medical Research

Competing Interests Statement DCW is a scientific consultant for OvaScience Inc (Cambridge MA) JLT discloses interest in intellectual property described in US Patent 7195775 US Patent 7850984 and US Patent 7955846 and is a co-founder of OvaScience

Bidirectional Communication Between the Oocyte and Surrounding Somatic Cells is Required for Successful Follicular Development and Oocyte MaturationJia Peng12 John J Eppig3 Martin M Matzuk124567

Follicular development and oocyte maturation are essential pro-cesses for successful ovulation to produce competent eggs ready for fertilization Historically it has been unclear whether the oocyte or the surrounding granulosa cells orchestrate the rate of follicular growth Elegant work from the last decade shed considerable light on this process Studies utilizing chimeric reaggregated mouse ovaries consisting of half-grown 12-day-old oocytes and newborn granulosa cells revealed that these stage-mismatched ovaries could undergo follicular development from primary to antral follicle stage at twice the normal rate However recent studies also uncovered the importance of ovarian somatic cells in the regulation of folliculogen-esis and oogenesis Here we discuss bidirectionalcommunication between oocytes and their companion somatic cells during follicular development and oocyte maturation which ensures normal female fertility

inTRODuCTiOnIn mammals primordial germ cells (PGCs) divide and migrate to the germinal ridge to become oogonia which begin meiotic divi-sion during prenatal life Around the time of birth a cohort of oo-cytes recruits a single layer of flattened pre-granulosa cells to form the primordial follicles (1) Folliculogenesis starts with activation of a primordial follicle which progresses through primary secondary and antral follicle stages and finishes by either generating a fertiliz-able oocyte or follicular atresia It was not clear previously that the rate of ovarian follicular development was dominantly controlled by either oocytes or neighboring somatic cells until a key study was done in the last decade (2) In this study mid-sized oocytes from the secondary follicles of 12-day-old mice were isolated and combined with somatic cells from primordial follicles of newborns to produce reaggregated ovaries As a result the 12-day-old oocyte prompt-ed the proliferation and differentiation of both cumulus and mural granulosa cells and doubled the rate of follicular development to generate competent oocytes ready for fertilization Thus this study demonstrated that a developmental program intrinsic to the oocyte is crucial for orchestrating the rate of follicular growth and funda-mentally changed our understanding of the regulation of ovarian follicular development

OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing gROWTh FACTOR β (TgF-β) PAThWAyIn follicular development the mammalian oocyte modulates granu-losa (directly) and thecal cell (indirectly) proliferation and differen-tiation rates through multiple oocyte-secreted factors (3) Among those factors growth differentiation factor 9 (GDF9) and bone mor-phogenetic protein 15 (BMP15) are well known as two of the most important paracrine factors to prompt primary follicle growth as well as subsequent participation in the various stages of folliculogenesis (4) GDF9 and BMP15 are close paralogs in the TGF-β superfamily and signal via the downstream P-SMAD23 or P-SMAD158 path-way respectively to stimulate proliferation and differentiation of granulosa cells Animal models deficient in these two ligands have been used to study their in vivo functions at the early stages of fol-licular development between poly- and mono-ovulatory mammals (Table 1) In mice Gdf9 null females are sterile due to an arrest of follicular growth at the primary follicle stage (5) Bmp15 null mice exhibit later defects in ovulation and fertilization rates which lead to reduced litter size (6) In sheep homozygous BMP15 mutants ex-hibit a block in folliculogenesis at the primary follicle stage resulting

Thisarticlebasedontheoriginalpublication J J Eppig et al Proc Natl Acad Sci USA 99 2890 (2002)

Departments of 1Pathology amp Immunology 2Molecular amp Human Genetics 4Molecular and Cellular Biology and 5Pharmacology 6Center for Drug Discovery and 7Center for Reproductive Medicine Baylor College of Medicine Houston TX 3The Jackson Laboratory Bar Harbor MECorresponding Author mmatzukbcmedu

Figure 1 Bidirectional communication between oocyte and surrounding somatic cells There are three major ways of oocyte-somatic cells communication in follicular development and oocyte maturation (i) oocyte-secreted factors GDF9 BMP15 and GDF9BMP15 heterodimer (ii) granulosa cell-derived kit ligand and its receptor on the oocyte and (iii) secondary messengers cAMP and cGMP

Produced by the ScienceAAAS Custom Publishing Office

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

14 15

in sterility similar to that of Gdf9 null mice (7) A homozygous point mutation in sheep GDF9 also results in abnormalities at early stages of follicular development (8) In humans although there is no direct evidence to prove that BMP15 and GDF9 are major causes of ovar-ian insufficiency multiple studies have found that these two genes are associated with premature ovarian failure (9ndash11) or spontane-ous dizygotic twinning (12) Therefore these genetic data indicate that both GDF9 and BMP15 play important roles in the transition between primary and secondary follicles as well as in ovulation However which of the two proteins is the dominant regulator might differ due to varying protein activities and expression patterns among species

To further resolve the functions of these two growth factors and their potential synergistic functions in later follicle developmental stages recombinant GDF9 orand BMP15 were used to treat granu-losa cells in vitro In a recent study our group purified human (h) and mouse (m) recombinant GDF9 and BMP15 homodimers and het-erodimers to define their activities in pre-ovulatory follicles (13) We showed that mGDF9 homodimer was a potent trigger of the TGF-β signaling cascade and upregulated downstream cumulus expansion-related gene expression to induce the proliferation and differentia-tion of granulosa cells while mBMP15 homodimer was inactive By contrast hGDF9 homodimer was inactive and hBMP15 homodimer had a relatively low activity compared to mGDF9 In our studies mhGDF9BMP15 heterodimers were the most biopotent ligands in the regulation of ovarian function

negATiVe OOCyTe RegulATORS in The PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAyGranulosa cell-derived kit ligand is a positive regulator stimulat-ing oocyte growth and somatic cell proliferation After binding with kit ligand the kit receptor mediates PI3K signaling cascades in the oocyte via many downstream regulators Among these regulators several key components of the PI3K pathway function as suppres-sors of follicular activation at the early stages of follicular develop-ment Phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a major negative regulator Mice lacking Pten in the oocyte exhibit premature ovarian failure due to depletion of primor-

dial follicles in early adulthood (14) FOXO3A a forkhead transcrip-tion factor is another important downstream negative effector of the PI3K pathway Phosphorylation of FOXO3A leads to its nuclear export and the regulation of genes involved in primordial follicle ac-tivation Foxo3a knockout female mice are phenotypically similar to Pten oocyte-specific knockout mice (15) These animal models could help unravel the etiology of infertility and premature ovarian failure in women and enable clinical intervention

MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn ViA gAP JunCTiOnSSynchrony between oocyte meiosis and follicular ovulation is required for female fertility Granulosa cells maintain oocyte meiotic arrest via the natriuretic peptide Cnatriuretic peptide receptor 2 (NPPCNPR2) system (16) Mural granulosa cells produce the NPPC ligand which binds to the receptor NPR2 a guanylyl cyclase in cumulus cells resulting in cyclic guanosine monophosphate (cGMP) generation cGMP diffuses from cumulus cells to the oocyte via gap junctions and then inhibits PDE3A an oocyte-specific phosphodiesterase to maintain elevated cyclic adenosine monophosphate (cAMP) in the oocyte (1718) High levels of cAMP require the orphan Gs-linked receptor GPR3 to prevent precocious gonadotropin-independent resumption of meiosis (19) After the LH surge PDE3A becomes activated decreasing the oocyte cAMP concentration and thus trig-gering the downstream pathways to resume meiosis during ovulation (20) As a dominant factor the oocyte controls meiotic arrest not only by maintaining high cAMP levels but also by secreting certain para-crine growth factors (including GDF9 and BMP15) to elevate NPR2 expression in cumulus cells (16) Furthermore a novel oocyte pro-tein named meiosis arrest female 1 (MARF1) has been identified as an oocyte maturation regulator by recent ENU mutagenesis screen-ing (21) Mutations of Marf1 leads to infertility characterized by the failure of meiotic resumption upregulation of protein phosphatase 2 catalytic subunit beta (PPP2CB) and an increase in DNA damage in the ovulated oocytes

COnCluSiOnS AnD iMPliCATiOnSIn summary genetic data generated over the past few decades

An Introduction to Human Embryo DevelopmentRenee A Reijo-Pera

Human embryo development begins with an oocyte-to-embryo tran-sition a process that includes the fusion of the egg and sperm migra-tion of the germ cell pronuclei into the uterus pronuclear membrane breakdown and a series of cleavage divisions This culminates in the activation of the unique human embryonic genome compaction of the blastomeres to form a morula and subsequent differentiation of the first cell lineagesmdashthe trophectoderm and inner cell mass (1) In humans there is a minor wave of transcriptional activation early in development and a major wave that constitutes embryonic genome activation (EGA) on day three of embryogenesis (2) Human embryo development is remarkable in that the oocyte provides the major-ity of the required resources to direct the complex developmental pathways during the first three days of development in the absence of large-scale transcriptional activity (2) Consider further that native

Figure 1 LAMIN-B1 (green) and CENP-A (orange) expression in a DAPI-stained (blue) cleavage-stage human embryo by confocal microscopy reveals chromosome-containing micronu-clei in the blastomeres and cellular fragments of human embryos Image from (6)

Thisarticlebasedontheoriginalpublication C C Wong et al Nature Biotech 28 1115 (2010)Institute for Stem Cell Biology amp Regenerative Medicine Department of Obstetrics and Gynecology Stanford University School of Medicine Stanford CAreneerstanfordedu

have identified three major pathways that mediate the communica-tion between oocyte and somatic cells (Fig 1) (i) oocyte-secreted GDF9 BMP15 and GDF9BMP15 heterodimer which stimulate TGFβ signaling in granulosa cells (ii) granulosa cell-derived kit ligand that binds to kit receptor on the oocyte and triggers down-stream PI3K signaling and (iii) secondary messengers which are transferred via oocyte-cumulus cell gap junctions Although the oo-cyte is the dominant factor orchestrating the onset of folliculogen-esis and regulating somatic cell proliferation and differentiation dur-ing follicular development other regulators in the oocyte-somatic cell regulatory loop are also indispensable for normal female fer-tility (22) Thus progression through successive stages of fol-licular development and oocyte maturation requires bidirectional communication between the oocyte and surrounding somatic cells

REFERENCES 1 H Peters Acta Endocrinol (Copenh) 62 98 (1969) 2 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002) 3 P G Knight C Glister Reproduction 132 191 (2006) 4 J A Elvin C Yan M M Matzuk Mol Cell Endocrinol 159 1 (2000) 5 J Dong et al Nature 383 531 (1996) 6 C Yan et al Mol Endocrinol 15 854 (2001)

7 S M Galloway et al Nat Genet 25 279 (2000) 8 J P Hanrahan et al Biol Reprod 70 900 (2004) 9 T T Wang et al Hum Reprod 28 2473 (2013)10 E Di Pasquale P Beck-Peccoz L Persani Am J Hum Genet 75 106 (2004)11 H Dixit et al Hum Genet 119 408 (2006)12 G W Montgomery et al Twin Res 7 548 (2004)13 J Peng et al Proc Natl Acad Sci USA 110 E776 (2013)14 P Reddy et al Science 319 611 (2008)15 D H Castrillon L Miao R Kollipara J W Horner R A DePinho Science 301 215 (2003)16 M Zhang Y Q Su K Sugiura G Xia J J Eppig Science 330 366 (2010)17 S Vaccari J L Weeks 2nd M Hsieh F S Menniti M Conti Biol Reprod 81 595 (2009)18 R P Norris et al Development 136 1869 (2009)19 L M Mehlmann et al Science 306 1947 (2004)20 F J Richard A Tsafriri M Conti Biol Reprod 65 1444 (2001)21 Y Q Su et al Science 335 1496 (2012)22 M M Matzuk K H Burns M M Viveiros J J Eppig Science 296 2178 (2002)

Acknowledgments Studies on oocyte-somatic cell bidirectional communication have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through R01 grants HD33438 (MMM) and HD23839 (JJE)

Species Gene Mutant Phenotype

MouseGdf9

Heterozygous Normal fertility (5)Homozygous Sterility due to block at primary follicle stage (5)

Bmp15Heterozygous Normal fertility (6)Homozygous Subfertility due to defects in cumulus expansion (6)

SheepGDF9

Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (8)Homozygous Sterility due to block at primary follicle stage (8)

BMP15Heterozygous Increased ovulation rate resulting in increased offspring (eg dizygotic twins) (7)Homozygous Sterility due to block at primary follicle stage (7)

HumanGDF9

Heterozygous Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)Homozygous Not reported

BMP15Heterozygous Associated with premature ovarian failure (1011)Homozygous Not reported

Table 1 GDF9 and BMP15 mutants in mouse sheep and human

human reprogramming from gametic pronuclear programs to embry-onic programs involves the epigenetic remodeling of one of the most challenging sets of chromosomes to reprogram those inherited from

Produced by the ScienceAAAS Custom Publishing Office

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

16 17

analysis of single embryos and embryonic blastomeres (6) Results indicated that euploid embryos could not be accurately distinguished from those with aneuploidies when only cell cycle parameters were analyzed By adding additional confocal microscopy analysis (which included reconstruction of all blastomere karyotypes to the four-cell stage) however we concluded that the complexity in human em-bryonic aneuploidy is not simply due to errors on the meioticmitotic spindle Rather results suggested that generation of aneuploidy may also occur during interphase and can be explained by the contain-ment of a subset of missing chromosomes within cellular fragments which bud off from embryonic blastomeres and may persist or be-come reabsorbed We also noted that chromosome-containing frag-ments may arise from micronuclei or embryonic structures distinct from primary nuclei with encapsulated chromosome(s) detected only in the blastomeres of cleavage-stage human embryos (Fig 1) These findings suggest that individual human blastomere behavior is diagnostic of chromosomal status and provides the first example of the use of non-invasive imaging and automated fragmentation track-ing to distinguish euploid and aneuploid cells These studies predict

Figure 2 Automated tracking of cellular fragmentation for embryo assessment Time sequence of cumulative segment lengths in pixels for each frame of the time-lapse image analysis for three embryos was observed embryo C3 with high fragmentation embryo B2 with low fragmentation and embryo C1 with no fragmentation Note that the embryo with high fragmentation exhibits much larger cumulative length of segments than that of embryos with low or no fragmentation Image from (6)

the sperm (1) These chromosomes are highly condensed highly methylated and organized around a protamine scaffold rather than a histone scaffold Yet it is clear from several studies that native re-programming in the human oocyte-to-embryo transition succeeds in over 75 percent of embryonic blastomeres (3) moreover advances in somatic cell nuclear transfer (SCNT) procedures have revealed the robust ability of human oocytes to reprogram somatic DNA (4)

uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCTDeVelOPMenTAl FATeOver the last decade key studies in human embryo development have used time-lapse imaging of embryogenesis to elucidate the role of various cell cycle parameters and factors that may predict success or failure in the embryo reaching the blastocyst stage and beyond In 2010 Wong etal reported the use of time-lapse microscopy and gene expression profiling at the single embryo and blastomere level to document the growth of human embryos from zygote to blasto-cyst (3) Key developmental features were extracted and algorithms generated to predict success and failure in development leading up to the blastocyst stage morphological assessment was augmented by a comparison of gene expression in embryos that succeeded or failed to develop These studies indicated that the characteristics of the first cytokinesis and the first two cleavage divisions could be used as markers for a positive outcome Successful development was shown for the first time to be highly regulated following strict timing in pre-implantation development even prior to EGA In normal development degradation of maternal mRNA was shown to precede

cell autonomous EGA In contrast arrested embryos most often dis-played aberrant cytokinesis during the first cleavage divisions prior to EGA accompanied by equally aberrant gene expression in the embryo or individual blastomeres Since the studies by Wong and colleagues suggested that human embryo fate could be predicted accurately prior to EGA it was proposed that success and failure is often inherited Embryos that begin life with defective maternal pro-grams or inherit defective paternal components are likely to display aberrant cytokinesis embryo fragmentation and arrest of individual blastomeres or the entire embryo

The methods and algorithms described by Wong etal provided a platform for improved and earlier diagnosis of embryo potential to hopefully allow for the transfer of fewer embryos earlier in develop-ment during assisted reproduction procedures Subsequent reports have documented that the parameters identified are able to provide predictive power beyond the blastocyst stage to implantation and that other parameters may be added for ease of use or to adapt the technology to clinics that differ in terms of cell culture media patient base or other characteristics (5)

exTenSiOn OF STuDieS TO unDeRSTAnDing geneSiS OF AneuPlOiDyBased on the results of Wong et al we hypothesized that measure-ment and analysis of specific parameters in preimplantation human embryos could provide insight into fundamental characteristics of the embryo including ploidy We first sought to correlate normal and ab-normal development by correlating continual imaging with genetic

Figure 3 Proposed model for the origin of human embryonic aneuploidy based on fragmentation timing fragment re-sorption and underlying chromosomal abnormalities Human embryos with mei-otic errors (monosomies and trisomies) and those that appear to be triploid typically exhibit fragmentation at the one-cell stage while fragmentation is most often detected at the two-cell stage in embryos with mitotic errors We also demonstrated that missing chromosome(s) are contained within frag-ments termed ldquoembryonic micronucleirdquo and for those embryos with mitotic errors pro-pose that the embryo likely divided before these chromosomes properly aligned on the mitotic spindle The correlation between the timing of fragmentation and the type of em-bryonic aneuploidy suggests that the em-bryo can sense chromosomal abnormalities and undergoes fragmentation as a survival mechanism As development proceeds these fragments either remain or are reab-sorbed by the blastomere from which they originated or a neighboring blastomere to generate the complex human aneuploidies observed Image from (6)

that a combination of time-lapse parameter analysis along with as-sessment of blastomere fragmentation dynamics (Fig 2) may re-duce the inadvertent transfer of aneuploid embryos with implications for decreasing miscarriage risk and improving in vitro fertilization success Based on this work we have offered a model of aneuploidy development during human embryogenesis that is currently being further examined (Fig 3)

SuMMARyNumerous basic and clinical laboratories are now investigat-ing the use of noninvasive time-lapse imaging to provide reli-able information regarding the development of human embryos to the blastocyst stage implantation and beyond It is hoped that advances will enable the separation of those embryos that are most likely to develop successfully from those that like-ly would result in miscarriage or still-birth due to severe birth defects

REFERENCES 1 K Niakan J Han R Pedersen C Simon R R Pera Development

139 829 (2012) 2 Z Xue et al Nature 500 593 (2013) 3 C Wong et al Nat Biotechnol 28 1115 (2010) 4 M Tachibana et al Cell 153 1228 (2013) 5 M Meseguer et al Hum Reprod 26 2658 (2011) 6 S Chavez et al Nat Commun 3 1251 (2012)

