at SciVerse ScienceDirect

Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22

Contents lists available

Placenta

journal homepage: www.elsevier .com/locate/placenta

IFPA Meeting 2011 workshop report III: Placental immunology; epigenetic andmicroRNA-dependent gene regulation; comparative placentation; trophoblastdifferentiation; stem cellsq

W.E. Ackerman IV a, J.N. Bulmer b, A.M. Carter c, J.R. Chaillet d, L. Chamley e, C.P. Chen f, E.B. Chuong g,S.J. Coleman h, G.P. Collet i, B.A. Croy j, A.M. de Mestre k, H. Dickinson l, J. Ducraym, A.C. Enders n,N.M.E. Fogarty o, M. Gauster p, T. Golos q, S. Haider r, A.E. Heazell h, O.J. Holland e, B. Huppertz p,A. Husebekk s, R.M. John t, G.M. Johnsen u, C.J.P. Jones h, B. Kalionis v, J. König p, A.R. Lorenzonw,A. Moffett o, J.C. Moreira de Mello x, A.M. Nuzzo y, P. Parhamz, O. Parolini aa, M.G. Petroff bb, G. Pidoux cc,M.P. Ramírez-Pinilla dd, W.P. Robinson ee, A. Rolfo y, Y. Sadovsky d, H. Soma ff, J.H. Southcombe i,T. Tilburgs gg, G.E. Lash b,1,*

a Laboratory of Perinatal Research, Department of Obstetrics and Gynecology, The Ohio State University, Columbus, OH, USAbReproductive and Vascular Biology Group, Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UKc Institute of Molecular Medicine, University of Southern Denmark, Odense, DenmarkdMagee-Womens Research Institute, University of Pittsburgh School of Medicine, Pittsburgh, PA, USAeDepartment of Obstetrics and Gynaecology, Faculty of Medical & Health Sciences, The University of Auckland, New ZealandfDivision of High Risk Pregnancy, Mackay Memorial Hospital, TaiwangDepartment of Genetics, Stanford University, CA, USAhMaternal and Fetal Health Research Centre, School of Medicine, University of Manchester, UKiNuffield Department of Obstetrics and Gynaecology, University of Oxford, UKjDepartment of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, CanadakDepartment of Veterinary Basic Sciences, The Royal Veterinary College, London, UKl The Ritchie Centre, Monash Institute of Medical Research, Monash University, Melbourne, AustraliamDepartment of Basic Medical Sciences, Durban University of Technology, Durban, South AfricanDepartment of Cell Biology and Human Anatomy, University of California, Davis, CA, USAoCentre for Trophoblast Research, University of Cambridge, Cambridge, UKp Institute of Cell Biology, Histology and Embryology, Medical University of Graz, Graz, AustriaqDepartment of Comparative Biosciences, School of Veterinary Medicine, University of Wisconsin-Madison, Madison, WI, USArDepartment of Obstetrics and Gynecology, Reproductive Biology Unit, Medical University of Vienna, Vienna, Austrias Faculty of Health Sciences, University of Tromsø, Laboratory Medicine, University Hospital North Norway, Division of Immunology, Tromsø, NorwaytCardiff School of Biosciences, University of Cardiff, Cardiff, UKuDepartment of Obstetrics and Gynaecology, Oslo University Hospital-Ulleval, The Biotechnology Centre of Oslo and NCMM - Centre for Molecular Medicine Norway,Nordic EMBL Partnership, University of Oslo, NorwayvDepartment of Perinatal Medicine Pregnancy Research Centre and University of Melbourne Department of Obstetrics and Gynaecology, Royal Women’s Hospital,Melbourne, AustraliawDepartment of Cell and Developmental Biology, Institute of Biomedical Sciences, University of São Paulo, São Paulo, BrazilxDepto de Genética e Biologia Evolutiva, Instituto de Biociencias, Sao Paulo, BrazilyDepartment of Obstetrics and Gynecology, University of Turin, Turin, Italyz Stanford University School of Medicine, USAaaCentro di Ricerca E.Menni, Fondazione Poliambulanza, Brescia, ItalybbDepartment of Anatomy and Cell Biology, University of Kansas Medical Centre, Kansas City, KS, USAcc Inserm U767, Faculté de Pharmacie, Paris Descartes, Paris, Francedd Escuela de Biología, Universidad Industrial de Santander, Bucaramanga, ColombiaeeUniversity of British Columbia, Vancouver, British Columbia, Canadaff Saitama Medical School, JapanggDepartment of Stem Cell and Regenerative Biology, Harvard University, Cambridge MA, USA

q Working title: IFPA 2011 workshops III.* Corresponding author. Institute of Cellular Medicine, 3rd Floor, William Leech Building, Newcastle University, Newcastle upon Tyne NE2 4HH, UK. Tel.: þ44 191 222 8578;

fax: þ44 191 222 5066.E-mail address: g.e.lash@ncl.ac.uk (G.E. Lash).

1 GEL edited this manuscript based on contributions from the other authors.

0143-4004/$ e see front matter � 2012 Published by IFPA and Elsevier Ltd.doi:10.1016/j.placenta.2011.11.022

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22S16

a r t i c l e i n f o

Article history:Accepted 24 November 2011

Keywords:ImmunologyEpigeneticsComparative placentationTrophoblast differentiationStem cellsPlacenta

a b s t r a c t

Workshops are an important part of the IFPA annual meeting as they allow for discussion of specialisedtopics. At IFPA meeting 2011 there were twelve themed workshops, five of which are summarized in thisreport. These workshops related to various aspects of placental biology: 1) immunology; 2) epigenetics;3) comparative placentation; 4) trophoblast differentiation; 5) stem cells.

� 2012 Published by IFPA and Elsevier Ltd.

1. Placental immunology

Chairs: Larry Chamley and Tamara TilburgsSpeakers: B Anne Croy, Judith N Bulmer, Ted Golos, Anne

Husebekk, GuroM Johnsen, AshleyMoffett, Peter Parham,MargaretG Petroff, Jennifer H Southcombe, Tamara Tilburgs,

1.1. Outline

The placenta is one of nature’s great mysteries. Since the fetusand its placenta are derived from bothmother and father this organis in essence a tissue transplant. In an uncomplicated pregnancy theplacenta is a semi-allograft in which the paternally derived anti-gens are incompatible with the maternal immune system. Incontrast, in a donor oocyte or surrogate pregnancy the placentamay be a full allograft which is genetically unrelated to the hostmother. The immune paradox of pregnancy, in which the mothermust tolerate the placenta and fetus yet not succumb to infection,was first elaborated by Sir Peter Medawar more than half a centuryago. However, how much is actually known about how maternaltolerance is achieved and how mother and fetus survive whatshould be an immune conflict?

The workshop focused on three specific questions: 1) Areuterine natural killer (uNK) cells required for human pregnancy;2) Where and what is the maternalefetal interface: uterus, lymph,maternal blood?; 3) Antigen specific tolerance: can we generatetolerance - are there antigens to do this?

1.2. Summary

The Chairpersons opened the workshop by briefly presentinga case history of an abdominal pregnancy in a woman after totalhysterectomy. Although 22 such pregnancies had been previouslyreported, this was the first in which the fetus was live-born. Thisappropriately introduced the first question e “in the light of thiscase are uNK cells essential for pregnancy”?

Anne Croy reviewed immune deficient mouse strains with veryfew to no uNK cells but successful gestation, indicating that uNKcells are not essential for murine pregnancy. The benefit of uNKcells, defined by murine studies, lies in protection of maternal andfetal hearts from abnormal compensation to the elevated cardio-vascular demands of pregnancy. Judith Bulmer indicated that whilethe role of uNK cells in normal pregnancy has been intensivelyresearched, decidual macrophages and T lymphocytes are receivingmuch less attention. Uterine NK cells potentially interact directlywith extravillous trophoblast (EVT). In addition they secrete manycytokines, angiogenic growth factors and proteases, levels of whichdiffer with gestational age. In tubal pregnancy EVT invades thefallopian tube wall and arteries, ultimately resulting in tuberupture/haemorrhage. Tubal implantation sites contain T cells andmacrophages but, unless there is decidualisation of tubal mucosa,