Produced by the ScienceAAAS Custom Publishing Office

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

18 19

inTRODuCTiOnWe have investigated the effects of two environmental toxicants (vinclozolin and methoxychlor) following exposure of rats during fetal gonadal sex determination and the impact on gonadal devel-opment and function (12) A ser-endipitous observation was made when F1 generation animals were mistakenly bred to generate the F2 generation offspring the vast majority of the testes in the F2 generation carried a spermato-genic cell defect that induced apoptosis This prompted a mul-tiple year study that demonstrated the phenomenon of environmen-tally induced epigenetic transgen-erational inheritance of disease for the first time (3) This study showed an increase in apoptosis in testis spermatogenic cells in the F1 F2 F3 and F4 generations as well as in outcrossed offspring through non-Mendelian genetic inheritance affecting 90 of the male population The causative epi-genetic effects were traced to specific DNA methylation sites The impact of this study on our understanding of the molecular control of inheritance disease etiology and environmental toxicology is signifi-cant Over the past ten years a series of studies have further inves-tigated this phenomenon including environmental factor specificity germline epigenetics and disease etiology (45)

ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCeThe definition of epigenetic transgenerational inheritance is ldquogerm-line mediated inheritance of epigenetic information between gen-erations that leads to phenotypic variation in the absence of direct environmental influencesrdquo (6) This is in contrast to epigenetic in-heritance that involves direct environmental germline or somatic cell exposure and epigenetic responses during early development that influence phenotypes in later life An example is the Agouti mouse

model which responds to an environmental signal in utero through a change in an epiallele thus shifting the coat color of the offspring (57) Environmental epigenetic inheritance responses may be correct-ed in subsequent generations during epigenetic reprogramming of the germline or early embryo such that the phenotype is lost (89)

In order for environmentally induced epigenetic transgenerational inheritance to occur the external insult must act on a gestating fe-male during fetal gonadal sex determination influencing the epigen-etic programming through altered DNA methylation patterns in the germline (Fig 1) The epigenetic information is then carried through the epigenome of the germline into the embryonic stem cell and thus to all somatic cells of the offspring Susceptible tissues in off-spring will potentially develop disease and the defect will continue to be transgenerationally inherited (Figs 1 and 2) (3ndash5) Several stud-ies described below support this hypothesis of the molecular etiol-ogy of epigenetic transgenerational inheritance

geRMline ePiMuTATiOnS AnD exPOSuRe (TOxiCAnT) SPeCiFiCiTyThe environmental compound (toxicant) administered in the initial study was vinclozolin one of the most widely used agricultural fun-gicides (3) The outcross information from this work indicated that

the transgenerational phenotype was transmitted through the male sperm (3) A genome-wide promoter analysis identified approxi-mately 50 differentially methylated regions (DMRs) in the sperm of the F3 generation from vinclozolin-treated animals when compared with sperm from matched control (vehicle exposed) F3 rats (10) The DMRs carry epimutations that can transmit this epigenetic informa-tion transgenerationally (4)

A critical question was whether this phenomenon was unique to vinclozolin A series of studies investigated the actions of dioxin (11) a pesticide and an insect repellent (permethrin and DEET) (12) plastics (BPH and phthalates) (13) and hydrocarbons (JP8 jet fuel) (14) All were found to promote the transgenerational inheritance of sperm epimutations and of disease (15) Interestingly each toxicant promoted a unique set of epimutations with negligible overlap (Fig 3) (15) These studies did not measure true environmental risk but provide a good foundation for future analysis to determine the envi-ronmental hazards of exposure Others have now shown transgen-erational inheritance of disease as a result of exposure to various environmental factors including nutrition (16) stress (17) and other toxicants (18) Combined observations suggest that epigenetic bio-markers for ancestral toxicant exposure and adult onset disease may exist

TRAnSgeneRATiOnAl inheRiTAnCe OF DiSeASe AnD PhenOTyPiC VARiATiOnThe first transgenerational abnormality observed was an increase in spermatogenic cell apoptosis (319) Other abnormal patholo-gies seen (Fig 1) included prostate disease (11ndash15 20 21) kid-ney disease (11ndash14 20) mammary tumors (20) immune system abnormalities (14 20) and behavioral effects such as anxiety (22) Other laboratories have shown transgenerational effects on reproduction (2324) stress response (25) and obesity (1426) Some transgenerational diseases were induced by a variety of different environmental exposures (15) including polycystic ova-ries and reduction of primordial ovarian follicle pool size which were found in the majority of females from all exposure groups examined (27)

Environmentally Induced Epigenetic Transgenerational Inheritance of DiseaseMichael K Skinner

Thisarticlebasedontheoriginalpublication M D Anway et al Science 308 1466 (2005)Center for Reproductive Biology School of Biological Sciences Washington State University Pullman WA skinnerwsuedu

Epigenetic transgenerational inheritance may also impact other ar-eas of biology One study found that the F3 generation of vinclozolin-treated rats had significant altered mate preference behavior (28) Since sexual selection is a major determinant in evolutionary biol-ogy this work suggests that transgenerational epigenetics may play a critical role in evolution (4)

TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeSAnD The eTiOlOgy OF RePRODuCTiVe DiSeASeTransgenerational germline transmission of epimutations results in all somatic cells in offspring carrying an altered methylation pattern and transcriptome (4 6) Building upon earlier transgenerational transcriptome studies in fetal rat testis (29) we examined the

Figure 1 Environmenally induced epigenetic transgenerational inheritance of disease The environmental factors such as toxicants nutrition and stress influence a gestating female during the period of fetal gonadal sex determination to then affect disease incidence (percentage) in the F1 F2 F3 and F4 generations The disease incidence for vinclozoiin lineage males compared to control lineage animals is shown for each generation The incidence of tumor development prostate dis-ease kidney disease testis disease and immune abnormalities are presented Modified from (20)

Figure 2 Role of germline in epigenetic transgenera-tional inheritance An envi-ronmental factor acts on the F0 generation gestating fe-male (left) to influence the de-veloping F1 generation fetus resulting in reprogramming of the methylation state of the primordial germ cell DNA This altered DNA methyla-tion pattern becomes fixed similar to an imprinted gene and is transferred through the germline to subsequent gen-erations Somatic cells in the offspring in later generations also carry the altered epig-enome generating an aber-rant transcriptome and pro-moting adult-onset disease Modified from (4)

Figure 3 Venn diagram of transgenerational epimutations associated with differ-ent exposure groups showing a number of common epimutations in the F3 genera-tion of rats due to ancestral exposure of F0 generation gestating females to dioxin pesticide plastics or hydrocarbons Modified from (15)

Produced by the ScienceAAAS Custom Publishing Office

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

20 21

transgenerational transcriptomes of 11 different tissues in male and female vinclozolin-treated versus control lineage rats (30) All tissues had a transgenerational transcriptome that was unique to the specific tissue with negligible overlap between tissues It is intriguing that a relatively small number of epimutations can produce such a large number of specific transcriptome changes (30) The identification of epigenetic control regionsmdashareas of 2ndash5 megabases with an overrepresentation of regulated genes close to DMRs and long noncoding RNA regionsmdashmay provide a clue (Fig 4) (30) suggesting unique molecular regulatory mechanisms that will require further investigation

To further elucidate the role of epimutations in adult onset disease the molecular etiology of ovarian diseases were studied (27) Follicu-lar somatic granulosa cells were found to have an altered epigenome and transcriptome that suggested specific signaling pathways were affected Similarly somatic Sertoli cells in the testis were found in a separate study to have transgenerational alterations in their epig-enomes and transcriptomes affecting genes previously shown to be involved in male infertility (31)

iMPACT AnD FuTuRe STuDieSOur original publication (3) described the phenomenon of envi-ronmentally induced epigenetic transgenerational inheritance of a disease Subsequent publications confirmed and clarified the mo-lecular and physiological parameters involved The research shows the existence of a non-genetic (ie epigenetic) form of transgen-erational inheritance that impacts the etiology of certain diseases as well as a potential molecular mechanism describing how envi-ronmental factors could directly influence gene expression and therefore disease (4 5) Further studies clearly are needed to clarify the role of epigenetics and transgenerational inheritance in disease etiology evolutionary biology and other areas of cell and developmental biology The specific next steps needed include in-vestigation into why specific sites are more susceptible to becom-ing transgenerationally programmed and the mechanism by which this occurs as well as the translation of this work from animals to humans These and other studies will undoubtedly have signifi-cant impact on our understanding of normal biology and disease etiology

REFERENCES 1 A S Cupp et al J Androl 24 736 (2003) 2 M Uzumcu H Suzuki M K Skinner Reprod Toxicol 18 765 (2004) 3 M D Anway A S Cupp M Uzumcu M K Skinner Science 308 1466 (2005)

4 M K Skinner M Manikkam C Guerrero-Bosagna Trends Endocrinol Metab 21 214 (2010) 5 R L Jirtle M K Skinner Nat Rev Genet 8 253 (2007) 6 M K Skinner Epigenetics 6 838 (2011) 7 M E Blewitt N K Vickaryous A Paldi H Koseki E Whitelaw PLoS Genet 2 e49 (2006) 8 R R Newbold Toxicol Appl Pharmacol 199 142 (2004) 9 R A Waterland M Travisano K G Tahiliani FASEB J 21 3380 (2007)10 C Guerrero-Bosagna M Settles B Lucker M Skinner PLoS ONE 5 e13100 (2010)11 M Manikkam R Tracey C Guerrero-Bosagna M K Skinner PLoS ONE 7 e46249 (2012)12 M Manikkam R Tracey C Guerrero-Bosagna M Skinner Reprod Toxicol 34 708 (2012)13 M Manikkam R Tracey C Guerrero-Bosagna M Skinner PLoS ONE 8 e55387 (2013)14 R Tracey M Manikkam C Guerrero-Bosagna M Skinner Reprod Toxicol 36 104 (2013)15 M Manikkam C Guerrero-Bosagna R Tracey M M Haque M K Skinner PLoS ONE 7 e31901 (2012)16 R A Waterland Horm Res 71 Suppl 1 13 (2009)17 P Jensen Prog Biophys Mol Biol (Epub ahead of print) (2013) doi 101016jpbiomolbio20130100118 V Bollati A Baccarelli Heredity (Edinb) 105 105 (2010)19 M D Anway M A Memon M Uzumcu M K Skinner J Androl 27 868 (2006)20 M D Anway C Leathers M K Skinner Endocrinol 147 5515 (2006)21 M D Anway M K Skinner Prostate 68 517 (2008)22 M K Skinner M D Anway M I Savenkova A C Gore D Crews PLoS ONE 3 e3745 (2008)23 E E Nilsson M D Anway J Stanfield M K Skinner Reproduction 135 713 (2008)24 T J Doyle J L Bowman V L Windell D J McLean K H Kim Biol Reprod 88 112 (2013)25 D Crews et al Proc Natl Acad Sci USA 109 9143 (2012)26 R Chamorro-Garcia et al Environ Health Perspect 121 359 (2013)27 E Nilsson et al PLoS ONE 7 e36129 (2012)28 D Crews et al Proc Natl Acad Sci USA 104 5942 (2007)29 M D Anway S S Rekow M K Skinner Genomics 91 30 (2008)30 M K Skinner M Manikkam M M Haque B Zhang M Savenkova Genome Biol 13 R91 (2012)31 C Guerrero-Bosagna M Savenkova M M Haque I Sadler- Riggleman M K Skinner PLoS ONE 8 e59922 (2013)

Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic PregnancyXiaofei Sun and Sudhansu K Dey

According to the US Centers for Disease Control and Prevention ectopic pregnancy is the leading cause of pregnancy-related death during the first trimester (1) In the United States around 100000 women encounter ectopic pregnancies each year and 6 of maternal deaths are attributable to ectopic pregnancies (2) Although ectopic pregnancy adversely affects female reproductive health its etiology remains elusive It was discovered that cannabinoid receptor 1 (CB1)-deficient mice demonstrated spontaneous oviductal retention of embryos at the blastocyst stage (3) making these mice a good model to study certain aspects of ectopic pregnancy

In normal pregnancy eggs are fertilized and undergo fur-ther development to embryos within the oviduct in mice or the Fallopian tubes in humans Mouse embryos develop within the oviduct for about 72 hours and then transit through the utero-tubal junction to enter the uterus The normal passage of embryos through the oviduct into the uterus requires coor-dinated oscillation of the cilia on the oviductal epithelium as well as muscle contractions Upon arrival in the uterine cav-ity embryos develop to the blastocyst stage and become activated (implantation competent) they can then implant in the uterus if the uterine environment is conducive to implantation

In ectopic pregnancies embryos implant outside of the uterine cavity in humans 98 of ectopic pregnancies occur in the Fallo-pian tubes and are known as tubal pregnancies Two prerequisites of tubal pregnancy are Fallopian tube retention of embryos and an abnormal tubal environment that is conducive to implantation A di-agnosis of tubal pregnancy often requires multiple visits to the doc-tor and there is the danger that the implanted embryo will rupture within the Fallopian tube potentially resulting in maternal death (4) Although some of the risk factors for tubal pregnancy are known such as tubal damage from surgery or infection cigarette smoking and in vitro fertilization procedures the precise mechanism by which embryo implantation in the Fallopian tube occurs is not completely understood (5)

Tubal pregnancy occurs rarely in other mammals and the lack of animal models has made it difficult to study the underlying mecha-nisms Extra-uterine implantation usually occurs in the abdominal cavity in animals and just a few cases of tubal pregnancy in primates have been reported (67) There are no known cases of full ectopic pregnancy in mice possibly because the walls of mouse oviducts are thin compared with human Fallopian tubes It is also possible that implantation-specific genes expressed in the uterus are not dis-

played in the mouse oviduct Nonetheless mouse models have been used to study certain aspects of oviductal embryo retention (8) One such rodent model CB1-deficient mice was the first to demonstrate spontaneous oviductal retention of embryos providing a useful tool to study certain mechanisms of ectopic pregnancy (9)

CB1 is a G protein-coupled cannabinoid receptor that is targeted by cannabis and endocannabinoids a group of endogenous canna-bis-like compounds that act as lipid mediators The two most exten-sively studied endocannabinoids are anandamide (AEA) and 2-ara-chidonoylglycerol (2-AG) (9) Levels of endocannabinoids in vivo are tightly regulated by enzymes controlling their synthesis and degrada-tion The magnitude and duration of AEA signaling is regulated by a membrane-bound fatty acid amide hydrolase (FAAH) which is also capable of degrading 2-AG to arachidonic acid and glycerol (9) In mice endocannabinoids and CB1 are present in the oviduct FAAH protein levels in the isthmus are lower than those in the ampullary region (1011) suggesting a gradient of AEA in the oviduct

In mice the movement of embryos within the oviducts is aided mainly by the beating of cilia located on the oviductal epithelium (1213) meanwhile the transport of embryos from the oviduct to the uterus is facilitated by a wave of oviductal muscle movement (14) Muscular movement is controlled by the peripheral sympathetic nervous system (15) Stimulation of the β2 adrenergic receptor (β2-AR) causes sphincter muscle relaxation whereas stimulation of the α1-AR produces muscle contraction It is believed that reciprocal stimulation of these two receptors causes a wave of contraction and relaxation which is conducive to the passage of embryos from the oviduct into the uterine lumen (1415)

Our group has shown that oviductal retention of embryos occurred inCB1-deficient mice (3) Reciprocal embryo transfer experiments

Thisarticlebasedontheoriginalpublication H Wang et al Nature Med 10 1074 (2004)Division of Reproductive Sciences Perinatal Institute Cincinnati Childrenrsquos Research Foundation Cincinnati OHCorresponding Author skdeycchmcorg

Figure 4 Schematic showing a 2ndash5 Mbp epigenetic control region containing multiple distal gene regulatory units (black boxes) under the control of a differentially methylated region (white triangle DMR) and a long noncoding RNA (white box lncRNA) Modified from (30)

Figure 1 Impaired oviductal embryo transport causes pregnancy loss in Cb1-- mice (A) Percentage of embryos recovered from oviducts or uteri in Cb1-- and wild-type (WT) fe-males mated with WT males on day 4 of pregnancy Oviductal retention of embryos is observed in Cb1-- females (B) A representative histological section of a day 7 pregnant Cb1-- oviduct showing a trapped blastocyst (Bl arrow) at the isthmus Bar 100 microm Mus muscularis S serosa Mu mucosa The figure is adapted from (3)

Produced by the ScienceAAAS Custom Publishing Office

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

22 23

further confirmed that maternal CB1 is critical for appropriate oviductal transport of embryos since only CB1-deficient recipients displayed oviductal retention irrespective of embryo genotype (Fig 1) (3) Our laboratory also found that higher cannabinoid signaling affects oviductal embryo transport Wild-type (WT) mice exposed to Δ9-tetrahydrocannabinol (THC the active psychoactive component of marijuana) or methanandamide (a stable AEA analog) showed similar oviductal embryo retention which was rescued by treatment with a CB1 antagonist (11) Studies using FAAH-deficient mice further confirmed that higher oviductal AEA levels disrupt normal oviductal embryo transport Taken together the results demonstrated that either higher or lower endocannabinoid signaling impairs normal wave movement through the oviduct and consequently derailed normal oviductal transportation of embryos

Our studies in mice stimulated research on ectopic pregnancy in women and many of the findings in humans were consistent with the results of the mouse studies In women with ectopic pregnan-cies the Fallopian tubes and endometria showed lower levels of CB1 compared with those in women with normal pregnancies (16) AEA levels in ectopic Fallopian tubes were significantly higher than those in normal pregnancies (17) and increased AEA lev-els were associated with low FAAH activity in ectopic Fallopian tubes In addition treatment with oleoylethanolamide an endo-cannabinoid significantly decreased cilia oscillation frequency displayed on the Fallopian tube epithelium (18) These results suggest that disturbances in cannabinoid signaling during early pregnancy could result in ectopic pregnancy in humans It would be interesting to explore whether polymorphisms in the gene en-coding the CB1 receptor are associated with ectopic pregnancy in humans

A national survey of marijuana use from 1991 to 2011 in students grades 9ndash12 showed a rise in usage of this Schedule I drug (19) Since synthetic cannabinoids appeared in 2008 the popularity among high school seniors and persons under 25 years of age has increased sharply (2021) Synthetic cannabinoids are found in illicit ldquofake marijuanardquo with street names such as ldquoSpicerdquo or ldquoK2rdquo Many synthetic cannabinoids have much higher affinities for cannabinoid receptors and are more potent compared to natural cannabis This increase in marijuana and synthetic cannabinoid use is potentially troubling when combined with the aforementioned statistic that ectopic pregnancy accounts for about 6 of all pregnancy-related deaths in the United States (2) Therefore our studies not only provide a useful animal model to study ectopic pregnancy but also

draw more public attention to the adverse effects of maternal usage of marijuana and synthetic cannabinoids