uNK cells are absent. In placenta accreta, decidua (and uNK cells) isat least focally absent but trophoblast invades myometrium andarteries. Uterine NK cells are present in ectopic decidua in preg-nancy and in decidualised tissue. Successful extra-uterine preg-nancy may reflect areas of extra-uterine decidualisation with uNKcells in ectopic decidua interacting with EVT. Alternatively, there islikely to be redundancy within the system and other cell types mayfulfill uNK cell function at extra-uterine sites. Peter Parhamdescribed how interactions of killer cell immunoglobulin-likereceptors (KIR) with human leukocyte antigen (HLA) class I mole-cules regulate these functions and are specific to simian primates,where co-evolution of receptors with ligands is evident. Emergingin hominoids as superior KIR ligands, the C1 and C2 epitopes ofHLA-C are the only variable epitopes on trophoblast cells and thusable to bind uNK cell KIR. The combination of fetal C2 and maternalKIR A haplotypes correlates with pregnancy disorders that causedeath, non-competitive offspring, and barrenness. The selectionpressure they impose is writ in the impressive inverse correlationbetween C2 and KIR A frequencies worldwide. Human survival hasclearly benefited from the contribution of uNK cells. This does notrule out rare occurrences of human reproduction without uNK cellfunction, but suggests it would not be a competitive evolutionarystrategy in the long run. In some circumstances dramatic changecan be effective, as illustrated in gibbons that lost HLA-C, HLA-Gand all but one KIR. Such ‘simultaneous’ losses also point to theimportance of uNK cell receptors and ligands in human reproduc-tion. Ted Golos described how CD56þ/perforinþ uNK cells in therhesus monkey have morphological heterogeneity, including smallcompact cells, cells with expanded cytoplasm and compact perforinstaining, and large cells of amorphous morphology, with widelydispersed staining. Immunodepletion of peripheral blood NK cellsreduces uterine NK cell numbers, however abundant decidualKi67þ NK cells suggest that trafficking from peripheral blood to thedecidua only partly contributes to dramatic increases in thedecidua. Immunodepletion in the luteal phase increases endome-trial blood vessel wall thickness, and delays stromal maturation. NKcell function in the transition from secretory phase endometrium todecidua of pregnancy is targeted in this model.

The focus of the workshop then moved to the question “whereand what is the maternalefetal interface?” Jennifer Southcombediscussed the immunomodulatory role of syncytiotrophoblastmicrovesicles. In the later stages of pregnancy, the syncytio-trophoblast layer of the placenta is in direct contact with maternalblood. Vesicles less than 1 micron in size, syncytiotrophoblastmicrovesicles (STBM), are continually shed into the circulation.They engage B cells and monocytes, and trigger the release of, orinhibit, various cytokines and chemokines. By altering cytokineresponses the STBM may contribute to the changes in maternalimmunity that are important for a successful pregnancy. STBMextend the local placental syncytiotrophoblast-maternal bloodinterface as they can interact with immune cells systemically. Guro

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22 S17

Mørk Johnsen discussed the distribution of different T cell pop-ulations in decidua and maternal peripheral blood from normal3rd trimester pregnancy as determined by flow cytometry.The percentages of activated Tregs (CD4þCD45RA�FOXP3þþ)and cytokine-secreting T cells (CD4þCD45RA�FOXP3þ) wereincreased in decidua compared to blood, while resting Tregs(CD4þCD45RAþFOXP3þ) were more abundant in blood than indecidua. Both CD4þCD45RA�FoxP3� and CD8þCD45RA� memorycells were significantly increased in decidua compared to blood,whereas CD4þCD45RAþ and CD8þCD45RAþ naive Tcells weremoreabundant in the circulation. This suggests a role for Tregs in 3rdtrimester decidua in pregnancy.

The workshop then addressed the final question “Can wegenerate tolerance - are there antigens to do this?” Anne Husebekkreminded the workshop that fetal/neonatal alloimmune thrombo-cytopenia is caused by antibodies to fetal human platelet antigen 1ain greater than 80% of the cases. These antibodies cross the placentaand sensitize fetal platelets, which are phagocytosed, rendering thefetus at risk of bleeding. HPA-1a is an epitope on the b3 integrin ofthe fibrinogen receptor on platelets. b3 integrin is also present oninvading trophoblast cells as part of the vitronectin receptor.Immunization may occur during the first trimester. Recent datashow an association between maternal anti-HPA-1a antibody leveland reduced birth weight. As the vitronectin receptor carrying theHPA-1a epitope is also present on activated endothelial cells andtrophoblast cells, it is likely that these cells may be influenced bythe antibodies and reduced birth weight may indicate placentaldysfunction. Tamara Tilburgs discussed HLA-C, which is the onlypolymorphic classical major histocompatibility antigen expressedby fetal trophoblast cells. There is increased activation of decidual Tcells in pregnancies containing a fetal-maternal HLA-C antigenmismatch. However the increase in activated CD4þCD25dim T cellscoincides with an increase in functional CD4þCD25bright regulatoryTcells that can suppress this response. In addition, preliminary datashow the presence of Y chromosome specific CD8þ T cells indecidual tissue obtained from women carrying a male fetus. Theseobservations indicate that maternal T cells actively recognize andtolerate fetal alloantigens. The future challenge is to determinetheir role in pregnancy complications. Peggy Petroff reminded theworkshop that in pregnancy, T cells that recognize paternally-inherited major and minor histocompatibility antigens areexpanded. Less certain, however, are the cellular sources of theseantigens and the mechanisms by which they access maternallymphoid organs. It has been shown that the human placenta isreplete with minor histocompatibility antigens and that cellulardebris exiting the placenta also possesses these antigens. Thepresence of minor antigens in the trophoblast changes the way weshould view its immunologic potential: rather than regarding thesyncytiotrophoblast as immunologically inert due to the absence ofmajor histocompatibility complex (MHC), theymay instead accountfor the fetal antigen-specific response in women.

1.3. Conclusion

The first part of the workshop emphasized the unique role uNKcells play during pregnancy. No definitive conclusion was reachedas to whether or not uNK cells are essential for pregnancy; theirimportance may differ from species to species. The discussion anddata presented supported the idea that uNK cells may not benecessary in an individual pregnancy. However, uNK cells mayemerge as important and naturally selected components ofsuccessful pregnancy when considered on a long-term populationscale. In the second part of the workshop there was generalagreement that the fetal-maternal interface extends beyond thedecidua, to the lymph nodes, the syncytiotrophoblast layer and the

syncytiotrophoblast microvesicles that are continually shed intothe maternal circulation. Novel data was presented that bothextravillous trophoblast cells and syncytiotrophoblasts expresspaternally-inherited minor histocompatibility antigens and bothmay be able to elicit an antigen specific response by maternalimmune cells. The presence of decidual CD4þCD25brightFOXP3þ

regulatory T cells and their potential to generate antigen specificimmune suppression may position these cells as crucial immuneregulators during pregnancy. This data was mainly obtained inuncomplicated pregnancies and implications for pregnancypathology are unclear. The future challenge is to determine the roleof antigen specific immune recognition and regulation of fetaltrophoblast cells by maternal immune cells during pregnancycomplications.

2. Epigenetic and microRNA-dependent gene regulation

Chairs: Rosalind John and Yoel Sadovsky.Speakers: Richard Chaillet, Edward Chuong, Rosalind John, Joana

Moreira de Mello, Wendy Robinson, Yoel Sadovsky.

2.1. Outline

Exciting new researchfindings implicate epigeneticmechanismsas central regulators of placental development and function. Post-translational modifications of histones, nucleosome remodeling,changes in DNAmethylation, and the imprinted marking of specificparental alleles are different epigenetic mechanisms of transcrip-tional control that affect awide variety of cellular processes.Manyofthese mechanisms likely contribute to feto-placental develop-mental defects. Chromatin remodeling complexes and cytosinemethyltransferases facilitate epigenetic reprogramming during pre-implantation development. Correct reprogramming producesdevelopmentally competent embryos and trophoblast cells. Defectsin reprogramming, such as loss of critical genomic imprints, lead toplacental dysfunction as evidenced by specific changes in geneexpression of imprinted genes in the mouse placenta. Trans-generational inheritance of both normal and disease states may bemediated by maintenance of epigenetic modifications throughmeiosis. In addition, non-heritable expressionpatterns ofmicroRNAtranscripts modulate gene expression in the developing placenta,promoting stability of the transcriptomic and proteomic landscapeas well as providing a mechanism to dynamically respond to exog-enous stimuli. Many aspects of this diverse collection of epigeneticcontrols seem unique to the placenta, with mis-expressionpossibly contributing to diseases. The recent development of high-throughput analysis of chromatin modifications, RNA, microRNAand protein output, has enabled comprehensive determinations ofthe cellular epigenome in distinct physiological and pathologicalconditions. The identification of changes in this operational form ofthe epigenome may lead to a better understanding of underlyingcellular pathophysiology.

This workshop aimed to provide state-of-the-art presentationson DNA methylation, imprinting and microRNA-dependent generegulation in the placenta, and their impacts on placental devel-opment and fetal growth.

2.2. Summary

Richard Chaillet discussed the role of DNA (cytosine-5)-meth-yltransferase 1 (DNMT1) enzymes in imprinting mechanisms.Genomic imprinting is a process of gene regulation in whichparental alleles of imprinted genes are distinguished by DNAmethylation. Imprinted genes are formed through a developmentalreprogramming process. A cardinal feature of this process is the

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22S18

inheritance of unique gametic imprinted methylation patterns bythe action of maternal and zygotic DNMT1 cytosine methyl-transferase enzymes. These enzymes co-ordinately function duringpre-implantation to ensure the uniform presence of methylationimprints in all embryonic and extraembryonic cells.