REFERENCES 1 CDC MMWR Morb Mortal Wkly Rep 44 46 (1995) 2 K W Hoover G Tao C K Kent Obstet Gynecol 115 495 (2010) 3 H Wang et al Nat Med 10 1074 (2004) 4 S Senapati K T Barnhart Fertil Steril 99 1107 (2013) 5 J L Shaw S K Dey H O Critchley A W Horne Hum Reprod Update 16 432 (2010) 6 C P Jerome A G Hendrickx Vet Pathol 19 239 (1982) 7 J K Brown A W Horne Curr Opin Obstet Gyn 23 221 (2011) 8 J Zenteno C Silva H Cardenas H B Croxatto J Reprod Fertil 86 545 (1989) 9 H Wang S K Dey M Maccarrone Endocr Rev 27 427 (2006)10 Y Guo et al J Biol Chem 280 23429 (2005)11 H Wang et al J Clin Invest 116 2122 (2006)12 S A Halbert P Y Tam R J Blandau Science 191 1052 (1976)13 S A Halbert D R Becker S E Szal Biol Reprod 40 1131 (1989)14 G R Howe D L Black J Reprod Fertil 33 425 (1973)15 R D Heilman R R Reo D W Hahn Fertil Steril 27 426 (1976)16 A W Horne et al PLoS ONE 3 e3969 (2008)17 A K Gebeh J M Willets E L Marczylo A H Taylor J C Konje J Clin Endocr Metab 97 2827 (2012)18 A K Gebeh et al J Clin Endocr Metab 98 1226 (2013)19 CDC (The National Youth Risk Behavior Survey 2011) http wwwcdcgovhealthyyouthyrbspdfus_drug_trend_yrbspdf20 ONDCP (Office of National Drug Control Policy Fact Sheets - synthetic drugs 2012) httpwwwwhitehousegovondcpondcp- fact-sheetssynthetic-drugs-k2-spice-bath-salts21 httpwwwaceporguploadedFilesACEPNews_RoomNewsMedi aResourcesMedia_Fact_SheetsSynthetic20Drugs20 Fact20 Sheet-Finalpdf

Acknowledgments We thank Serenity Curtis for her careful editing of the manuscript Work in the Dey laboratory was funded in part by the National Institute on Drug Abuse and National Institute of Child Health and Human Development at the US National Institutes of Health XS was supported by a Lalor Foundation postdoctoral fellowship in reproductive biology

The wide application of in vitro fertilization has caused a dramatic in-crease in multiple births associated with adverse neonatal outcome mainly as a result of the transfer of several embryos per cycle Ran-domized controlled trials observational studies and meta-analyses were carried out to investigate pregnancy and live-birth rates after single (SET) or double embryo transfer (DET) as well as obstetrical outcomes Results showed that the live-birth rate was significantly higher after DET than SET if only fresh cycles were compared If one additional frozenthawed SET was added to the SET group live-birth rates did not differ substantially Obstetrical outcomes were consid-erably improved with the SET strategy We therefore conclude that SET is the more desirable option for a majority of patients

inTRODuCTiOnThe rapid development of different in vitro fertilization (IVF) tech-niques has resulted in increasing pregnancy and live-birth rates per cycle In parallel a dramatic increase in multiple births has occurred Numerous publications have demonstrated higher risks for an ad-verse outcome including preterm birth (PTB lt37 weeks) low birth weight (LBW lt2500 g) and perinatal mortality for children born after IVF compared with children born after spontaneous concep-tion mainly due to the high rate of multiple births following IVF (1ndash3) Also for IVF singletons increased rates of PTB and LBW were noted (3ndash5) likely caused by inherent characteristics of the infertile couple such as the motherrsquos age and nulliparity An increase in the frequen-cy of neurological sequele has also been found strongly associated with PTB and multiple births (6) An increased rate of physical malfor-mations has been reported for children born following IVF (7)

The most important factor affecting the rate of multiple births is the number of embryos transferred during IVF Reducing the number of transferred embryos from three to two had little effect on live births but decreased the rate of multiple births the observed twin rate was however still high (8) Data from large registry studiesmdashmany per-formed in Swedenmdashon obstetrical outcomes after IVF showed an in-creased rate of severe medical problems in IVF children generating an intensive debate in Sweden among obstetricians paediatricians doctors in reproductive medicine and politicians There was a call for transferring only one embryo during IVF a strategy supported by some reports that single embryo transfer results in satisfactory pregnancy rates (9)

The aim in the trial discussed here (10) was to investigate whether a similar live-birth rate could be achieved if two embryos were trans-ferred one at a timemdashone fresh embryo and if no live birth was de-

tected one additional frozenthawed embryomdashrather than the stan-dard practice of transferring two embryos on a single occasion and whether this would decrease the multiple birth rate

MeThODSBetween 2000 and 2003 eleven clinics in Sweden Denmark and Norway participated in the trial in which 661 women under 36 years of age performing their first or second IVF cycle and having at least two good quality embryos available for transfer were randomly as-signed to two groups SET and DET The study was double blind so neither the physician nor the patient knew whether one or two embryos had been transferred The number of patients needed in each group (330) was based on the assumptions that the true live-birth rate in both groups would be 030 and the probability is 080 that the upper limit of the 95 confidence interval (CI) for the difference in live-birth rates between the groups did not exceed 010 (a=005)

MAin ReSulTSThe results showed that the live-birth rate was 128330 (388) in the SET group and 142331 (429) in the DET group (95 CI for the difference 03-116) Thus equivalence between the two groups could not be demonstrated but the live birth rate did not differ sub-stantially between SET and DET Moreover the multiple birth rate was dramatically reduced in the SET compared to the DET group 08 (1) vs 333 (47) (plt0001) Most important the neonatal out-come was considerably better for children born to the SET group with significantly lower rates of PTB (116 vs 291 p=0002) and LBW (78 vs 275 plt00001) The SET group showed less severe neonatal complications demanding neonatal care in hospital (178 vs 339 p=0003) and also a lower rate of complex mor-bidity (78 vs 243 p=00001) (11) A financial analysis indicated that the SET strategy was cost-effective the incremental cost-effec-tiveness-ratio (ICER) was euro73307 ($99089) per additional delivery in the DET alternative

FuRTheR DeVelOPMenTSeveral randomized trials and meta-analyses (12 13) have com-pared the delivery rates in SET and DET groups The studies have shown significantly higher pregnancy and delivery rates after DET compared to SET when only fresh cycles were compared Perform-ing an additional frozenthawed SET in the SET group resulted in a delivery rate comparable with that in the DET group (10) The Scan-dinavian countries particularly Sweden have played a pioneering role in reducing multiple births by introducing SET on a large scale In Sweden this strategy has resulted in no change in delivery rates for fresh or frozenthawed cycles while the rate of multiple births has decreased dramatically from 25 to around 5 The overall SET rate is 70ndash80 (Fig 1) Delivery-and multiple birth rates after SET and DET according to age are illustrated in Figure 2 A gradual increase in SET rates has been seen worldwide including

Thisarticlebasedontheoriginalpublication A Thurin et al N Engl J Med 351 2392 (2004)Department of Obstetrics and Gynaecology Institute of Clinical Sciences Sahlgrenska Academy Gothenburg University Reproductive Medicine Sahlgrenska University Hospital Gothenburg Swedenchristinaberghvgregionse

Single Embryo Transfer in IVF Live Birth Rates and Obstetrical OutcomesChristina Bergh

Produced by the ScienceAAAS Custom Publishing Office

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

24 25

PeR

Cen

T

yeAR

Scandinavia and other northern European countries such as Bel-gium and the Netherlands as well as in Australia New Zealand and Japan The United States has lagged behind in applying SET (14) Although the rate of multiple births has decreased in recent years it is still high in many countries The latest report indicates that the multiple birth rate in Europe is 22 (2008) and in the USA is 31 (2009) (1516)

RATiOnAle FOR nOT APPlying SeTFears among both patients and clinicians that SET will lower preg-nancy and live-birth rates is probably the most common reason given for not applying SET more broadly A focus on short-term higher suc-cess rates (pregnancy rate per cycle) rather than optimal outcomes ie the birth and health of a singleton has slowed progress in this area

Despite the overwhelming data indicating increased risks associ-

ated with multiple births there is still an ongoing debate whether in par-ticular twins are a de-sired outcome for IVF It has been claimed that the risks associated with IVF twins are dramatical-ly exaggerated and that twins should be encour-aged (17)

There are also financial variables for patients in-volved to be considered and large differences ex-ist in reimbursement sys-tems for IVF in different countries which influ-ences utilization of SET In countries where IVF is

completely or partially covered by public funding SET has been ac-cepted more readily If patients are self-funded they might choose more embryos to be transferred in order to maximize the chance of becoming pregnant

Other factors impacting SET acceptance include a lack of knowl-edge concerning the risks of multiple births failure to consider cumu-lative live birth rates that include frozenthawed cycles differences in legislation between countries and impediments stemming from cultural beliefs

COnCluDing ReMARkSThe most important risk associated with IVF is the higher rate of multiple births resulting in increased child morbidity and mortality In addition to problems in the neonatal period preterm birth and low birth weight may have long-term consequences for future health Ac-cording to the Barker hypothesis (18) these adverse outcomes may

Figure 1 Delivery rate multiple-birth rate and single embryo transfer rate in Sweden 2000ndash2011 (fresh cycles)

Figure 2 Percent delivery per SET and DET percent multiple birth per SET and DET and the overall SET rate according to age of the mother National data from the Swedish Quality Registry for Assisted Reproduction (wwwucruuseqivf) for 2011 (n=9317 fresh IVF cycles)

lead to increased risk of type 2 diabetes and cardiovascular diseases in adulthood

The association between IVF and a poorer obstetric outcome for singletons has also been widely documented but is quantitatively a much smaller problem Maternal characteristics seem to be the larg-est contributors to these adverse outcomes (19) while the contribu-tion of IVF-related factors is less clear

Recent data from Sweden indicates that it is possible to have two consecutive singleton births in the same or simi-lar number of fresh IVF cycles using the SET strategy as with the DET strategy by simply adding a small number of frozenthawed cycles while concomitantly avoiding the risk of multiple births (20)

Current knowledge strongly supports SET as an IVF strategy that is safe and effective for mothers and children

REFERENCES 1 T Bergh A Ericsson T Hillensjouml KG Nygren U B Wenner

holm Lancet 354 1579 (1999) 2 L A Schieve et al N Engl J Med 346 731 (2002) 3 F M Helmerhorst D A Perquin D Donker M J Keirse BMJ 328

261 (2004) 4 R A Jackson K A Gibson Y W Wu M S Croughan Obstet

Gynecol 103 551 (2004)

5 S D McDonald et al Eur J Obstet Gynecol Reprod Biol 146138 (2009) 6 B Stroumlmberg et al Lancet 359 461 (2002) 7 C Bergh U B Wennerholm Best Pract Res Clin Obstet Gynecol 26 841 (2012) 8 A Templeton J K Morris N Engl J Med 339 573 (1998) 9 S Vilska A Tiitinen C Hyden-Granskog O Hovatta Hum Reprod 14 2392 (1999)10 A Thurin et al N Engl J Med 351 2392 (2004)11 A Thurin-Kjellberg P Carlsson C Bergh Hum Reprod 21 210 (2006)12 Z Pandian S Bhattacharya O Ozturk G Serour A Templeton Cochrane Database Syst Rev 15 CD003416 (2009)13 D J McLernon et al BMJ 341 epub c6945 doi101136bmjc6945 (2010)14 A Maheshwari S Griffiths S Bhattacharya Hum Reprod Update 17 107 (2011)15 AP Ferraretti et al Hum Reprod 27 2571 (2012)16 S Sunderam et al MMWR Surveill Summ 61 1 (2012)17 N Gleicher D Barad Fertil Steril 91 2426 (2009)18 D J Barker BMJ 311 171 (1995)19 A Pinborg et al Hum Reprod Update 19 87 (2013)20 A Sazonova K Kaumlllen A Thurin-Kjellberg U BWennerholm C Bergh Fertil Steril 99 731 (2013)

Oocyte Vitrification A Major Milestone in Assisted ReproductionAna Cobo

The cryopreservation of female gametes has been pursued since the onset of assisted reproductive technology (ART) After a lengthy period of failures and sporadic successes the introduction of refined vitrification methods has meant a huge step forward in infertility prac-tice as it allows efficient egg cryostorage Today the clinical use of oocytes that have been stored in cryogenic tanks is an accepted practice The very broad scope of this technology encompasses vari-ous new areas such as fertility preservation for social or oncological reasons the possibility of creating egg banks for ovum donation and the opportunity to avoid ovarian hyperstimulation syndrome or to ac-cumulate oocytes in low-responder patients Even segmented infer-tility treatments can be offered by stimulating ovaries vitrifying em-bryos and then transferring them during a natural cycle All of these options were not available just a few years ago but are now being routinely applied in increasingly more infertility centers worldwide

inTRODuCTiOnThe development of efficient cryopreservation techniques for female gametes has been a real challenge as both freezing and thawing expose oocytes to severe stress that can result in cell death The introduction of improved vitrification methods has meant increas-ingly successful outcomes where the most commonly applied meth-odology since the onset of ARTmdashslow freezingmdashhas led to limited success (1 2) Among the reasons for failure resulting from slow freezing chilling injury and ice formation are recognized as being the most deleterious (3) Chilling injury is avoided with vitrification because cooling occurs extremely rapidly and ice formation is cir-cumvented very efficiently by using high concentrations of cryopro-tectants (CPAs) Application of vitrification has been limited since it was first employed in embryology in the early 1980s (4) because the concentration of CPAs required for the earliest vitrification protocols can have dramatic toxic effects Significant changes made to these protocols have reduced the concentration of CPAs to circumvent del-eterious consequences to cells (5) The efficiency of modified pro-tocols has been successfully tested clinically which is an essential requisite for fertility preservation ovum donations or other clinical applications and also to overcome certain clinical obstacles that ART posed such as the absence of a partnerrsquos semen sample on

Thisarticlebasedontheoriginalpublication A Cobo et al Fertil Steril 89 1657 (2008)Instituto Valenciano de Infertilidad University of Valencia Valencia Spainacoboivies

Produced by the ScienceAAAS Custom Publishing Office

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

26 27

inTRODuCTiOn Given a prevalence of one in six couples infertility is one of the largest contemporary public health issues in Western countries In vitro fertilization (IVF) is now widely available and is the most ef-fective treatment for infertile couples While clinical outcomes have improved since IVF was first introduced the efficiency of the process remains low In the most recent US Centers for Disease Control and Prevention summary of IVF outcomes in the United States few-er than 17 of embryos deemed of sufficient quality to merit transfer actually implanted and progressed to delivery This fact reflects that morphologic and temporal markers of in vitro embryonic develop-ment have only modest predictive value when trying to assess the true reproductive potential of any single embryo As a result of this uncertainly patients and IVF providers elect to transfer multiple em-bryos in the hope that one with true reproductive potential will be included While improving delivery rates unacceptable rates of mul-tiple gestations with significantly increased maternal and neonatal morbidity ensue

ACCuRATe ASSeSSMenT OF eMBRyOniCRePRODuCTiVe POTenTiAlAssessing embryonic competence brings special challenges not en-countered elsewhere in clinical science The reality is that embryos deemed to lack reproductive potential are discarded as biologic waste Thus a misdiagnosis leads embryologists to throw out an embryo which might have been capable of becoming a baby Some couples only rarely produce a competent embryo or may have lim-ited access to care and thus such a misdiagnosis may lead them to discard their only meaningful opportunity for delivery There is no other area of medicine where a single abnormal laboratory result causes clinicians or scientists to discontinue care with such irrefut-able finality

VAliDATing ASSAyS FOR eMBRyOniCAneuPlOiDy SCReeningScreening embryos for the correct number of chromosomes (euploi-dy) represents a well-founded concept given the direct correlation between increasing maternal age decreased fecundity and oocyte aneuploidy (1) Developing and validating a single cell embryonic aneuploidy assay is more complex than for other cell types in part because access to biologic material for validation is difficult and limit-ing There are no cell lines available of embryonic cells at this stage of development but discarded embryos can be used for validation

Accurate Single Cell 24 Chromosome Aneuploidy ScreeningRichard T Scott Jr12 and Nathan R Treff12

Thisarticlebasedontheoriginalpublication N R Treff et al Fertil Steril 94 2017 (2010)1Reproductive Medicine Associates of New Jersey Basking Ridge NJ2Division of Reproductive Endocrinology Department of Obstetrics Gynecology and Reproductive Sciences Robert Wood Johnson Medical School Rutgers University New Brunswick NJCorresponding Author rscottrmanjcom

It might seem logical that if test results are consistent among all the cells in an embryo then the test is valid However embryonic mo-saicism is a real phenomenon impacting approximately one in five embryos and therefore when testing embryonic cells from the same embryo some disagreement is expected When that happens does it represent an error in the assay or mosaicism A number of investi-gators have attributed these differences to mosaicism but there is no scientific rationale to preferentially attribute them to either mosaicism or laboratory

Given the critical impact that screening has on management of individual embryos the challenges of mosaicism the fact that the as-say may be required to work with only a single cell and the complex-ity in demonstrating clinical benefit validation of embryonic aneuploi-dy screening is a complex process requiring multiple studies at the basic laboratory and translationalclinical level The study reviewed herein describes the first step in the complex process of validating a toolmdashsingle nucleotide polymorphism (SNP) arraysmdashfor embryonic aneuploidy screening namely standardizing the assay

TARgeT DnA AMPliFiCATiOnSNP array technology provides an opportunity to simultaneously in-vestigate the copy number of all 24 chromosomes Unlike fluores-cence in situ hybridization (FISH)-based testing arrays require the incorporation of DNA amplification for which a variety of methodolo-gies could be employed Methods for whole genome amplification (WGA) were compared for performance on SNP arrays (2) Results demonstrated superior performance of a PCR-based strategy when evaluating copy number accuracy (most relevant to aneuploidy screening)

A FOunDATiOnAl STuDy OF SnP ARRAy SCReeningThis study was conducted in three phases (3) Single cells from a variety of cell lines with known chromosomal abnormalities were evaluated and data analysis settings were established to give the highest level of consistency This calibration was necessary to define the methodology Next the same cell lines were used to prepare and blind single cell samples for analysis Blinded analysis provided a rigorous test of the accuracy of the method on positive controls Finally multiple single cells from human embryos available for re-search were evaluated to demonstrate consistency of diagnoses in the relevant tissue type

AnalysesofCellLinesThe SNP array approach analyses the relative intensity of binding for a large number of SNPs spread across the genome Average intensi-ties are considered to have a copy number of two Those with lower intensity are assigned copy numbers of one (monosomy) and those with higher intensities are assigned copy numbers of three (trisomy) Given that embryonic aneuploidy typically involves the gain or loss of an entire chromosome the results on a single chromosome should

the day of ovum collection A comparison of embryo development using vitrified and fresh oo-

cytes was performed (6) to provide valuable information about the further potential of vitrified oocytes and to serve as a foundation for additional studies

OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTyA prospective cohort study was used to perform a first of its kind analysis of the quality of embryos emerging from simultaneously generated vitrified and fresh donor oocytes (6) All of the metaphase II oocytes assigned to a single recipient were randomly allocated one of two groups vitrified (oocytes were vitrified and then warmed for one hour before implantation) or fresh (maintained in culture at 37degC) Intracytoplasmic sperm injection (ICSI) using the husbandrsquos semen sample was performed simultaneously on both vitrified and fresh oocytes A high overall survival rate was achieved (97 N=224 of 231 oocytes) No significant differences in fertilization rates for day one (763 vs 822) day two (942 vs 978) or day three (776 vs 846) embryo cleavage or blastocyst formation rates (487 vs 475) were detected for the vitrified and fresh oocytes respectively The ratios of good quality embryos on day three (808 vs 805) and in the blastocyst stage (811 vs 700) were simi-lar in both groups Additionally promising clinical outcomes were obtained following the transfer of embryos developed from vitrified oocytes (408 for the implantation rate and 478 for the ongoing pregnancy rate) These outcomes are in contrast to those previously reported by other authors using a slow freezing methodology (1) The widespread introduction of oocyte vitrification into clinical prac-tice has been greatly strengthened by these findings which have demonstrated that both embryo developmental capability and im-plantation potential are essentially unaltered by oocyte vitrification

COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn TO CuRRenT CliniCAl PRACTiCeOur findings were further replicated by other authors with both do-nor and own oocytes [see (7) for review] In a follow up controlled randomized clinical trial we used a refined methodology to compare clinical outcomes utilizing both cryopreserved and fresh oocytes in an ovum donation program (8) In this study vitrified oocytes could not be shown to be inferior to fresh oocytes in terms of the ongoing pregnancy rate (odds ratio = 0921 95 CI 0667 to 1274) This finding definitively confirms our previous observations regarding the unimpaired potential of vitrified oocytes to develop into competent embryos capable of generating ongoing pregnancies in a proportion comparable to that of fresh oocytes

Most of the difficulties posed by fresh donations such as long wait-ing lists the need for donorrecipient synchronization and the lack of quarantine can be overcome by oocyte cryobanking Stored oocytes are available anytime they are needed significantly alleviating dona-tion logistics

The benefits of oocyte cryostorage have also been demonstrat-ed in autologous in vitro fertilization (IVF) cycles (8ndash10) leading to

high cumulative pregnancy rates (11) Rienzi and coworkers have mirrored our findings on embryo quality and clinical outcomes in a prospective randomized sibling-oocyte study conducted with infertile patients undergoing own-oocyte IVF cycles (9) The con-sistency and reliability of oocyte vitrification for this population was further demonstrated in a multicentric study (12) Addition-ally important findings regarding the number of oocytes vitrified the patientrsquos age and the developmental stage of the embryo at the time of transfer have been published and have proven useful for patient counseling (12) Oocyte vitrification has also been sug-gested to be an interesting therapeutic alternative for low respond-ers (13) improving outcomes for a group that has been very difficult to treat successfully and may even save patients the disappoint-ment of failure after IVF cycles with a poor response using fresh oocytes (14)

The extensive research carried out to date has laid the ground-work for the establishment of successful protocols for extremely ef-ficient oocyte cryopreservation These preservation programs have provided a viable alternative that avoids the downside of an unfavor-able uterine environment resulting from administration of exogenous gonadotrophins during controlled ovarian hyperstimultation (15ndash18)

The research described here has fundamentally changed the clinical landscape and strongly supports the position that oo-cyte vitrification offers advantageous clinical results in diverse populations opening up a wide range of possibilities in the field of ART

REFERENCES 1 D A Gook D H Edgar Hum Reprod Update 13 591 (2007) 2 A Cobo C Diaz Fertil Steril 96 277 (2011) 3 G Vajta M Kuwayama Theriogenology 65 236 (2006) 4 W F Rall G M Fahy Nature 313 573 (1985) 5 M Kuwayama G Vajta O Kato S P Leibo Reprod Biomed Online 11 300 (2005) 6 A Cobo et al Fertil Steril 89 1657 (2008) 7 A Cobo J A Garcia-Velasco J Domingo J Remohi A Pellicer Fertil Steril 99 1485 (2013) 8 A Cobo M Meseguer J Remohi A Pellicer Hum Reprod 25 2239 (2010) 9 L Rienzi et al Hum Reprod 25 66 (2010)10 C C Chang et al Fertil Steril 99 1891 (2013)11 F Ubaldi et al Hum Reprod 25 1199 (2010)12 L Rienzi et al Hum Reprod 27 1606 (2012)13 A Cobo N Garrido J Crespo R Jose A Pellicer Reprod Biomed Online 24 424 (2012)14 J Cohen G Grudzinskas M Johnson Reprod Biomed Online 24 375 (2012)15 B S Shapiro et al Fertil Steril 96 344 (2011)16 B S Shapiro et al Fertil Steril 96 516 (2011)17 A Maheshwari S Bhattacharya Hum Reprod 28 6 (2013)18 A Aflatoonian H Oskouian S Ahmadi L Oskouian J Assist Reprod Genet 27 357 (2010)

Produced by the ScienceAAAS Custom Publishing Office

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

28 29

all be the same The reality is that there is not uniform intensity for the SNPs on a single chromosome and individual SNPs may be assigned copy numbers of one two or three This is overcome by application of a Gaussian smoothing algorithm that evaluates the intensities amongst adjacent SNPs and averages them to enhance accuracy However the smoothing algorithms used for conventional karyotyping of the cell lines or clinical samples where hundreds of millions of cells are available do not function well following whole genome amplification of a single cell

The optimal smoothing distance was determined on cells with known karyotypes It was important to validate smoothing on chro-mosomes that were monosomic and trisomic and not just disomic Tolerances were established for the level of discordance amongst individual SNPs on a single chromosome The upper limit of discor-dance meaning the percentage of SNPs on a given chromosome that produce varying results was established for each chromosome If that level was exceeded the assay was deemed ldquonon-concurrentrdquo and the result considered unreliable The non-concurrent rate per chromosome should be lt01 and per cell should be lt2 These data were all generated on samples where the karyotype of the cell being tested was known Functionally this is starting with the answer and working backwards a critical step in the process but far from sufficient to validate the assay

In the second phase the prospective blinded evaluation the testing algorithm was applied to blinded samples The accuracy per chromosome was gt999 More importantly the overall ac-curacy of single cell aneuploidy screening using this methodology was 986

AnalysesofSingleCellsfromArrestedEmbryosIn the third phase individual cells from arrested cleavage-stage em-bryos were evaluated Given the underlying mosaicism there was no expectation that the results would be uniform However when dis-crepancies occurred it was logical that they would typically involve reciprocal errors of the same chromosomes For example a trisomy X in one cell might be accompanied by a monosomy X in another cell from the same embryo Additionally the level of non-concurrence of the SNP assignments on each chromosome should be similar to that found with the cell line samples

This prospective blinded assessment indeed found non-concur-rence rates similar to those in single cells taken from cell lines indi-cating similar accuracy levels Additionally the majority of the errors were reciprocal which was highly reassuring

This study provided the foundation for validating clinical embry-onic aneuploidy screening but represents only the first of several critical steps The predictive values of both euploid and aneuploid results cannot be determined solely in the laboratory Clinical studies assessing those predictive values as well as the overall impact on implantation delivery rates and clinical decision-making were now possible

CliniCAl STuDieS eMPOWeReD By SnP ASSAyDNA FingerprintingSNP arrays provide data on genetic polymorphisms that may be used for DNA fingerprinting (4) This provides a unique opportunity for a paired randomized controlled trial (RCT) of sibling embryos

since two embryos could be simultaneously transferred and result-ing fetuses or newborns could be precisely linked to the embryo from which they developed This has enabled powerful studies on the impact of oocyte vitrification (5) cleavage and blastocyst stage embryo biopsy (6) and other ongoing clinical trials

BenchmarkforAlternativeMethodsofCCSA validated method for comprehensive chromosome screening (CCS) provides an opportunity to test other screening methodolo-gies For example SNP arrays were used to demonstrate that FISH-based blastomere analysis is inconsistent (7) and poorly predictive of aneuploidy in the developing blastocyst (8) indicating that FISH is limited by more than just the inability to test all 24 chromosomes In addition it served as a standard when developing quantitative real-time (q)PCR (9) and next generation sequencing based meth-odologies (10) Finally SNP arrays were used to demonstrate signifi-cantly reduced concurrence of CCS by array comparative genomic hybridization compared to qPCR based on blinded analysis across multiple laboratories (11)

ClinicalTrialsValidation of the assay and DNA fingerprinting was followed by a prospective trial to determine the predictive value of euploid and an-euploid screening results for actual clinical outcome (12) Embryos selected by standard morphological criteria were biopsied and im-mediately transferred prior to any analysis of the sample SNP ar-ray predictions of aneuploidy and euploidy were compared to the actual clinical outcomes and used to directly calculate the predictive values of the assay Approximately 4 of embryos designated as aneuploid actually implanted and developed euploid fetuses Sub-sequent improvements have reduced this number to lt2 Embryos that screened euploid were more than twice as likely to implant and deliver This study (12) met a critical requirement for validating em-bryonic aneuploidy screeningmdashto directly measure the predictive values of an aneuploid result so that the risk for discarding a euploid embryo may be specifically determined

Given the demonstrated equivalence of qPCR- and SNP-arrayndashbased CCS described above SNP arrays indirectly provided an op-portunity to test qPCR clinical efficacy in subsequent RCTs The first RCT demonstrated that in women under the age of 42 with no more than one prior failed IVF cycle and a normal ovarian reserve CCS improves the success of IVF (13) A second trial further established the utility of CCS by demonstrating equivalent success rates with the transfer of a single euploid blastocyst compared to the transfer of two untested blastocysts while eliminating twins (14) and improving obstetrical outcomes (15)

ADDiTiOnAl APPliCATiOnSSNP array CCS has also been demonstrated effective at identifying sub-chromosomal imbalances associated with reciprocal transloca-tions (1617) These studies showed the importance of evaluating aneuploidy of all 24 chromosomes in parallel with analysis of the derivative chromosomes from the translocation since many other-wise balancednormal embryos were aneuploid for chromosomes not involved in the translocation

SNP arrays have also provided an opportunity to use loss of

heterozygosity to address the clinical relevance of uniparen-tal disomy [found to be extremely rare in the blastocyst (006)] to confirm the parthenogenetic status of embryos derived af-ter swapping nuclei between oocytes to eliminate mitochondrial DNA variants (18) to investigate the parental origins of aneu-ploidy using genotype comparisons (19) and to confirm the ge-netic status of human embryos derived from somatic cell nuclear transfer (20)

REFERENCES 1 T Hassold P Hunt Nat Rev Genet 2 280 (2001) 2 N R Treff J Su X Tao L E Northrop R T Scott Jr Mol Hum Reprod 17 335 (2011) 3 N R Treff J Su X Tao B Levy R T Scott Jr Fertil Steril 94 2017 (2010) 4 N R Treff et al Fertil Steril 94 477 (2010) 5 E J Forman et al Fertil Steril 98 644 (2012) 6 R T Scott Jr K M Upham E J Forman T Zhao N R Treff

Fertil Steril 100 624 (2013) 7 N R Treff et al Mol Hum Reprod 16 583 (2010) 8 L E Northrop N R Treff B Levy R T Scott Jr Mol Hum Reprod 16 590 (2010) 9 N R Treff et al Fertil Steril 97 819 (2012)10 N R Treff et al Fertil Steril 99 1377 (2013)11 A Capalbo et al 69th Annual Meeting of the American Society for Reproductive Medicine (2013) 12 R T Scott Jr et al Fertil Steril 97 870 (2012)13 R T Scott Jr et al Fertil Steril 100 697 (2013)14 E J Forman et al Fertil Steril 100 100 (2013)15 E J Forman Hong KH Scott RT Jr 61st American College of Obstetricians and Gynecologists Annual Clinical Meeting (2013)16 N R Treff et al Fertil Steril 96 e58 (2011)17 N R Treff et al Fertil Steril 95 1606 (2010)18 D Paull et al Nature 493 632 (2013)19 N R Treff et al Fertil Steril 90 S37 (2008)20 Y Chung et al Cloning Stem Cells 11 213 (2009)

What Is a ldquoNormalrdquo Semen AnalysisDavid S Guzick

Reference values for semen characteristics have been based on the distribution of values in fertile men In an effort to provide more meaningful reference values in 2001 members of the National In-stitutes of Healthrsquos (NIH) Reproductive Medicine Network (RMN) reported values for semen measurements based on a comparison of fertile and infertile men Instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnor-malrdquo data analysis yielded ranges of values that delineated three groupsmdashfertile indeterminate and subfertile These results can be more meaningfully applied in clinical practice and research than pre-vious measurements

inTRODuCTiOnDuring a standard infertility evaluation both male and female part-ners undergo a series of diagnostic tests in an effort to shed light on etiology (1) In the male partner a semen specimen is typically eval-uated for sperm concentration motility and morphology The results are presented to the patient usually with a statement about whether each of these parameters is ldquonormalrdquo

But what is ldquonormalrdquo Most clinicians refer to values published in the World Health Organization (WHO) Manual for Semen Analysis

the last edition of which was published in 2010 (2) Recognizing that there is substantial overlap in sperm parameters between fertile and infertile men (3ndash5) beginning with the 1999 edition of the WHO man-uals the nomenclature for semen characteristics was changed from ldquonormalrdquo to ldquoreferencerdquo values

Two prospective studies of semen parameters and fertility among couples who discontinued contraception published in 1998 (6) and 2000 (7) indicated that a reexamination of the WHO criteria was warranted In an effort to provide more meaningful reference values the NIH Reproductive Medicine Network (RMN) conducted a study to identify values for semen measurements that best discriminate between fertile and infertile men (8)

MeThODSIn the RMN study two semen specimens from each of the male partners in 765 infertile couples and 696 fertile couples were evalu-ated using uniform and rigorous methods The infertile couples were drawn from a randomized clinical trial in which the female partners all had normal results in an infertility evaluation (9) Fertile men (con-trols) were recruited from prenatal classes at the same hospitals in which the infertile couples were recruited Within each of nine RMN sites fertile couples were frequency matched to infertile couples ac-cording to the five-year age groups of both partners

Technicians from each of the nine RMN sites attended training sessions in semen analysis at a central laboratory Semen specimens were stained at the clinical sites and shipped to the central laboratory for assessment of sperm morphology by a single technician trained

Thisarticlebasedontheoriginalpublication D S Guzick et al N Engl J Med 345 1388 (2001)University of Florida Gainesville FLdguzickufledu

Produced by the ScienceAAAS Custom Publishing Office

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

30 31

SpermMeasurementRange OddsRatio(95CI)

MorphologicalFeatures Motility Concentration

Fertile Fertile Fertile 10

Subfertile Fertile Fertile 29 (22ndash37)

Fertile Subfertile Fertile 25 (16ndash42)

Fertile Fertile Subfertile 22 (13ndash36)

Subfertile Subfertile Fertile 72 (43ndash122)

Subfertile Fertile Subfertile 63 (38ndash103)

Fertile Subfertile Subfertile 55 (30ndash102)

Subfertile Subfertile Subfertile 158 (87ndash290)

Variable SemenMeasurement

Concentration Motility Morphology

(x106mL) () ( normal)

Fertile range gt480 gt63 gt12

Indeterminate range 135ndash480 32ndash63 9ndash12Univariate odds ratio for infertility (95 CI) 15 (12ndash18) 17 (15ndash22) 18 (14ndash24)

Subfertile range lt135 lt32 lt9

Univariate odds ratio for infertility (95 CI) 53 (33ndash83) 56 (35ndash83) 38 (30ndash50)

Table 2 Odds ratios for infertility for combinations of sperm measurements The ranges of the sperm measurements were defined by the thresholds derived from CART analysis The reference group for the odds ratios is the mean with all three measurements in the fertile range CI denotes confidence interval

Table 1 Fertile indeterminate and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility CI denotes confidence interval

by the developer of a strict morphology assessment (10) All data were then sent to a central site for data coordination

and statistical analysis Classification and regression tree (CART) analysis (11) was used to estimate thresholds for each sperm measurement that would discriminate between fertile and infertile men Two thresholds were estimated for each sperm characteristic one between the subfertile and indeterminate ranges and the other between the indeterminate and fertile ranges

ReSulTSInfertile men were slightly older than fertile controls (mean age 347 vs 335 plt0001) and were more likely to be white (910 vs 852 plt0001) and to smoke (179 vs 125 p=002) but were less likely to have completed college (55 vs 64 plt0001) As shown in Table 1 the CART-defined subfertile range for sperm mea-surements was a concentration of less than 135 million per milliliter less than 32 motility and less than 9 normal morphology Table 1 also shows CART-identified thresholds that identify the indetermi-nate and fertile ranges and associated odds ratios

Table 2 shows that the odds of infertility increase with an increasing number of sperm measurements in the subfertile range An increase

by a factor of two to three in the likelihood of infertility when one sperm measure-ment was in the subfertile range can be compared with an increase by a factor of five to seven in the risk of infertility when two sperm measurements were sub-fertile and an increase by a factor of 16 when all three measurements are subfer-tile

The frequency distribu-tions of fertile and infertile men with regard to the three sperm measurements are shown in Figure 1 Overall the degree of overlap be-tween fertile and infertile men is striking Indeed ap-proximately equal propor-tions of fertile and infertile

men had values in the indeterminate ranges of the three sperm mea-surements There was an excess of infertile men with values in the subfertile ranges of these variables and a corresponding excess of fertile men with values in the fertile ranges

The area under receiver-operating-characteristic (ROC) curves (12) which assesses the level of discrimination between fertile and infertile men was significantly greater for morphology (066) than for concentration (060 plt0001) and motility (059 plt0001)

DiSCuSSiOn AnD uPDATeA case can be made that the case-control methodology used in the RMN study is the best of an imperfect set of options Follow up of couples who have discontinued contraception has the advantage of prospectively defining couples as fertile and infertile but such an ap-proach requires large samples given the low incidence of infertility and is complicated by the unknown extent of female infertility among the couples so defined as infertile To date reference values from such a methodology have not been proposed

The WHO manual uses a statistical cut-point among fertile men defined as the 5th percentile in the last edition Such men are at the statistical tail of the normal distribution of fertile men but they are still fertile Using a population of fertile men to predict male infertil-ity makes little sense biologically or methodologically Moreover the databases from which cut-points were chosen in the first four editions were constrained by imprecisely defined reference popula-tions of fertile men and there was variability in the standards and protocols among andrology laboratories (13) Reference values de-fined in the latest edition are based on a pooled database with im-proved laboratory consistency and a better characterized group of recent fathers The 5th percentile of semen characteristics among these fertile men were sperm concentration lt15 million per millili-ter motility lt40 and normal morphology lt4

Interestingly the WHO cut-point for sperm concentration is quite

close to the cut-point found in the RMN study that separated sub-fertile from indeterminate Comparing fertile and infertile men the RMN study identified a somewhat lower threshold for motility (32 vs 40) This is reflected in Figure 1 which shows no difference between fertile and infertile men in the range of 32ndash42 motility Additionally Figure 1 shows a clear difference between fertile and infertile men in the range of 5ndash8 normal morphology which justi-fies an RMN-defined cut-point for morphology that is higher than the WHO manual