Rosalind John discussed the epigenetic regulation of maternaladaptation to pregnancy. Placenta-derived factors act to modify thematernalefetal interface and to regulate maternal adaptation topregnancy for optimal fetal benefit. By directly regulating specificplacental lineages, genomic imprinting can modulate the supply ofnutrients to the embryo. For example, the imprinted Igf2 gene actswithin the labyrinth to regulate the diffusion exchange character-istics of the mouse placenta. Imprinting could also modulateadaptations to pregnancy by controlling cell types that secretepregnancy-specific factors into the maternal blood stream. Phlda2specifically regulates a key endocrine lineage of the mouseplacenta, the spongiotrophoblast, which produces many factorsthat are critical for successful pregnancy including prolactin-related hormones and pregnancy-specific glycoproteins. Conse-quently, Phlda2 may regulate adaptation to pregnancy by manip-ulating the amount of these factors secreted into the motherssystem. Esx1, Cited1 and Nrk, which also regulate this lineage, arematernally expressed by virtue of their location on the X-chromo-some, which is preferentially paternally inactivated in the mouseplacenta. It has recently been reported that the imprinted Ascl2gene negatively regulates both the spongiotrophoblast and theparietal trophoblast giant cell lineages of the mouse placenta. Thusderegulated imprints in the placenta can act directly to restrict thegrowth of the fetus through restricting the flow of nutrients butmay also cause inappropriate adaptations to pregnancy, withadditional adverse consequences for maternal health.

Yoel Sadovsky discussed the search for function of placentalmicroRNAs in relation to studies on trophoblast response to injury.MicroRNAs are germane for modulation of gene expression throughRNA degradation and inhibition of translation. Human trophoblastcells express common and unique microRNA species, but the func-tional targets of most microRNAs remain unknown. The search forrelevant targets of trophoblastmicroRNAs includes high-throughputexpression assays, in silico analyses for integration of complex datawith or without a priori assumption of target complementarity,reporter assays, and mutagenesis of microRNA binding elements.

Wendy Robinson discussed clonal patterns of placental devel-opment as determined by evaluating X-chromosome inactivationprofiles. The data support a model whereby each villous treerepresents a clone derived from one or very few progenitor cells.Trophoblast also shows distinct developmental origin from amnionand chorion, as expected. Comparing patterns of clonal growth inthe placenta to the distribution of other types of genetic andepigenetic variation can help distinguish the relative contributionof variation that arises in the pre-implantation embryo from thatarising predominantly after tissue differentiation.

Joana Moreira de Mello discussed a candidate gene for X-linkedgenomic imprinting in human full-term placenta. Genomicimprinting and X-chromosome inactivation are major epigeneticphenomena in mammals. Female humans display random X inac-tivation in every somatic tissue but the existence of imprintedgenes on the X, which is subject to transcriptional inactivation,remains enigmatic. Imprinted X-linked genes have been describedin mice, but there is only indirect evidence in humans. A human X-linked gene that undergoes genomic imprinting has been recentlyreported. This gene is not clustered with imprinted genes and themethylation pattern of one of its promoter’s CpG islands does notcorrelate with preferential maternal monoallelic expression. This islikely the first human X-linked gene that is subject to imprinting inthe term placenta.

Edward Chuong discussed the epigenomics of trophoblastdifferentiation. Chromatin immunoprecipitation and high-throughput sequencing (ChIP-Seq) was used to identify genome-wide epigenetic changes that occur during in vitro differentiationof mouse trophoblast stem cells to primary trophoblast giant cells.These data highlighted thousands of genomic regions that wereclassified into active/inactive regulatory elements based onenrichment of specific histone modifications. Focusing on regula-tory elements that become active in trophoblast giant cells mayprovide insight into how epigenetic modifications help establishtrophoblast lineage commitment.These types of analyses are beingextended to trophoblast derived from rat and rabbit to investigatethe evolutionary conservation of trophoblast gene regulation.

2.3. Conclusions

Deciphering the role of imprinting, epigenetic marks and posttranscriptional miRNA-based regulation of RNA and proteinexpression is clearly essential in order to understand normalplacental development and function. Moreover, placental tropho-blast cells exhibit unique types of epigenomic andmiRNA pathwaysthat do not exist in embryo-derived lineages. Recent developmentsin high throughput and sequencing technologies enable newinsights intomechanisms bywhich environmental stressors impacttrophoblast epigenomic mechanisms to induce placental injury.Aberrant patterns of DNA methylation or miRNA expression maysuggest biomarkers of placental dysfunction that contribute toclinically relevant diseases during pregnancy. Of equal importance,these advances allow us to gain a better insight into the role thatepigenetic and miRNA-related processes might have played inplacental evolution, resulting in the diversity of placental types thatexist in modern day mammals.

3. Comparative placentation

Chairs: Anthony Carter and Mandi de Mestre.Speakers: Anthony Carter, Hayley Dickinson, Allen Enders, Car-

olyn Jones, Mandi deMestre, Martha Ramirez-Pinilla, Hiroaki Soma.