Semen characteristics are biologically continuous variables Whether they be sperm measurements such as concentration motility and morphology or sperm function tests such as zona binding assays egg penetration tests acrosome reaction testing or others no measurement has a clear cut-point that separates fertile from infertile Thus clinicians cannot convey with certainty to an infertile couple that a semen measurement below a given threshold value is responsible for their infertility nor that a value above such a threshold excludes a male factor etiology This is essentially a probabilistic matter In the RMN study to capture this idea instead of a single value for each semen measurement that would distinguish between ldquonormalrdquo and ldquoabnormalrdquo we defined the two values that best delineated three groups fertile indeterminate and subfertile

Since these values have been defined by comparing fertile and infertile men they provide ranges that can be meaningfully applied in clinical practice and research

REFERENCES 1 M A Fritz Clin Obstet Gynecol 55 692 (2012) 2 WHO Laboratory Manual for the Examination of Human Sperm- Cervical Mucus Interaction (World Health Organization Geneva ed 5 2010) 3 J MacLeod R Z Gold J Urol 66 436 (1951) 4 J MacLeod R Z Gold Fertil Steril 2 187 (1951) 5 J MacLeod R Z Gold Fertil Steril 2 394 (1951) 6 J P Bonde et al Lancet 352 1172 (1998) 7 M J Zinaman C C Brown S G Selevan E D Clegg J Androl 21 145 (2000) 8 D S Guzick et al N Engl J Med 345 1388 (2001) 9 D S Guzick et al N Engl J Med 340 177 (1999)10 T F Kruger et al Fertil Steril 49 112 (1988)11 L Breiman J H Friedman R A Olshen C J Stone Classification and regression trees (Wadsworth Belmont CA 1984)12 J A Hanley B J McNeil Radiology 143 29 (1982)13 T G Cooper et al Hum Reprod Update 16 231 (2010)

Produced by the ScienceAAAS Custom Publishing Office

Figure 1 Percentage of men from infertile and fertile couples with values in the subfertile indeterminate and fertile ranges for sperm concentration (A) percentage of motile sperm (B) and percentage of sperm with normal morphologic features (C) as defined by classification and regression tree (CART) analysis Arrows indicate the thresholds between subfertile and indeter-minate ranges (red) and indeterminate and fertile ranges (green) Adapted from (8)

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

32 33

Circulating Angiogenic Factors and the Risk of PreeclampsiaS Ananth Karumanchi12 and Ravi Thadhani3

Preeclampsia remains a major cause of maternal and fetal mor-bidity and mortality worldwide We have previously suggested that aberrant vascular endothelial growth factor signaling due to excess circulating soluble fms-like tyrosine kinase 1 plays a causal role in the pathogenesis of preeclampsia This article summarizes the key pathophysiological consequences of these angiogenic abnormalities and discusses our efforts to translate this knowledge into novel diag-nostic and therapeutic strategies

inTRODuCTiOnAs a leading cause of premature birth and maternal and infant mortality worldwide preeclampsia remains a tragically unmet pub-lic health challenge Preeclampsia and related hypertensive disor-ders conservatively cause over 75000 maternal and 500000 infant deaths globally each year (wwwpreeclampiaorg) mostly in develop-ing countries

The mechanisms that initiate preeclampsia in humans have long been elusive leading to its description in medical textbooks as an id-iopathic ldquotoxemiardquo of pregnancy whose definitive treatment remains delivery of the placenta Nearly a decade ago we proposed that el-evated plasma soluble fms-like tyrosine kinase (sFlt1) an anti-an-giogenic protein and low levels of placental growth factor (PlGF) a pro-angiogenic protein were key pathophysiological characteristics of preeclampsia (12) and showed robust correlations with disease presence and severity (3) Moreover alterations in these angiogenic factors were noted prior to onset of clinical signs or symptoms (14) In addition we demonstrated that alterations in circulating sFlt1 may explain a number of risk factors for preeclampsia such as multiple gestation trisomy 13 nulliparity and molar pregnancies (5) These findings led us to hypothesize that preeclampsia is an anti-angiogen-ic state due to excess circulating sFlt1 (6)

BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiAIn 2003 we identified upregulated sFlt1 mRNA in preeclamptic pla-centas (3) sFlt1 protein a splice variant of the vascular endothelial growth factor (VEGF) receptor Flt1 or VEGFR1 is made by the syn-cytiotrophoblast layer of the placenta and is secreted into maternal circulation (7) inhibiting VEGF signaling in the vasculature (Fig 1) Circulating sFlt1 levels are markedly increased in women with es-tablished preeclampsia while free VEGF levels are decreased (3) even prior to onset of clinical signs and symptoms (8) Furthermore removal of sFlt1 or addition of exogenous VEGF can reverse the anti-angiogenic phenotype noted in preeclamptic plasma (39) Het-

erologous expression in rodents produces a syndrome of hyperten-sion proteinuria and glomerular endotheliosis in the kidney resem-bling human preeclampsia (31011) Recombinant VEGF therapy or lowering of sFlt1 ameliorates the preeclampsia phenotype in ro-dent models (101213) Reduction of VEGF levels by 50 in the glomeruli using genetically modified mice leads to proteinuria and glomerular endothelial damage that resemble preeclampsia (14) Furthermore VEGF antagonists used in cancer patients can pro-duce a preeclampsia-like phenotype including hypertension protein-uria and cerebral edema that is often noted in severe preeclampsia (15ndash17) These data support the hypothesis that inhibition of VEGF signaling in an adult may lead to preeclampsia-like signssymptoms

The studies on sFlt1 and VEGF in preeclampsia biology have also led to new insights into basic biology of vascular and placental ho-meostasis in health and disease The microvasculature of organs af-fected in preeclampsia such as the glomeruli and hepatic sinusoidal vasculature are more permeable due to the presence of intracellu-lar perforations referred to as fenestrations While it was previously known that VEGF induces endothelial fenestrae in culture (18) ex-perimental data in VEGF knockout mice demonstrated that fenestral density is regulated by constitutive expression of VEGF (14) There-fore it is not surprising that the microvascular damage in preeclamp-sia is largely concentrated in vasculature that constitutively express VEGF and that loss of endothelial fenestrae in preeclampsia is due to excess circulating sFlt1 While the placenta is the major source of sFlt1 production recent studies have suggested that syncytial knots (degenerating syncytiotrophoblast tissue) in the placenta are the ma-jor site of sFlt1 production (19) These syncytial knots easily detach from the syncytiotrophoblast resulting in free multinucleated aggre-gates (50ndash150 mm diameter) laden with sFlt1 protein and mRNA which are secreted into the maternal circulation Studies using au-topsy material have confirmed that shed syncytial knots contribute to circulating sFlt1 in preeclampsia (20)

It is likely that other synergistic anti-angiogenic proteins such as soluble endoglin and semaphorin 3B may also contribute to the pathogenesis of preeclampsia (21 22) Patients presenting with high levels of both soluble endoglin and sFlt1 had a more severe and premature form of the disease (21 23) Animals exposed to high soluble endoglin and high sFlt1 developed the most severe features of preeclampsia including cerebral edema (2124) More research into the role of these synergistic factors in preeclampsia is needed

BiOMARkeR STuDieS in PReeClAMPSiAAlthough VEGF is secreted it acts as a juxtacrine or autocrine growth factor locally and circulating VEGF levels are relatively low during pregnancy (lt10 pgmL) Furthermore measurement of VEGF in serum is compromised by platelet VEGF release during clotting and hence circulating VEGF is not a useful biomarker for

Thisarticlebasedontheoriginalpublication R J Levine et al N Engl J Med 350 672 (2004)1Beth Israel Deaconess Medical Center Harvard Medical School Boston MA 2Howard Hughes Medical Institute Boston MA 3Massachusetts General Hospital Harvard Medical School Boston MACorresponding Author sananthbidmcharvardedu

clinical use As a surrogate of VEGF im-pairment placental growth factor levels (PlGF) in the plasma has emerged as an attractive biomarker PlGF a VEGF fam-ily member is a pro-angiogenic protein abundant during pregnancy which binds to the Flt1 receptor but not other VEGF receptors such as the kinase insert do-main receptor (KDR) As sFlt1 levels rise free PlGF levels drop and the sFlt1PlGF ratio correlates with preeclampsia phe-notypes (25 26) Working with industry partners we and others have developed highly sensitive specific and robust high throughput assays to quantitate levels of sFlt1 and free PlGF in plasma and serum (27ndash29)

Several groups have demonstrated that sFlt1 and PlGF levels can be used to differentiate preeclampsia from dis-eases that mimic preeclampsia such as chronic hypertension gestational hypertension kidney disease and ges-tational thrombocytopenia (30) We led a large prospective study that dem-onstrated that the plasma sFlt1PlGF ratio at presentation predicts adverse maternal and perinatal outcomes (oc-curring within two weeks) in the pre-term setting This ratio alone outperformed standard clinical diag-nostic measures including blood pressure proteinuria uric acid and others Importantly sFlt1PlGF levels in the triage setting cor-related with preterm delivery an observation that has now been confirmed by several groups (31ndash33) In addition we demon-strated that women diagnosed with preeclampsia but with a nor-mal angiogenic profile are not at risk for adverse maternal or fetal outcomes (34)

TheRAPeuTiC STuDieSHuman and animal studies as outlined above have strongly sug-gested that targeting the sFlt1 pathway may be a viable strategy to prevent or treat preeclampsia Below are some strategies that have been evaluated in either preclinical or early clinical studies

TherapeuticApheresissFlt1 has a large volume of distribution and circulating plasma lev-els represent less than 20 of the total body sFlt1 burden (35) We hypothesized that a selective adsorption column such as dextran sulfate (a polyanionic derivative of dextran) could augment sFlt1 re-moval Dextran sulfate binds sFlt1 efficiently (36) and these columns have been used previously to bind apolipoprotein B-containing lipo-protein LDL in pregnant patients with familial homozygous hypercho-lesterolemia without adverse consequences to mother or fetus (37) In a proof-of-concept study we demonstrated that a ~30 reduction in circulating sFlt1 levels using dextran sulfate apheresis may be sufficient to ameliorate preeclamptic signs and symptoms and pro-

long pregnancy duration We have successfully extended (in a single arm unblinded study) three preeclamptic pregnancies by two to four weeks all of which resulted in healthy deliveries with no neonatal or maternal morbidity (36) During treatment sFlt1 levels were lowered by ~30 on average indicating that modestly lowering circulating sFlt1 levels may be sufficient to successfully prolong preeclamptic pregnancies

RelaxinandStatinsExperimental studies suggest that the hormone relaxin a natu-rally occurring ovarian hormone may be used to ameliorate pre-eclampsia by improving vascular compliance and upregulating locally produced VEGF (38) A phase I study to test the safety of human relaxin in women with preeclampsia is ongoing (39) Com-pounds that upregulate pro-angiogenic factors such as statins also have been demonstrated to reverse preeclampsia in animal models (11) A clinical trial to test the safety of statins in women with established preeclampsia has been initiated in the United Kingdom (40)

SuMMARyDuring the past decade epidemiological experimental and thera-peutic studies from several laboratories have provided compelling evidence that an anti-angiogenic state due to excess circulating sFlt1 and related angiogenic proteins leads to preeclampsia Further work on the regulation of sFlt1 production may improve our understanding of the fundamental molecular defect in preeclampsia We anticipate

Figure 1 Schematic of Impaired VEGF Signaling During Preeclampsia During normal pregnancy vascular homeostasis is maintained by physiological levels of VEGF signaling in the vasculature In preeclampsia excess placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF receptors (Flt1 and KDR)

Produced by the ScienceAAAS Custom Publishing Office

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

34 35

that in the ensuing decade these molecular discoveries will lead to improvement of care of patients suffering from preeclampsia and its devastating complications

REFERENCES 1 R J Levine et al N Engl J Med 350 672 (2004) 2 R Thadhani et al J Clin Endocrinol Metab 89 770 (2004) 3 S E Maynard et al J Clin Invest 111 649 (2003) 4 R Romero et al J Matern Fetal Neonatal Med 21 9 (2008) 5 C E Powe R J Levine S A Karumanchi Circulation 123 2856 (2011) 6 I Agarwal S A Karumanchi Pregnancy Hypertens 1 17 (2011) 7 D E Clark et al Biol Reprod 59 1540 (1998) 8 B M Polliotti et al Obstet Gynecol 101 1266 (2003) 9 S Ahmad A Ahmed Circ Res 95 884 (2004)10 A Bergmann et al J Cell Mol Med 14 1857 (2010)11 K Kumasawa et al Proc Natl Acad Sci USA 108 1451 (2011)12 J S Gilbert et al Hypertension 55 380 (2010)13 Z Li et al Hypertension 50 686 (2007)14 V Eremina et al J Clin Invest 111 707 (2003)15 V Eremina et al N Engl J Med 358 1129 (2008)16 P Glusker L Recht B Lane N Engl J Med 354 980 (2006)17 T V Patel et al J Natl Cancer Inst 100 282 (2008)18 W G Roberts G E Palade J Cell Sci 108 (Pt 6) 2369 (1995)19 A Rajakumar et al Hypertension 59 256 (2012)20 A J Buurma et al Hypertension 62 608 (2013)21 S Venkatesha et al Nat Med 12 642 (2006)22 Y Zhou et al J Clin Invest 123 2862 (2013)23 R J Levine et al N Engl J Med 355 992 (2006)

24 A S Maharaj et al J Exp Med 205 491 (2008)25 R J Levine et al JAMA 293 77 (2005)26 M Noori A E Donald A Angelakopoulou A D Hingorani D J Williams Circulation 122 478 (2010)27 S J Benton et al Am J Obstet Gynecol 205 469 e461 (2011)28 S Sunderji et al Am J Obstet Gynecol 202 40 e41 (2010)29 S Verlohren et al Am J Obstet Gynecol 202 161 e161 (2010)30 H Hagmann R Thadhani T Benzing S A Karumanchi H Stepan Clin Chem 58 837 (2012)31 T Chaiworapongsa et al J Matern Fetal Neonatal Med Epub ahead of print (2013) doi 10310914767058201380690532 A G Moore et al J Matern Fetal Neonatal Med 25 2651 (2012)33 S Rana et al Circulation 125 911 (2012)34 S Rana et al Hypertens Pregnancy 32 189 (2013)35 F T Wu M O Stefanini F Mac Gabhann A S Popel PLoS ONE 4 e5108 (2009)36 R Thadhani et al Circulation 124 940 (2011)37 R Klingel B Gohlen A Schwarting F Himmelsbach R Straube Ther Apher Dial 7 359 (2003)38 J T McGuane et al Hypertension 57 1151 (2011)39 E Unemori B Sibai S L Teichman Ann NY Acad Sci 1160 381 (2009)40 A Ahmed Thromb Res 127 Suppl 3 S72 (2011)

Disclosures SAK is co-inventor of multiple commercial patents related to diagnosistreatment of preeclampsia that have been licensed to multiple companies SAK has served as a consultant to Roche Beckman Coulter and Siemens Diagnostics and has financial interest in Aggamin LLC RT is co-inventor on patents related to licensed biomarkers in preeclampsia RT also discloses financial interest in Aggamin LLC

Functional ovaries are necessary in mammalian females for propaga-tion of the species and maintenance of the hormone milieu Through-out most of the postnatal life of a female oocytesmdashthe female germ cellsmdashare encased in somatic cells in structures called follicles The resting pool of follicles called primordial follicles either die or are recruited into the growing pool of more developed follicles Although there is much debate about postnatal oocyte stem cells it is believed that the rate of demise of primordial follicles determines the repro-ductive longevity of an individual within a species While there are many mutations that have been identified that disrupt female fertility in model organisms (mainly mice) few mutations in ovary-specific genes have been identified in women with fertility problems How-ever new strategies for genome sequencing may help to identify causative fertility associated mutations Effective treatments exist for all types of ovarian failure though ovulation dysfunction is more dif-ficult to detect and empirical treatments have less certain benefits

The BASiC SCienCe Over the last 25 years we have witnessed amazing and rapid ad-vances in our understanding of the genetics of mammalian repro-duction in particular the genes and factors important for ovarian development and physiology [reviewed in (1ndash5)] In large part this progress has been possible because of breakthroughs in DNA se-quencing and transgenic mouse technology Knockout mouse strate-gies developed in the late 1980s and early 1990s allowed research-ers to address the question ldquoWhat is the essential role of a gene productrdquo Prior to this point the reproductive genetics community relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mu-tagenesismdasha challenging process akin to finding a proverbial needle in a haystack Embryonic stem (ES) cell technology allowed for the reverse genetics approach in which precise mutations could be cre-ated to disrupt a gene and subsequently elucidate its function

This technology has permitted identification of over 100 pro-teins that function in the formation of female embryonic germ cells at every stage of ovarian folliculogenesis in post-ovulation events and at various stages of meiosis and DNA repair These pioneering studies prompted work in other mammalian species including sheep with relevant and important translational follow up in humans Some of the pertinent studies will be described in depth below

geRMline STeM CellSThe pioneering transgenic mouse studies of Brinster and Palmiter (6) and others led not only to the advances in mouse reproductive genetics but also to advances in oocyte and embryo culture condi-tions for human assisted reproduction [elegantly reviewed in (7)] and in manipulation of the male germline (8) Spermatogonial stem cell transplantation techniques identified factors required for the main-tenance of male stem cells in an undifferentiated state and also un-covered genes that mark the differentiated state (8) More recently other groups have attempted to identify and define female germline stem cells generating enormous controversy in the literature the lay press and at scientific meetings While not wanting to join in the con-troversy especially since there may be some differences between animal models and humans we will comment on two intriguing stud-ies published recently First Liu and colleagues (9) used a genetic trick to make DDX4-positive germ cells in the postnatal ovary and thereby follow the differentiation and proliferation of these cells over time The authors could not detect mitotically active germline pre-cursors in contrast to the previous studies in mice (10) or humans (11) Second Saitou and colleagues (12) differentiated mouse XX ES cells and induced pluripotent stem (iPS) cells into cells resem-bling primordial germ cells (PGCs) reconstituted these cells with go-nadal somatic cells and showed that they could undergo meiosis In vivo these cells with meiotic potential could develop to the germinal vesicle stage and could give rise to fertile offspring when subjected to in vitro maturation and fertilization Thus oocyte precursors could be created from pluripotent stem cells and could be manipulated for fertilization

The above studies and other exciting research being performed in defining sex divergent steps in the differentiation of PGCs and the intrinsic and extrinsic factors necessary for prenatal entry into and arrest of meiosis will continue to influence the scientific pursuit of the elusive oocyte stem cell These studies will also aid in the in vitro pro-duction of functional oocytes from human ES cells andor iPS cells

BiDiReCTiOnAl COMMuniCATiOn DuRing FOlliCulOgeneSiS Understanding the control of the recruitment of follicles into the grow-ing pool has been an actively pursued area of research The Eppig and Nalbanov laboratories have postulated that oocyte-secreted fac-tors could influence somatic cell function in vitro [reviewed in (1)] and later work of Eppig showed that the oocyte dictated the pheno-type of the postnatal granulosa cells (13) In 1996 knockout mouse studies from the Matzuk laboratory uncovered for the first time the identity of one of these oocyte-secreted germline factors growth dif-ferentiation factor 9 (GDF9) a member of the transforming growth factor β (TGF-β) superfamily (14) Mice lacking GDF9 showed a