3.1. Outline

The focus of this workshop was on placental diversity in verte-brates ranging from fish to mammals. Laboratory rodents anddomesticated mammals having been well covered at previousmeetings, the intentionwas towiden thenet to lesserknownspecies.

3.2. Summary

Amanda de Mestre explained that placentation in the horse(Equus caballus) had the unique feature of combining a non-invasive epitheliochorial placenta for nutrient and gaseousexchange with invasive trophoblast that performs endocrinefunctions. Invasive trophoblast cells establish themselves in theendometrial stroma in structures termed endometrial cups.Trophoblast cells survive despite a hostile uterine immunologicalenvironment.When transplanted to sites outside the uterus in non-pregnant mares, invasive trophoblast cells survive and function fora similar period to normal endometrial cups. This observation,together with in vitro studies of trophoblastelymphocyte interac-tions, suggests invasive trophoblast possesses innate immuno-modulatory capabilities essential to their survival. Furthermore,molecules specific to the uterine environment are not critical to thesurvival of invasive trophoblast.

Hayley Dickinson characterized the placenta of the spiny mouse(Acomys cahirinus), a small rodent species native to North Africa

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22 S19

and the Middle East. Precocial offspring are born after a 39-daygestation and litter sizes are small (1e4). The placenta is discoidhemotrichorial with emergence of the labyrinth at mid-gestation.The junctional zone is made up of glycogen cells, spongiotropho-blast, giant cells and large maternal blood spaces. Unlike conven-tional laboratory rodents, micro-CT imaging revealed a largevascular contribution to the fetal membranes originating from theumbilical vessels close to the fetal surface of the placenta. Sexdifferences in placental development have also been noted.

Anthony Carter introduced the Laotian rock rat (Laonastesaenigmamus), a relict species of rodent. First described in 2005, thisspecies belonged to a lineage basal to the hystricognaths that hadbeen thought to have gone extinct in the Miocene. Unlike in hys-tricognaths, the labyrinth was neither folded nor lobulated andthere was no subplacenta. These two characteristics had beenregarded as adaptations related to the reproductive strategy ofhystricognaths. Therefore it is noteworthy that Laonastes alsocarried a single fetus which was born in a precocial state. Theimplication might be that the reproductive strategy shared byLaonastes and hystricognaths evolved before the placental adap-tations characteristic of the latter group.

Allen Enders said the endotheliochorial placenta of the three-toed sloth (Bradypus tridactylus) was generally considered similarto that of the short tailed shrew (Blarina brevicauda) because of theperforated trophoblast. Both placentas have a highly irregularheterogeneous interstitial lamina and both have openings intointratrophoblast spaces from the maternal side. However maternaland fetal endothelia show several differences and the sloth did nothave complete discontinuities as seen in shrew trophoblast. It ispossible that such discontinuities might develop in the slothplacenta later in pregnancy since only a relatively early pregnancyspecimen had been examined by electron microscopy.

Evidence for trophoblast secretory activity in non-eutherianepitheliochorial placentae was presented by Carolyn Jones. Epi-theliochorial placentae from the Tammar wallaby (Macropus euge-nii) and the three-toed skink (Chalcides chalcides) exhibitmicrovillous surfaces on both the trophoblast and uterine epithe-lium that do not appear to interdigitate closely, unlike those ofeutherian epitheliochorial placentae. Trophoblast cells containingheavily glycosylated granules are present in both species and it isspeculated that some of thesemight be secretory in nature, possiblyfor the production of material to enhance intercellular contact.

Martha Ramírez-Pinilla said that the chorioallantoic placenta ofthe Andean skink (Mabuya sp.), a squamate reptile, exhibited thehighest placental complexity and the highest level of placento-trophy known among Reptilia, which meant that virtually all of thenutrients for embryonic development were obtained across theplacenta. A massive nutrient transfer (water, lipids, proteins, inor-ganic ions) was known to occur during embryonic development,especially when the chorioallantoic placenta was mature. Itpossessed areas functionally related to gas exchange and highlyspecialized structures related to nutrient transfer (placentome,paraplacentome, chorionic areolas, and absorptive plaques). Someof these morphological specializations showed an impressiveconvergence with those of eutherian mammals.

Hiroaki Soma compared placentation in the scallopedhammerhead shark (Sphyrna lewini) and the manta ray (Mantabirostris). The scalloped hammerhead shark has a gestationalperiod of approximately 1 year and bears about 20 pups nourishedby yolk sac placentation. The chorionic epithelium of hammerheadshark placenta contains immunopositive placental-like proteinswhen anti-human antibodies are used. A ductus vitellointestinaliswas positioned between two umbilical cord vessels. On the otherhand, the Manta ray has a 1 year long gestation but bears only onepup nourished by placental analogues such as trophonemata that

are characterized by crypts with vascularized villous strings. Thereare abundant secretory vesicles with the granules of proteins andlipids being released from the uterine epithelial surface into theuterine cavity for the pup.