The Ovarian Factor Cutting-Edge Basic Science Combined with Classic Clinical Insights

Richard S Legro1 and Martin M Matzuk2

1 Department of Obstetrics and Gynecology Pennsylvania State University College of Medicine Hershey PA 2Department of Pathology and Immunology Baylor College of Medicine Houston TXrsl1psuedu (RSL) mmatzukbcmedu (MMM)

Produced by the ScienceAAAS Custom Publishing Office

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

36 37

Figure 1 Follicular ovarian and endometrial ultrasonographic measurements at the be-ginning and end of the one-month baseline period and at their maximum during r-metHu-Leptin treatment Each symbol represents one subject Reprinted with permission from (16)

block in folliculogenesis at the primary fol-licle stage and later studies identified an oocyte-secreted paralog bone morphoge-netic protein 15 (BMP15) which also influ-ences fertility in mice sheep and women (5) More recent studies demonstrated that mouse and human GDF9BMP15 heterodi-mers are far more biopotent than active homodimers (10ndash30-fold higher activity in mice ~3000-fold in humans) (15) Howev-er oocyte-secreted growth factors are only one-half of an ongoing bidirectional dialog that regulates key metabolic synchrony of oocytes and granulosa cells throughout fol-liculogenesis (see also page 13) The im-plications of these studies are far-reaching for animal and human fertility and include the development of new defined culture media supplemented with cocktails of oocyte and granulosa cell proteins and metabolites that take advantage of this crosstalk

PiTuiTARy COnTROl OFOVARiAn FunCTiOnFollicle-stimulating hormone (FSH) and luteinizing hormone (LH) are key regulators of ovarian function (16) Mice lacking FSH and women with homozygous mutations in the FSHβ or FSH receptor gene are infertile secondary to blocks in folliculogenesis at the mul-tilayer preantral follicle stage Mice or women with mutations in the LHβ or LH receptor genes have blocks prior to ovulation resulting in infertility The ability to predict defects at the hypothalamic or pituitary levels based on alterations in serum FSH or LH levels has enabled the identification of other genes that are important regulators of FSH and LH and that indirectly influence gonadotropin synthesis (16) An understanding of these key ovarian regulators has been critical for assisted reproduction

The CliniCAl SCienCeWe will now shift focus to highly cited articles that relate to clinical interventions that have improved (or replaced) ovarian function in an infertile population (Table 1) The choice is highly subjective but the goal is to include articles that have led to a paradigm shift in the treatment of female infertility We will cover treatments for the most common causes of ovulation failure as well as lesser ovula-tory problems such as those found in undiagnosed or unexplained infertility The World Health Organization (WHO) provides a schema for ovulation failure with three categories (based on hypothalamicpituitary and ovarian function) Type I (hypogonadotropic hypoestro-genic) Type II (normogonadotropic normoestrogenic) and Type III (hypergonadotropic hypoestrogenic)

WHO Type I Anovulation-Hypothalamic AmenorrheaHypothalamic amenorrhea can arise from genetic conditions leading to failure of the GnRH neurons to develop or migrate to the hypo-thalamus or from disruption of genes relating to onset of puberty There are also common acquired conditions associated with stress excessive exercise and loss of body fatweight (ie anorexia nervo-sa or exercise-induced amenorrhea) which can cause hypothalamic shutdown and downstream loss of ovarian function While treatment with gonadotropins effectively bypasses the hypothalamic GnRH pulse generator and directly stimulates follicular development the factors leading to the hypothalamic failure remain in doubt Cessa-tion of activity and regain of weight should cure the condition but no high-quality clinical trials have yet demonstrated this

The study of Welt etal (17) looked at the effects of recombinant leptin administration on ovarian function in women with hypotha-lamic amenorrhea The intervention group was observed without treatment for one month then administered recombinant leptin for up to three months A control group received no treatment Five of eight patients in the intervention group either ovulated or had some resumption of ovarian activity (Fig 1) None of the control subjects or intervention subjects during the initial observation pe-riod showed any change This provided evidence that peripheral fat status was critical to maintaining reproductive function and sup-ported studies that a critical amount of fat mass is necessary to initiate menarche

WHO Type II Ovulatory Dysfunction-Polycystic Ovary SyndromeA study of Legro etal (18) occurred at a time of great debate as to the best treatment for polycystic ovary syndrome (PCOS) a hetero-geneous condition of hyperandrogenism and chronic anovulation with characteristic polycystic ovaries filled with preantral follicles PCOS was viewed by some as a metabolic syndrome due to dimin-

Figure 2 Regimens of ovar-ian stimulation and hormone replacement used to synchro-nize the development of ovar-ian follicles in the oocyte donor and endometrial cycle in the recipient Reprinted with per-mission from (20)

Figure 3 The time course and quantitative change in circulating concentrations (Mean plusmn SE) of lutein-izing hormone (LH) follicle-stimulating hormone (FSH) estradiol (E2) and progesterone (P) during the menstrual cycle of four normal women treated with a GnRH agonist for three days after menses onset (hatched box left) Data were centered around the midcycle gonadotropin peak (day 0) Reprinted with permission from (25)

Produced by the ScienceAAAS Custom Publishing Office

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

38 39

ished insulin actionmdashrequiring primary treatment with insulin-sen-sitizing effectsmdashand by others as a reproductive abnormality with inappropriate gondadotropin secretion and corresponding altered ovarian steroidal production and lack of development of functional follicles resulting in anovulation The study examined the effects of ovulation-inducing agents clomiphene alone vs metformin alone vs their combination in infertile women with PCOS using a double blind double dummy design

Contrary to expectations clomiphene was markedly superior to metformin achieving a twofold higher live-birth rate with the combi-nation offering a further marginal but non-significant improvement Importantly the quality of ovulation differed depending on ovulation-inducing agent the improved fecundity and ovulation rates on clomi-phene were associated with increased body weight waist circumfer-ence and insulin levels from baseline implying that improvement in these factors were not as critical to restoration of ovulation in these women as previously thought This study served to justify the search for ovulation-inducing agents that work primarily by altering ovarian steroid feedback to the hypothalamic pituitary axis such as aroma-tase inhibitors that block the conversion of excess androgens to bio-active estrogens

WHO Type III Anovulation Age Related to Diminished Ovarian ReservePrimary ovarian insufficiency can be due to genetic conditions such as Turnerrsquos Syndrome or single gene defects such galactosidase de-

ficiency or mutations acquired from exposure to radiation or alkylat-ing agents during cancer treatment Further while Type III anovula-tion occurs relatively rarely all women experience an age-related decline in ovarian function with increasing rates of embryo aneu-ploidy and miscarriage culminating in the complete loss of ovulation and menses at menopause While preservation of fertility is a rapidly expanding preventive treatment option there have been no real ad-vances in restoring ovarian function in women with age-related ovar-ian insufficiency A breakthrough occurred when it was shown that embryos created from donor oocytes could be placed in the uterus of a woman with ovarian insufficiency whose endometrium had been rendered receptive to embryo implantation using sex steroid treat-ment Case reports describing embryo flushing in inseminated oo-cyte donors (19) and IVF performed using donor oocytes (20) pro-vided evidence in support of this procedure

The broader clinical implication built upon this evidence was the extension of this therapy to women with advanced age and infertility due to what is now described as diminished ovarian reserve (DOR) Sauer etal (2122) studied women over 40 with DOR finding that implantation of donated oocytes yielded pregnancies and live-birth rates markedly superior to expected fertility rates based on age Manipulation of hormone levels to prepare the uterus for implanta-tion is shown in Figure 2 Rates of pregnancy after oocyte dona-tion routinely exceeded those after standard IVF in women of all age groups in registries throughout the world Such high success rates in a group that had previously experienced such dismal results reset

Category SpecificCondition Reference KeyFinding

WHO Type I Anovulation

Hypothalamic Amenorrhea due to exercise or excessive thinness

16Recombinant leptin was able to initiate ovarian function in five of eight women given the intervention three of whom ovulated implying that fat mass is critical to reproductive function

WHO Type II Anovulation

Polycystic Ovary Syndrome 17

Ovulation and live birth as well as fecundity per ovulation were better with clomiphene than metformin despite metforminrsquos improved effects on weight and insulin levels

WHO Type III Anovulation

Age-related diminished ovarian reserve

20

Oocyte donation is an effective treatment for women with diminished ovarian reserve with live-birth rates similar to those with primary ovarian insufficiency suggesting uterine function is not age-dependent

Ovulatory Dysfunction

Acquired luteal phase defect 25

A luteal phase defect characterized by a shortened luteal phase with inadequate circulating serum progesterone levels could be induced by the administration of a GnRH agonist in the early follicular phase in normally ovulating women

Subtle Ovulatory Dysfunction

Unexplained infertility 26

Empiric treatment with the ovulation induction agent clomiphene or with unstimulating intrauterine insemination is no better than expectant management for couples with unexplained infertility

expectations for subsequent therapies Further it underscored that the primary effects of aging impacted oocyte quality and number

OVulATORy DySFunCTiOnmdashluTeAl PhASe DeFeCTSMany women with infertility or recurrent pregnancy loss do not present with chronic or sporadic anovulation but are suspected to have some form of ovulation dysfunction leading to the absence of functional oocytes andor to implantation failure Several ovulatory disorders occur within the context of what appears to be regular ovulation but continue to defy clear definition and diagnosis These include extremely rare disorders such as luteinized unruptured fol-licle syndrome empty follicle syndrome and more common disor-ders such as luteal phase defects (LPDs) LPDs originally described by Georgeanna Jones are characterized by inadequate corpus luteum function following ovulation as measured by a variety of parameters including inadequate rise in basal body temperature secretion of progesterone or its metabolites an out-of-phase en-dometrial biopsy or a shortened luteal phase (23) Subsequent studies have found this disorder difficult to diagnose largely due to the occurrence of these ldquoabnormalitiesrdquo in normal fertile populations (2425)

However the proof of concept for LPD was demonstrated in a clas-sic study by Sheehan et al (26) in which normally ovulating women (verified by careful monitoring of gonadotropins and sex steroids prior to treatment) were administered a GnRH agonist in the early follicular phase that resulted in a temporary increase in gonadotropin secretion followed by a decrease in FSH levels and a shortened luteal phase (Fig 3) While this was seen at the time as a potential contraceptive it helped to justify the use of progesterone adminis-tration to supplement the luteal phase in IVF cycles where a GnRH agonist had been administered and was also used to treat women with suspected LPDs

unexPlAineD inFeRTiliTymdasheMPiRiC OVulATiOn inDuCTiOnUnexplained infertility remains the most heterogeneous of infertility diagnoses and contains subclinical etiologies related to ovulatory tubal and male factors Couples often receive empiric therapy which takes a shotgun approach trying to hit multiple targets with a single shot Empiric treatment with ovulation-inducing agents such as clo-miphene and gonadotropin has two intended goals to increase the quality of ovulation (difficult to quantify as noted above with LPDs) and to cause superovulation which results in multiple mature oo-cytes and both a greater chance for pregnancy and a higher risk of multiple pregnancies

The study by Bhattacharya et al (27) randomized couples with unexplained infertility into three treatment groups (with similar ini-tial pregnancy rates) empiric clomiphene citrate unstimulated in-trauterine insemination (IUI) or expectant management The trial was unique for the inclusion of the control expectant management group a ldquowatch and waitrdquo approach not usually regarded as accept-able in an infertility clinical trial Although subjects in the expectant management group were less satisfied with their participation than others the trial did meet its enrollment goals This trial examined the role of subtle subclinical ovulatory dysfunction in unexplained infertility and although there was no statistically significant benefit

to IUI it trended better than clomiphene This study raised the bar for empiric infertility treatments introducing the requirement that proof of benefit of the intervention over expectant management be shown before acceptance as a clinical treatment It further un-derscores the need for better diagnosis strategies in unexplained infertility

COnCluSiOnSThis review summarizes groundbreaking research that offers hope for new treatments of ovarian disorders that lead to ovulation dys-function and anovulation It highlights the critical role of the oocyte in its traditional responsive role to gonadotropins and ovarian factors but also its newer role in guiding ovarian function With a greater emphasis on translation of this work to the clinic we anticipate that the classic treatments of the past will soon be supplanted by newer and better evidence-based strategies

REFERENCES 1 M M Matzuk K Burns M M Viveiros J J Eppig Science 296 2178 (2002) 2 M M Matzuk D J Lamb Nat Cell Biol 8 (S1) S41 (2002) 3 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 4 M A Edson A K Nagaraja M M Matzuk Endocr Rev 30 624 (2009) 5 M M Matzuk K H Burns Annu Rev Physiol 74 503 (2012) 6 R D Palmiter R L Brinster Annu Rev Genet 20 465 (1986) 7 A R Shuldiner N Engl J Med 334 653 (1996) 8 R L Brinster Science 316 404 (2007) 9 H Zhang et al Proc Natl Acad Sci USA 109 12580 (2012)10 K Zou Z Yuan Nat Cell Biol 11 631 (2009)11 Y A White et al Nat Med 18 413 (2012)12 K Hayashi et al Science 338 971 (2012)13 J J Eppig K Wigglesworth F L Pendola Proc Natl Acad Sci USA 99 2890 (2002)14 J Dong et al Nature 383 531 (1996)15 J Peng et al Proc Natl Acad Sci USA 110 e776 (2013)16 A Roy M M Matzuk Nat Rev Endocrinol 7 517 (2011)17 C K Welt et al N Engl J Med 351 987 (2004)18 R S Legro et al N Engl J Med 356 551 (2007)19 J E Buster et al Lancet 2 223 (1983)20 P Lutjen et al Nature 307 174 (1984)21 M V Sauer R J Paulson R A Lobo N Engl J Med 323 1157 (1990)22 M V Sauer R J Paulson R A Lobo JAMA 268 1275 (1992)23 H W Jones Jr Fertil Steril 90 e5 (2008)24 C Coutifaris et al Fertil Steril 82 1264 (2004)25 H L Hambridge SL Mumford Hum Reprod 28 1687 (2013)26 K L Sheehan et al Science 215 170 (1982)27 S Bhattacharya et al BMJ 337 a716 (2008)

Acknowledgments Studies on the female reproductive tract in the Matzuk laboratory have been supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development through grants RO1 HD032067 RO1 HD033438 U54 HD007495 and the Ovarian Cancer Re-search Fund and in the Legro laboratory by grants R01HD056510 U54 HD034449 and U10 HD 38992

Produced by the ScienceAAAS Custom Publishing Office

Table 1 List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

40 41

Infertility The Male FactorDolores J Lamb123 and Larry I Lipshultz12

inTRODuCTiOnInfertility impacts the lives of 8 to 12 of couples attempting to conceive for the first time In roughly half of these cases a male factor is causative yet surprisingly in many assisted reproductive technology (ART) clinics the male is not clinically evaluated beyond a routine semen analysis While most textbooks state that primary infertility in approximately 30 of infertile men is idiopathic some experts speculate that 50 to 80 of all male infertility results from a (usually undiagnosed) genetic cause In these patients no explana tion can be found for their impaired semen quality In part this statistic reflects our poor understanding of the basic mecha-nisms regulating sex determination genital tract differentiation and development spermatogenesis sperm maturation in the genital tract and the molecular events required for fertilization and early embryonic development hence our inability to properly diagnose the cause of the infertility Indeed many of the commonly used di-agnostic categories for male infertility are descriptive rather than mechanistically based A diagnosis of non-obstructive azoosper-mia (NOA) testicular failure cryptorchidism or ductal obstruction provides no insight into the molecular cause of the infertility In this review we focus on the molecular defects that underlie the congeni-tal genitourinary birth defects associated with infertility endocrine deficiencies idiopathic spermatogenic failure and deficiencies of sperm function

STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe DeFeCTS ACCOunT FOR A SigniFiCAnT PeRCenTAge OF MAle inFeRTiliTyKaryotype abnormalities are present in about 6 of infertile men [reviewed in (1)] With severe spermatogenic deficiency the inci-dence of karyotype abnormalities increases At our tertiary referral clinic at the Baylor College of Medicine more than 11 of NOA men and nearly 17 of men with male genitourinary birth defects have karyotype anomalies (2) These anomalies can be numerical and af-fect only the sex chromosomesmdashfor example Klinefelter syndrome (46XXY-46XXXXY) which accounts for about 14 of all NOAmdashor structural affecting the autosomes or the sex chromosomes includ-ing complex chromosomal rearrangements deletions duplications inversions and translocations

A well-recognized structural chromosome defect causing male infertility is the occurrence of Y chromosome microdeletions en-compassing the azoospermia factor (AZF) region first described in 1995 (3) The DeletedinAzoospermia(DAZ) gene was the first AZFspermatogenesis gene identified within a Y chromosome microdele-

tion The AZF region is now further subdivided into smaller segments termed AZFa AZFb and AZFc Sperm are unlikely to be found in men with deletions in AZFa or b whereas men with AZFc deletions encompassing DAZhave a 75 chance of having rare sperm found on a testis biopsy [reviewed in (1)] For AZFcmicrodeletions the rare variant having a major effect on spermatogenesis is seen with the b2b4 deletion which accounts for about 6 of severe spermatogen-ic failure whereas the more commonly occurring grgr deletion has a modest affect doubling the risk of severe spermatogenic failure (4)

Upon homologous recombination and meiotic segregation auto-somal structural defects such as Robertsonian translocations can result in balanced or unbalanced translocations Infertility can arise due to the production of unbalanced gametes (nullisomic or disomic) resulting in aneuploid embryos or chromosomally unbalanced off-spring (5) Thus before proceeding with ART to attempt to achieve a pregnancy it is important to know if chromosome anomalies are present in the male patient

enDOCRine PAThWAyS ARe CRiTiCAl FOR nORMAl FeRTiliTy in MAleSAlthough endocrine defects account for only about 1 of all male infertility the molecular abnormalities that underlie these conditions are increasingly evident In short any genetic defect affecting the hypothalamic-pituitary-gonadal axis steroid hormone biosynthesis and metabolism steroid hormone receptors and their signaling pathways peptide hormones and their receptors and associated signaling pathways or paracrine factors and cytokines and their respective signaling pathways can affect male fertility [reviewed in (6ndash8)]

The best example of the relative complexity of genetic defects that underlie a seemingly discrete infertility phenotype is reveal by the studies of hypogonadotropic hypogonadism by Crowley and others (9ndash19) Gene defects causing GnRH deficiency vary in relation to their site of action on the ontogeny of the GnRH neurons and the clinical phenotype observed For patients with anosmia neurodevelopmen-tal gene functions that regulate GnRH neuronal migration or olfactory bulbtract development are affected due to gene deficiencies such as in KAL1andNELF In contrast GnRH-deficient patients with nor-mosmia may have defects in genes that encode members of the kis-speptin neurokinin B and GnRH signaling pathways such asKISSKISS1RTAC3TACR3 and GnRHR Interestingly defects in the prokineticin and FGF signaling pathways (FGF8 FGFR1 PROK2R) are variably present in patients and families with both anosmia and a normal sense of smell within the same pedigree [reviewed in (9)] De-spite the identification of numerous monoallelic bialellic and digenic mutations in the genes above a molecular cause for slightly less than half of isolated GnRH deficiency remains unknown and a topic of in-tense investigation It is clear that even for hypogonadotropic hypo-gonadism considerable genetic complexity underlies the causative defects