3.3. Conclusions

Features of placental types that occur in nine animals rangingfrom the domestic horse to reptiles and fish were presented. Thisworkshop highlighted not only placental morphological diversitythat exists across species but also illustrated interesting examplesof convergent evolution at either the cellular or functional level.A number of species displayed specific placental and/or pregnancyadaptations in placental structure and gestation length that couldbe exploited to investigate basic mechanisms of placental function.For example, the spiny mouse has a longer gestation period whencompared with conventional laboratory rodents and could be uti-lised to study immunological responses in pregnancy that involvememory. In summary, expanding our understanding of structureand function to a wider range of placental types will result ina better appreciation of the evolutionary aspects of placentationand may provide a novel approach for future studies of placentalfunction and pathology.

4. Trophoblast differentiation and tools for studying it ewitha focus on human villous trophoblast

Chairs: Alex Heazell and Berthold Huppertz.Speakers: William Ackerman, Sarah Coleman, Gavin Collett,

Jennifer Ducray, Norah Fogarty, Martin Gauster, Sandra Haider,Olivia Holland, Aline Lorenzon, Guillaume Pidoux.

4.1. Outline

Trophoblast differentiation remains a major topic in placentalresearch. The last decade has seen a number of discoveries shed-ding new light on the different subtypes of trophoblast, includingvillous cytotrophoblast and syncytiotrophoblast, extravilloustrophoblast with its interstitial, intramural, endovascular andendoglandular subtypes plus the multinucleated trophoblast X-cells. Although it has been a major focus of placental research fordecades, regulation of trophoblast differentiation is still a mysteryand a large number of groups present an even larger number ofdifferent points of view on how trophoblast differentiates andwhatis the fate of the different subtypes. This is often due to differentmethodological approaches with one of the major issues in thisfield being the cell types and culture conditions used to analyzetrophoblast differentiation.

This workshop aimed to discuss the latest developments inachieving insight into the differentiation pathways of villoustrophoblast, including morphological changes of the cells, directinsights into molecular pathways, changes in expression andlocalization patterns of proteins, etc. The workshop was dividedinto three main themes: 1) cytotrophoblast proliferation, differ-entiation and apoptosis; 2) cytotrophoblast fusion; 3) syncytio-trophoblast differentiation and syncytial knots.

4.2. Summary

Sandra Haider discussed the role of Notch signalling in humanplacental development. Notch signalling is a highly conservedpathway controlling proliferation, cell death and differentiation.The localization of all Notch receptors (Notch1-4) and their ligands(Jagged1 and 2, DLL1, 3, 4) was investigated in first trimester humanplacenta. In addition, the role of Notch signalling in trophoblast

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22S20

proliferation and invasionwas studied. Notch receptors and ligandswere widely expressed in different trophoblast subtypes. Notch1and Notch4 were specific markers for a subset of proliferative cellcolumn trophoblast cells and villous cytotrophoblast, respectively.Inhibition of Notch signalling enhanced outgrowth of villousexplant cultures and stimulated cytotrophoblast migration. Addi-tionally, inhibition of Notch signalling induced proliferation of cellcolumn trophoblast cells and villous cytotrophoblast.

Aline Lorenzon discussed stromal cell derived factor-2 (SDF-2)and trophoblast differentiation. It has been shown that SDF-2 iswidely expressed in human placenta, although its function is stillunknown. Primary cytotrophoblast cells were cultured for 30 min,24 and 48 h for differentiation assays in 21% O2, and for 24 h in 21%O2 followed by 1 h or overnight in 2% O2. Immunoblottingdemonstrated that SDF-2 was downregulated in hypoxia, withhypoxia inducible factor (HIF) and b-hCG confirming hypoxia anddifferentiation conditions, respectively. The focus is to nowunderstand the role played by SDF-2 and its biological meaning atthe maternalefetal interface.

Martin Gauster discussed lysophosphatidylinositol as a putativeinducer of villous trophoblast differentiation. Bioactive lipidshave been suggested to be involved in placental growth anddifferentiation. One such lipid is lysophosphatidylinositol (LPI),a lysophospholipid generated by enzymatic hydrolysis of phos-phatidylinositol. The ability of LPI to induce villous trophoblastdifferentiation was tested in BeWo and ACH-3P cell lines. LPIinduced increased b-hCG secretion, downregulated expression ofE-cadherin and impaired BrdU incorporation. These results suggestLPI as an endogenous inducer of villous trophoblast differentiation.The underlying mechanisms have not been defined so far, but thedetection of the LPI receptor, a G-coupled-protein receptor, inhuman placental tissues and primary trophoblast suggestsa possible role for receptor mediated signalling.