1Center for Reproductive Medicine 2Scott Department of Urology and 3Department of Molecular and Cellular Biology Baylor College of Medicine Houston TXCorresponding Author dlambbcmedu

geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTyUsing mouse models investigators were surprised to learn that genes never thought to be involved in male reproductive function when deleted cause infertility One of the most unexpected finds was that loss or pharmacological blockage of testis-expressed taste genes caused male infertility in the mouse (20) The targeted dele-tion of TAS1R3 a component of taste receptors and the gustducin α-subunit GNAT3 lead to male sterility due to immotile sperm with poor morphology (detached or amorphous heads tails flipped over heads and multiple kinks and loops in the tails) indicating a poten-tial role in spermiogenesis and sperm maturation (20)

In 2008 Matzuk and Lamb summarized the gene defects in mouse models and human patients associated with male infertility [see sup-plement to (7)] and a large number of additional gene defects have since been defined affecting genitourinary birth defects disorders of sexual determination and differentiation puberty spermatogenesis sperm function and other aspects of male reproductive health For example cationic channel of sperm (CatSper)mdashthe main flagellar Ca2+ channel in spermmdashwas shown to play a role in nongenomic progresterone signaling (21) Upon activation by progesterone calcium influx triggers hyperactivated motility and the acrosome reaction While CatSper mutations occur rarely they can result in severe asthenozoopsermia (22) Unfortunately mouse models are not informative because unlike humans the CatSper channel in the mouse is not sensitive to progesterone Nevertheless there are literally thousands of possible gene defects identified in mice and humans causing male infertility In the clinic however just one gene is analyzed on a routine basis in males with congenital bilateral ab-sence of the vas deferens (CBAVD)mdashthe CFTR gene (cystic fibrosis transmembrane regulatory protein) (23) CBAVD is a genital form of cystic fibrosis For patients of North African ethnicity patients may also be tested for mutations of the Aurora Kinase C gene specifically the c144delC mutation (24) This mutation results in large-headed polyploid multi-flagellar sperm because this protein is necessary for male meiotic cytokinesis As research continues additional genetic testing will no doubt become standard of care in the evaluation of the infertile male

FeRTiliTy PReSeRVATiOn FOR PReVenTiOn OF SeCOnDARy inFeRTiliTyIn the testis long-cycling isolated spermatogonia identified by mor-phological criteria have been proposed to be testicular stem cells (25ndash27) The pioneering work of Brinster and colleagues demon-strated with certainty that these testicular stem cells exhibit the unique properties of prolonged proliferation self-renewal generation of differentiated progeny maintenance of developmental potential and proliferation in response to injury (28) Although work in this area is predominantly in rodent models and more recently primates this represents a research area of significant promise for patients facing secondary infertility

Spermatogenesis is damaged by exposure to chemotherapeutic agents radiation toxic agents and occupational exposures to go-nadotoxins resulting in secondary infertility Prolonged periods of azoospermia or severe oligospermia may result While sterility ap-pears permanent spermatogenesis may spontaneously recover years after treatment (2930) when the surviving reserved stem cells

(type A spermatogonia) reinitiate the proliferative and differentiative events required for spermatogenesis Today cancer patients should be advised to bank cryopreserved semen prior to potentially harmful treatment Children who undergo cancer treatment cannot produce an ejaculate for cryopreservation and even for adults most cou-ples would prefer a natural conception Accordingly application of spermatogonial stem cell isolation and cryopreservation followed by germ cell transplantation to restore spermatogenesis after recovery from cancer treatment may provide a very useful approach to pres-ervation of fertility for these cancer patients

A significant roadblock to translation of these studies to clinical practice has been the concern that the testis may serve as a reser-voir for cancer cells (hematological cancers in particular) and trans-plantation of spermatogonial stem cells obtained prior to chemo-therapy could transmit the malignant cells back into the now healthy recipient Recently a novel cell sorting strategy was devised (21) to remove malignant contamination while simultaneously enriching the population of human spermatogonial stem cells A human to mouse xenotransplantation biological assay was used to define both the spermatogonial stem cell activity and malignant contamination of the enriched cell populations The development of these separation technologies will be a major advance towards eventual application of spermatogonial stem cell transplantation in the clinic However human studies demonstrating safety and efficacy will be needed as current purities of 988 to 998 are still insufficient even small numbers of malignant cells present during hematopoietic stem cell transplantation will prove problematic Importantly hematopoietic stem cell transplantation is required to treat a life-threatening condi-tion whereas fertility preservation is an elective procedure [reviewed in (31)]

Human spermatogonial stem cells can be cultured under condi-tions that allow them to exhibit pluripotent potential (32 33) The cells exhibit properties similar to embryonic stem cells and can pro-duce teratomas when transplanted into immunodeficient mice The human adult germline stem cells can differentiate into the full range of cell types including germline cells (3233)

In the future spermatogonial stem cells will not only offer the promise of seminiferous tubule rejuvenation after exposure to gonadotoxins but perhaps also the potential to regenerate or-gans and tissues Much further in the future these cells may be used for gene therapy to cure certain Mendalian inherited genetic diseases

heAlTh RiSkS FOR inFeRTile Men gOBeyOnD TheiR RePRODuCTiVe WOeSAlthough it is important to find methods to bypass or overcome se-vere male factor infertility to assist severe male factor patients in achieving a pregnancy bypassing naturersquos barriers to fertilization by utilizing potential defective sperm for in vitro fertilization by in-tracytoplasmic sperm injection (ICSI) may have unrecognized con-sequences for the patient and his offspring Today we know that Y chromosome microdeletions occur through events that generate isodicentric Y chromosomes as well as Y chromosomes with dele-tions duplications or inversions The first report of Y chromosome microdeletions and their association with NOA came from David Pagersquos laboratory (described above) (3) but it has been recognized

Produced by the ScienceAAAS Custom Publishing Office

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

42 43

more recently that these structural defects can be generated by a variety of mechanisms Indeed intrachromosomal homologous re-combination can generate pseudo-isoY chromosomes and Y chro-mosome pericentric inversions (34) At one time it was thought that about 95 of the Y chromosome (except the pseudoautosomal re-gions) did not undergo homologous recombination during meiosis However as a result of breakpoint hotspots as well as homologous recombination errors due to amplicons associated with palindromic

structures present on the human Y chromosome Y chromosome microdeletions may occur (35) These rearrangements can even occur due to amplicons on the short (Yp) and long (Yq) arms of the Y chromosome through intrachromosomal non-allelic homologous recombination (34)

These structural Y chromosome defects affect the male specific Y and although other genes not involved in spermatogenesis were affected the clinical consequence of these defects appeared

Figure 1 SHOX deletions and duplications in patients with Y chromosome microdeletions co-existing genomic syndromes Y chromosome microdeletions are usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach However when array comparative genomic hybridization is employed in addition to the Y chromosome microdeletions found additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified Some of these encompass the SHOX gene Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome microdeleted man Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene Both of these men had stature abnormalities as well It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluo-rescent in situ hybridization reveals that these variations are present on the Y chromosome

to predominantly affect spermatogenesis However in part due to isodicentric Y chromosome formation and complex rearrangements and deletionsduplications of the Y about 25 of men with NOA due to Y chromosome microdeletions have a co-existing genomic syndrome SHOX (short stature homeobox-containing gene) syndrome (36) (Fig 1) SHOX syndrome is the most common cause of short stature and results from mutations microdeletions or microduplications of the SHOX gene which is located in the pseudoautosomal region of the sex chromosomes SHOX is a dosage-sensitive gene that does not undergo X-inactivation While SHOX syndrome occurs in Turner and Klinefelter syndrome patients (deletion and duplication respectively) the clinical spectrum varies The associated Madelung deformity of the forearm and mesomelic disproportion of the limbs develops over time and appears during the second decade of life (or perhaps never) There are a number of other clinical indications (among them scoliosis microgathia and muscular hypertrophy of the calves) that are found in some but not all SHOX syndrome patients [reviewed in (37)] The presence of SHOX deletion or duplication in a subset of men with Y chromosome microdeletions suggests that this co-existing genomic syndrome could be inherited by the male offspring of men conceived by ICSI although to date there have been no reports of SHOX syndrome in ICSI- conceived offspring

There may be other associated health issues for the NOA male that are clinically unappreciated Our studies show a 29-fold increased risk of cancer in NOA patients (38) Walsh and colleagues (39) evaluated data on 22562 partners of infertile couples and showed the risk of testis cancer was nearly threefold higher in infertile men compared with normal men (38) Jacobsen etal similarly evaluated more than 32000 men who had their semen analysed over a 30-year period and found a 16-fold increased risk of testis cancer in men with reduced sperm counts (40) There was also an increased incidence of cancers of the peritoneum and other digestive organs in these men Walsh and colleagues performed an analysis of their patient cohort and showed that those with male factor infertility were 26 times more likely to be diagnosed with high-grade prostate cancer (3941)

Indeed a comprehensive male infertility evaluation identified a number of significant (and previously undiagnosed) medical patholo-gies In a landmark paper by Honig et al significant medical pa-thologies were uncovered in 11 of 1236 patients presenting to a male infertility clinic (42) Ten patients (08) had tumors (six testis three brain and one spinal cord) and their findings point to the need for urologic consultation when an infertile couple seeks treatment at an ART clinic It is believed that there are common etiologic factors for infertility and cancer development such as deficient DNA repair mechanisms (394043)

COnCluSiOnSWe have witnessed an exponential increase in advances in our understanding of the molecular controls of spermatogenesis and sperm function as well as increasing definition of the molecular de-fects associated with infertility in the male A major challenge that remains is to enhance our abilities to translate these research ad-vances into routine clinical practice Additionally it is essential to ensure that every male factor patient has a complete history and

physical performed by a urologist or andrologistendocrinologist as well as an in-depth assessment of the causes of his infertility Our goal is to have physicians recognize that there may be other health risks for the infertile male that go beyond their infertility We thus look with hope towards the future where the research ad-vances of today will be translated to clinical practice resulting in not only restoration of fertility for currently infertile men after gonado-toxin exposure but perhaps even regeneration of organs and tis-sues without the ethical constraints that limit embryonic stem cell research today Importantly infertile couples must understand that the cause of their infertility may remain unknown They also need to carefully consider the potential health risks for both the patient and any offspring conceived by ART that go beyond possible birth defects

REFERENCES 1 A W Pastuszak D J Lamb J Androl 33 1075 (2012) 2 A N Yatsenko et al J Urol 183 1636 (2010) 3 R Reijo R K Alagappan P Patrizio D C Page Lancet 347 1290 (1996) 4 S G Rozen et al Am J Hum Genet 91 890 (2012) 5 I Bernicot et al Eur J Med Genet 55 245 (2012) 6 K Hwang et al Ann NY Acad Sci 1214 E1 (2010) 7 M M Matzuk D J Lamb Nat Med 14 1197 (2008) 8 M M Matzuk D J Lamb Nat Cell Biol 4 Suppl s41 (2002) 9 W F Crowley Jr Mol Endocrinol 25 1989 (2011)10 B Mosinger et al Proc Natl Acad Sci USA Epub ahead of print (2013) doi 101073pnas130282711011 J F Smith et al Proc Natl Acad Sci USA 110 6823 (2013)12 M S Hildebrand et al Eur J Hum Genet 18 1178 (2010)13 A Anguiano et al JAMA 267 1794 (1992)14 K Dieterich et al Hum Mol Genet 18 1301 (2009)15 C Huckins Cell Tissue Kinet 4 335 (1971)16 C Huckins Cell Tissue Kinet 4 313 (1971)17 C Huckins Anat Rec 169 533 (1971)18 R L Brinster J W Zimmermann Proc Natl Acad Sci USA 91 11298 (1994)19 M L Meistrich G Wilson B W Brown M F da Cunha L I Lipshultz Cancer 70 2703 (1992)20 R M Pryzant M L Meistrich G Wilson B Brown P McLaughlin J Clin Oncol 11 239 (1993)21 S L Dovey et al J Clin Invest 123 1833 (2013)22 S Conrad et al Nature 456 344 (2008)23 N Kossack et al Stem Cells 27 138 (2009)24 J Lange et al Genomics 102 257 (2013)25 S Repping et al Am J Hum Genet 71 906 (2002)26 C J Jorgez et al J Clin Endocr Metab 96 E674 (2011)27 G Binder Horm Res Paediatr 75 81 (2011)28 M L Eisenberg P Betts D Herder D J Lamb L I Lipshultz Fertil Steril Epub ahead of print (2013) doi 101016j fertnstert20130502229 T J Walsh et al Cancer 116 2140 (2010)30 R Jacobsen et al BMJ 321 789 (2000)31 J M Hotaling T J Walsh Nat Rev Urol 6 550 (2009)32 S C Honig L I Lipshultz J Jarow Fertil Steril 62 1028 (1994)33 M R Maduro et al Mol Hum Reprod 9 61 (2003)

Produced by the ScienceAAAS Custom Publishing Office

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

44 45

Molecular Interplay in Successful Implantation

Jeeyeon Cha1 Felipe Vilella2 Sudhansu K Dey1 and Carlos Simoacuten2

ABSTRACTEmbryo implantation in the maternal womb is a fundamental event in mammalian procreation The phases of implantation are apposition attachment and finally penetration of the blastocyst trophectoderm through the uterine epithelium and into the stroma Both uterine re-ceptivity and blastocyst activation are required for successful implan-tation This review briefly highlights the molecular processes respon-sible for endometrial receptivity implantation and decidualization in mice and humans

inTRODuCTiOnImplantation requires an effective bidirectional communication be-tween the receptive endometrium and an implantation-competent blastocyst The duration of the window of implantation (WOI) to achieve optimal pregnancy outcome is short-lived Blastocyst attach-ment to the uterine lining (luminal epithelium) initiates the implanta-tion process which is followed by localized stromal cell proliferation and differentiation to generate specialized decidual cells at the site of implantation a process called decidualization Appropriate de-cidualization directs placentation in which the maternal circulation is brought into contact with the embryonic vascular system estab-lishing a functional placenta Maternal resources filtered across the placental barrier nourish and protect the fetus Implantation is the first significant encounter between the mother and embryo and is crucial for a successful pregnancy

MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnSOF enDOMeTRiAl ReCePTiViTyThe WOI a transient interphase between the prereceptive and re-fractory phases lasts approximately 24 hours (beginning day four of pregnancy or pseudopregnancy) in mice and three days in hu-mans during the mid-luteal phase of the menstrual cycle (1ndash3) Un-derstanding the ldquomolecular clockrdquo at play during the WOI could help to identify biomarkers of endometrial receptivity useful for increasing pregnancy rates in in vitro fertilization (IVF) programs or to develop novel contraceptives

The master regulators for the acquisition of endometrial receptivity are estrogen and progesterone (P4) In mice the transition from the prereceptive to the receptive phase requires priming with P4 super-imposed with a small amount of estrogen The P4 and estrogen nu-clear receptors receptor chaperones and co-regulators all play cru-

1Division of Reproductive Sciences Cincinnati Childrenrsquos Hospital Medical Center Cincinnati OH2Fundacioacuten Instituto Valenciano de Infertilidad (FIVI) Valencia University and Instituto Universitario IVIINCLIVA Valencia SpainThese authors contributed equallyCorresponding authors SKDeycchmcorg (SK) CarlosSimonivies (CS)

cial roles in regulating the WOI Progesterone receptors (PR-A and PR-B) and estrogen receptors (ERα and ERβ) are expressed in the uterus Studies in mice have shown that these isoforms serve as the principle modulators during specific stages of pregnancy ERα (en-coded by the Esr1 gene) is the primary mediator of estrogen activity in the uterus during implantation as the uteri of Esr1-- mice are hy-poplastic and unable to support implantation (4) Mice missing both PR-A and PR-B are also infertile and have multiple ovarian and uter-ine defects although PR-Bndashdeficient mice show normal fertility (56) indicating that PR-A is the key player in implantation FKBP52 an immunophilin co-chaperone for steroid hormone nuclear receptors optimizes PR activity and has profound effects during pregnancy (78) In the absence of Fkbp52 P4 signaling and responsiveness are significantly diminished resulting in implantation failure Depending on the genetic background of the mice this functional P4 resistance can be overcome by exogenous P4 administration (9) FKBP52 is also expressed in human endometrium (10)

ER and PR signaling during implantation are executed by jux-tacrine paracrine and autocrine factors orchestrated by various growth factors cytokines lipid mediators homeobox transcription factors and morphogens Mouse models have improved our mecha-nistic understanding of several factors critical for uterine receptiv-ity and implantation (Fig 1 also see reviews 1and2) Among the growth factors heparin-binding EGF-like growth factor (HB-EGF) stands out as a major player in mediating embryo-uterine interac-tions during the attachment reaction in both mice and humans (Fig 2 11ndash14) Membrane-anchored HB-EGF expressed in the luminal epithelium functions as a juxtacrine mediator for adhesion of the blastocyst expressing EGF receptors while soluble HB-EGF influ-ences embryo-uterine interactions in a paracrine manner (1) Juxta-crine interactions between the luminal epithelium and blastocyst in humans are also mediated by L-selectin ligands and their receptors displayed on the luminal epithelium and blastocyst cell surface re-spectively (15)

Leukemia inhibitory factor (LIF) a member of the interleukin (IL)-6 family of cytokines is expressed in mouse uterine glands early on day four of pregnancy (the day of uterine receptivity) and in the stroma surrounding the blastocyst upon attachment late on day four (1617) This factor is essential for uterine receptivity and implan-tation via binding to its receptor LIFR and co-receptor gp130 to activate STAT3 signaling LIF-gp130-STAT3 signaling is critical to implantation since uterine deletion of the genes encoding gp130 or Stat3has been shown to cause implantation failure (1819) More recently identified mediators critical for uterine receptivity are muscle segment homeobox genes (Msx1Msx2) conditional uterine deletion of these genes results in implantation failure (Fig 3 20)

In some respects implantation resembles a pro-inflammatory reaction and involves inflammatory mediators Cyclooxygenase-2 (Cox2) a critical enzyme for prostaglandin (PG) biosynthesis is

expressed in the uterus at the site of implantation (21ndash23) Cox2-derived PGs are thought to participate in implantation since ablation of the Ptgs2 (Cox2) gene leads to infertility (21ndash23) PGs also influence vascular permeability and decidualization in the stromal bed (24) Cox2-derived prostacyclin (PGI2) activates PPARδ which appears to influence implantation as Pparδ-- mice show deferred implantation and compromised pregnancy outcome (22) Cytosolic phospholipase A2α (cPLA2α encoded by Pla2g4a) which generates arachidonic acid for Cox2-generated PG synthesis is expressed in the uterus similarly to Cox2 Its critical role in implantation is evidenced by deferred implantation and related adverse effects in Pla2g4andashndash mice compromising pregnancy outcome (25)

Lpar3--mice exhibit a similar phenotype to Pla2g4andashndashmice exhibiting downregulated Cox2 (26 reviewed in 1and2)

Among other lipid mediators endogenous cannabinoid-like com-pounds (endocannabinoids) and their G-protein coupled recep-tors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation Endocannabinoids N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol (2-AG) bind to CB1 and CB2 Uterine anandamide levels are higher during the prereceptive phase but decrease during the receptive phase (27) Similarly increased anan-damide levels resulting from deficiency of fatty acid amide hydrolase (FAAH) which degrades anandamide lead to multiple pregnancy defects including disruption of on-time implantation (28) These re-

Figure 1 Signaling network for uterine receptivity and implantation Hybrid cartoon based on mouse and human studies portraying compartment- and cell-type specific expression of molecules and their potential functions critical for uterine receptivity implantation and decidualization Interplay of ovarian P4estrogen-dependent and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrineparacrineautocrine manner During attachment interactions between the blastocyst and luminal epithelium (LE) involve ErbB14 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF as well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst The other critical signaling pathways for uterine receptivity and implantation are also shown AA arachidonic acid COUP-TFII chicken ovalbumin upstream promoter transcription factor-2 E estrogens EC epithelial cell (luminal and glandular epithelia) ENaC epithelium sodium channel ErbB14 epidermal growth factor receptor 14 GE glandular epithelium Hand2 Heart- and neural crest derivatives-expressed protein 2 HB-EGF heparin-binding EGF-like growth factor ICM inner cell mass KLF5 Kruppel-like factor 5 LPA3 lysophosphatidic acid receptor 3 MSX1 muscle segment homeobox 1 PPARδ peroxisome proliferator-activating receptor δ Ptc Patched RXR retinoid X receptor SC stromal cell SGK1 serum- and glucocorticoid-inducible kinase 1 Smo Smoothened Tr trophectoderm Compartment colors blue stroma pink luminal epithelium orange glandular epithelium purple epithelium at the attachment site Adapted from (1)

Produced by the ScienceAAAS Custom Publishing Office

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

46 47

sults indicate that a tight regulation of endocannabinoid signaling is critical to uterine receptivity and oviductal transport of embryos (see page 21) Aberrant anandamide levels are also associated with re-current pregnancy loss and ectopic pregnancy in women (2930)

In humans receptive status is typically defined by histological characteristics known as the Noyes criteria (31) However the accu-racy and relevance of these criteria have been questioned (3233) Thus the search is on-going for specific biomarkers that define the WOI Morphological changes of the endometrial luminal epithelium include modifications of the plasma membrane (34) and cytoskel-eton (35ndash37) The presence of pinopodes was proposed as a recep-tivity marker (3839) but these ectoplasmic epithelial projections are not specific to the receptive state (40) Biochemical markers such as integrins (41) MUC1 (42) calcitonin (43) LIF (16ndash17) and HOXA10 (44) initially generated interest but none have proven to be effective in clinical practice

Microarray technology has advanced the search for biomarkers based on the transcriptomics signature during the WOI (45) Recent evidence has helped to characterize the human endometrial cycle by expression profiling with respect to receptive status (346) A di-agnostic tool has been developed to identify receptive endometrium using a specific transcriptomics signature in both natural and hor-monal replacement therapy cycles (47) The endometrial receptiv-ity array (ERA) is a customised microarray platform of 238 genes differentially expressed during the menstrual cycle that can identify the WOI of each patient (48) ERA appears superior to endometrial histology and results are reproducible in the same patient on the same day of her menstrual cycle up to 40 months later (48) ERA for WOI detection to schedule embryo transfer in patients with re-current implantation failure has recently been reported as a novel IVF therapeutic strategy (49) The proteomic profile of the human endometrium throughout the menstrual cycle has also been stud-ied (50ndash52) However despite the large amount of information gen-erated a consensus profile has not been established Analysis of endometrial secretions (secretomics) is an emerging noninvasive alternative since endometrial fluid aspiration in the same treatment cycle does not affect pregnancy rates (53)

DeCiDuAlizATiOn iS CRiTiCAl FOR PRegnAnCy SuCCeSSDuring decidualization in a normal pregnancy in mice the implanting blastocyst initiates changes in the surrounding stromal cells induc-ing stromal proliferation and differentiation into decidual cells (1) In humans the initiation of this process (predecidualization) does not require the presence of a blastocyst however the process becomes robust with implantation Predecidualization prepares the endome-trium for implantation and appears akin to the extensive stromal cell proliferation with expression of decidual markers seen prior to im-plantation in mice (1) Decidualization involves morphogenic mo-lecular and vascular modifications that are initiated by estrogen fol-lowing P4 priming Hoxa10 and Hoxa11 critical for patterning during development are also expressed in the uterus and have crucial roles in decidualization deletion of these genes produces compromised P4 signaling and fertility in mice (54ndash57) Morphologically deciduali-zation is characterized by elongated stromal cells transforming into enlarged round cells with specific ultrastructural modifications In hu-

mans this transformation is accompanied by the secretion of prolac-tin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1) (58) with changes in extracellular matrix proteins such as laminin type IV collagen fibronectin and heparin sulphate proteoglycans (59ndash61)

Proteomic analysis during human decidualization revealed 60 differentially expressed proteins including known markers (cathep-sin B) and new potential biomarkers (transglutaminase 2 and per-oxiredoxin 4) (59) Secretomics analysis during this process identi-fied 13 secreted proteins including established (IGFBP-1 and PRL) and novel (MPIF-1 and PECAM-1) markers (61)

APPROPRiATe TROPhOBlAST inVASiOn iS CRiTiCAl TO PlACenTATiOnTrophoblast invasion through the maternal decidua induces extra-cellular matrix (ECM) remodeling regulated by both the trophec-toderm and the stroma (62) Metalloproteinases (MMPs) play key roles in degrading ECM components (63) MMP-2 and MMP-9 are expressed and secreted by the human endometrium and participate in tissue remodelling during implantation and promote menstruation during normal cycles (64ndash67) Conversely decidual cells activate the expression of tissue inhibitors of MMPs (TIMPs) and downregu-late urokinase plasminogen activator (uPA) (65) It is thought that TIMPs limit ECM degradation by MMPs during decidualization re-stricting embryonic invasion A balance between these factors is crucial for a successful pregnancy outcome (1 68ndash70) Aberrant decidual display of ECM components is associated with preeclamp-sia or restricted intrauterine growth (71 72) It was also recently reported that histone acetylation derails the balance of ECM modu-lators inhibiting trophoblast invasion (73)

ADVAnCeS AnD ChAllengeSUterine transition through the prereceptive receptive and refrac-tory phases is dynamic and rapid Many genes show overlapping expression spanning more than one phase andor uterine tissue compartment making stage- and tissue-specific contributions of par-ticular genes difficult to ascertain with current tools For example Bmp2 which is expressed in the stroma during attachment and decidualization is considered to be critical for decidualization in mice (74 75) but its exact mechanism of action is still unknown Cre recombinase-expressing mouse lines have been used to de-lete or overexpress floxed genes but expression of these genes in multiple cell types from the uterus ovary and pituitary often limits interpretation of results Therefore there is still a need for the de-velopment of reproductive tissue- and cell-specific Cre lines Gen-eration of inducible conditional knockout mouse models could be valuable in addressing stage-specific effects of a gene with over-lapping expression patterns However the transient and dynam-ic expression of some genes makes the utility of such inducible systems questionable

Limited numbers of systems biological approachesmdashembracing genomics transcriptomics proteomics lipidomics and metabolo-micsmdashhave been used to study the intricate and dynamic changes underlying implantation These studies have produced a wealth of information but effective assimilation and rigorous validation of the results are lacking limiting their impact

Comparative research on implantation across species has become rare in recent years Studies contrasting different strategies andor hormonal requirements could help to identify common mechanisms underlying implantation better understand the causes of female infertility and improve pregnancy rates In particular the transition

Figure 3 Msx genes regulate uterine luminal epithelial cell polarity during receptivity Confocal images of E-cadherin and β-catenin colocalization Retention of the luminal epithelium (le) in Msx1Msx2dd mice at the site of blastocyst apposition on day 6 as opposed to luminal epithelium breakdown in Msx1Msx2ff mice with loss of E-cadherin Arrowhead indicates the location of blastocysts s stroma Scale bar 100 microm Reprinted from (20)

Figure 2 HB-EGF serves as a reciprocal mediator between the luminal epithelium and activated blastocyst during attachment reaction (A) In situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day four of pregnancy Note distinct hybridization signals in luminal epithelial cells surrounding two blastocysts in a longitudinal section Arrows demarcate location of blastocysts (B) Scanning electron microscopy of blastocysts co-cultured with 32D cells Zona-free day four (0900 hours) mouse blastocysts were co-cultured with (a) parental 32D cells (b) 32D cells displaying transmembrane form of HB-EGFTM or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours After extensive washing to remove loosely adhering cells blastocysts were fixed and examined by scanning electron microscopy Magnification 800X Arrows point to 32D cells (C) Bmp2 gene expression in response to beads preloaded with HB-EGF Beads preabsorbed either in BSA (control a) or HB-EGF (100 ngmL b) were transferred into uterine lumens of day four pseudopregnant mice Bmp2 expression at the sites of beads were examined on day five Arrows indicate the locations of the beads Note localized stromal expression of Bmp2 at the sites of beads preabsorbed with HB-EGF Bar 75 mm Reprinted from (12 13 74)

Produced by the ScienceAAAS Custom Publishing Office

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

48 iii

of the uterus from receptive to nonreceptive states requires special attention since the molecular interplay required remains largely unknown and may be critical to extend the WOI during IVF The complexities of pregnancy necessitate continued basic and clinical research since aberrant implantation leads to compromised pregnancy outcomes and may even impact the long term health of offspring

REFERENCES 1 J Cha X Sun S K Dey Nat Med 18 1754 (2012) 2 H Wang S K Dey Nat Rev Genet 7 185 (2006) 3 M Ruiz-Alonso D Blesa C Simon Biochim Biophys Acta 1822

1931 (2012) 4 D B Lubahn et al Proc Natl Acad Sci USA 90 11162 (1993) 5 J P Lydon et al Genes Dev 9 2266 (1995) 6 B Mulac-Jericevic et al Science 289 1751 (2000) 7 S Tranguch et al Proc Natl Acad Sci USA 102 14326 (2005) 8 Z Yang et al Mol Endocrinol 20 2682 (2006) 9 S Tranguch et al J Clin Invest 117 1824 (2007)10 H Yang et al Reproduction 143 531 (2012)11 H Xie et al Proc Natl Acad Sci USA 104 18315 (2007)12 S K Das et al Development 120 1071 (1994)13 G Raab et al Development 122 637 (1996)14 H J Yoo D H Barlow H J Mardon Dev Genet 21 102 (1997)15 O D Genbacev et al Science 299 405 (2003)16 C L Stewart et al Nature 359 76 (1992)17 H Song et al Mol Endocrinol 14 1147 (2000)18 J H Lee et al FASEB J 27 2553 (2013)19 X Sun Bartos A Whitsett J and Dey SK Mol Endocrinol Epub ahead of print (2013) doi101210me201320 T Daikoku et al Dev Cell 21 1014 (2011)21 H Lim et al Cell 91 197 (1997)22 H Lim et al Genes Dev 13 1561 (1999)23 H Wang et al J Biol Chem 279 10649 (2004)24 H Matsumoto et al J Biol Chem 277 29260 (2002)25 H Song et al Development 129 2879 (2002)26 X Ye et al Nature 435 104 (2005)27 B C Paria et al J Biol Chem 276 20523 (2001)28 H Wang et al J Clin Invest 116 2122 (2006)29 M Maccarrone et al Lancet 355 1326 (2000)30 A K Gebeh et al J Clin Endocrinol Metab 98 1226 (2013)31 R W Noyes A T Hertig J Rock Am J Obstet Gynecol 122 262 (1975)32 M J Murray et al Fertil Steril 81 1333 (2004)33 C Coutifaris et al Fertil Steril 82 1264 (2004)34 C R Murphy Cell Res 14 259 (2004)35 J C Martin et al Biol Reprod 63 1370 (2000)36 M Thie P Fuchs H W Denker Int J Dev Biol 40 389 (1996)37 M Thie et al Eur J Cell Biol 66 180 (1995)38 G Nikas Hum Reprod 14 Suppl 2 37 (1999)39 B A Lessey Fertil Steril 96 522 (2011)40 C E Quinn R F Casper Hum Reprod Update 15 229 (2009)41 B A Lessey et al J Clin Invest 90 188 (1992)42 M Meseguer A Pellicer C Simon Mol Hum Reprod 4 1089 (1998)43 S Kumar et al J Clin Endocrinol Metab 83 4443 (1998)

44 H S Taylor P Igarashi D L Olive A Arici J Clin Endocrinol Metab 84 1129 (1999)45 M Schena D Shalon R W Davis P O Brown Science 270 467 (1995)46 J A Horcajadas A Pellicer C Simon Hum Reprod Update 13 77 (2007)47 P Diaz-Gimeno et al Fertil Steril 95 50 e51-15 (2011)48 P Diaz-Gimeno et al Fertil Steril 99 508 (2013)49 M Ruiz-Alonso et al Fertil Steril Epub ahead of print (2013) doi 101016jfertnstert20130500450 T Parmar et al Fertil Steril 92 1091 (2009)51 F Dominguez et al Hum Reprod 24 2607 (2009)52 S Altmae et al Mol Hum Reprod 16 178 (2010)53 M H van der Gaast et al Reprod Biomed Online 7 105 (2003)54 H Lim et al Mol Endocrinol 13 1005 (1999)55 I Satokata G Benson R Maas Nature 374 460 (1995)56 H M Hsieh-Li et al Development 121 1373 (1995)57 A M Raines et al Development 140 2942 (2013)58 J C Irwin L de las Fuentes B A Dsupin L C Giudice Regul Pept 48 165 (1993)59 C L Dunn R W Kelly H O Critchley Reprod Biomed Online 7 151 (2003)60 S Tabanelli B Tang E Gurpide J Steroid Biochem Mol Biol 42 337 (1992)61 T Garrido-Gomez et al J Clin Endocrinol Metab 96 706 (2011)62 F van den Brule et al Chem Immunol Allergy 88 163 (2005)63 A Page-McCaw A J Ewald Z Werb Nat Rev Mol Cell Biol 8 221 (2007)64 W H Rodgers et al J Clin Invest 94 946 (1994)65 F Schatz et al Semin Reprod Endocrinol 17 3 (1999)66 L A Salamonsen G Nie Rev Endocr Metab Disord 3 133 (2002)67 S K Das et al Dev Genet 21 44 (1997)68 P K Lala C Chakraborty Placenta 24 575 (2003)69 M Knofler Int J Dev Biol 54 269 (2010)70 V Plaks et al Proc Natl Acad Sci USA 110 11109 (2013)71 J V Cockle et al Reprod Sci 14 629 (2007)72 C J Lockwood et al Biol Reprod 78 1064 (2008)73 C Estella et al PLoS ONE 7 e30508 (2012)74 B C Paria et al Proc Natl Acad Sci USA 98 1047 (2001)75 K Y Lee et al Mol Cell Biol 27 5468 (2007)

Acknowledgments Work of SKD and JC was funded by grants from the National Institute of Child Health and Human Development National In-stitute on Drug Abuse and National Institute of Aging of the US National Institutes of Health Work of CS and FV was funded by grants from the European Union FP7 People Programme namely the Marie Curie Actions Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM 324509) and through the Spanish Government (CDTI-EUREKA E6478 ndash NOTED and Ministerio de Economiacutea y Competitividad SAF2012-31017) and Valencia Government Generalitat Valenciana (Conselleria drsquoEducacioacute PROMETEOII2013018)

Produced by the ScienceAAAS Custom Publishing OfficeProduced by the ScienceAAAS Custom Publishing Office

Serono Symposia International Foundation | Improving the patients life through medical education

Are you finding it a challenge to keep up with thelatest developments in reproductive medicineRegistering with the SSIF reproductive medicinewebsite could be just the solution you need And it is absolutely freeAll you need to do to enjoy the wealth of informationthat SSIF provides about your specialist field is tovisit reproductive-medicineseronosymposiaorg andclick on the button to registerIt is quick and easy and you will have immediateaccess to bull Upcoming educational activities in reproductive

medicinebull Online video lectures by international expertsbull Online courses to increase your knowledgebull Slide and video presentations from SSIF symposiahellip And much more

follow us on

httptwittercomSSIF_RM

Take a look at the reproductive medicine section of the Serono SymposiaInternational Foundation (SSIF) website and register to access free e-learningactivities and receive updates on next live educational events and resources

Reproductive medicine section of the SSIF websitewwwreproductive-medicineseronosymposiaorg

SSIF_RM

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

iv viv v

Speaker (L) Renee A Reijo-PeraChallenger (R) Carlos Simoacuten

Main Presentation bitly1iuUGk1Challenger Response bitly1g1QE0q

Non-invasiveimagingofhumanembryosbeforeembryonicgenomeactivationpredictsdevelopmentto

theblastocyststage

Speaker (L) Martin M Matzuk Challenger (R) Antonio PellicerMain Presentation bitlyIInMfT

Challenger Response bitlyIDt5ws

Themammalianoocyteorchestratestherateofovarian

folliculardevelopment

Speaker (L) Sudhansu K DeyChallenger (R) Hilary OD Critchley

Main Presentation bitlyIIoMABChallenger Response bitly18dFTDn

AberrantCannabinoidsignalingimpairsoviductal

transportifembryos

Speaker (L) Christina BerghChallenger (R) Robert FischerMain Presentation bitly1jfJVjf

Challenger Response bitly1ciKKEISpeaker Response bitly1cW8tJ2

Electivesingle-embryotransferversusdouble-embryo

transferininvitrofertilization

Speaker (L) Richard T Scott JrChallenger (R) Carlos Simoacuten

Main Presentation bitlyIBTdrk Challenger Response bitly1bbRoN5

Accuratesinglecell24chromosomeaneuploidyscreeningusingwholegenome

amplificationandsinglenucleotidepolymorphismmicroarrays

Speaker (L) David S GuzickChallenger (R) Richard T Scott Jr

Main Presentation bitly1iuWf1iChallenger Response bitly1atpl7m

Spermmorphologymotilityandconcentrationinfertile

andinfertilemen

Speaker (L) S Ananth Karumanchi Challenger (R) Randy Levinson

Main Presentation bitly1c8zXJKChallenger Response bitly18X6y88

Circulatingangiogenicfactorsandtheriskofpreeclampsia

Speaker Ana CoboMain Presentation bitly1bFx3S2

Comparisonofconcomitantoutcomeachievedwithfreshand

cryopreserveddonoroocytesvitrifiedbytheCryotopmethod

Conference Videos Link to The Top Ten in Reproductive Medicine conference on the SSIF website here

SSIF_Advertising_210x267_Layout 1 141113 1708 Pagina 1