Jennifer Ducray discussed increased cytotrophoblast fusion inpre-eclampsia. Using term placentae from normotensive andpre-eclamptic pregnancies, double immunofluorescence wasemployed to compare the distribution of cytotrophoblast cellstogether with staining of these cells for cleaved caspase 8. Initialfindings suggest a higher rate of cleaved caspase 8 positive cyto-trophoblast cells in the pre-eclamptic cohort. If not matched bydownstream caspase activity, a dysfunctional trophoblast turnovercould result. In addition, punctate autofluorescence was prevalentin the syncytiotrophoblast of pre-eclamptic samples. It is temptingto speculate that these could be lipofuscin-like material resultingfrom inadequate autophagic processes of trophoblast turnover inpre-eclampsia which in turn may result in necrotic shedding.

Guillaume Pidoux discussed the role of ezrin and gap junctionsin trophoblast fusion. A limited number of human cell types canfuse to form a multinucleated syncytium. In the differentiation ofhuman placenta, mononuclear cytotrophoblasts aggregate and fuseto form a syncytiotrophoblast. The cAMP/phosphokinase A (PKA)signaling pathway is indispensable for syncytial formation and inparticular a role for the AKAPs (A-kinase anchoring proteins) hasbeen implicated. Ezrin was shown to be an AKAP required for hCGregulation of cellecell fusion. Ezrin organizes a supramolecularcomplex containing PKA, connexin 43 and zona occludens 1 (ZO1),and targets PKA to connexin 43 and controls its phosphorylation toenhance gap junctional communication, which is a prerequisite fortrophoblast cellecell fusion.

William Ackerman discussed cellular events involved in syncy-tialization. The syncytiotrophoblast is a remarkably specializedepithelium that regulates many functions crucial to the humanplacenta. Following fusion of a progenitor cytotrophoblast with thesyncytiotrophoblast, the newly merged cellular organelles undergodramatic modifications. To gain initial insights into the complicated

cellular events that accompany syncytialization a BeWo model hasbeen used. These cells can be induced to undergo fusion events that,in many respects, recapitulate those that occur in vivo. Dramaticstructural and biochemical alterations in a number or organelles,including the Golgi/trans-Golgi system, endocytotic pathways,mitochondria, and endoplasmic reticulumwere discussed.

Gavin Collett discussed the role of RhoE in regulating fusion ofBeWo choriocarcinoma cells. The role of Rho GTPases in trophoblastdifferentiation and fusion is not known. Treatment of BeWo cellswith dibutyryl cyclic AMP (dbcAMP) resulted in a protein kinaseA-dependent upregulation of RhoE concomitant with cell fusion.Silencing of RhoE expression by siRNA inhibited dbcAMP-inducedfusion but had no effect on functional differentiation, as deter-mined by expression of human chorionic gonadotrophin andplacental alkaline phosphatase. Upregulation of RhoE by dbcAMPwas attenuated under hypoxic conditions in which cell fusion wasimpaired.

Norah Fogarty discussed differences between syncytial knotsand sprouts. It is pertinent that features of syncytial knots andsprouts are defined so that researchers can distinguish betweenthese functionally different regions. In serial sections of normalterm placenta, syncytial nuclear aggregates were traced through toidentify sprouts, false knots and true knots. Sprouts containedPCNA-positive nuclei whereas knots did not. Knots containednuclei with heterogeneous DNA damage. Nuclei that were 8’OHdGor TUNEL-positive also displayed morphological features ofdegeneration, including a dramatically smaller nuclear area anda heavily condensed ring of DNA around the periphery. These datasuggest that knots may be formed by shepherding of aged nuclei,which may drag in functional nuclei as collateral. Alternatively,functional nuclei are gathered into knots prior to condensation oftheir chromatin.

Sarah Coleman discussed nuclear condensation in syncytialnuclear aggregates. Syncytiotrophoblast nuclei show variableamounts of heterochromatin, hypothesised to be more condensedin syncytial nuclear aggregates. Heterochromatin was analysed inan archive of electron micrographs. Nuclei in syncytial nuclearaggregates were more often pyknotic than those in syncytio-trophoblast; one possibility is that as nuclei condense they aregathered into syncytial nuclear aggregates. The frequency ofapoptosis was assessed in different types of syncytial nuclearaggregates by immunohistochemistry of the cytokeratin M30neoepitope. Knots weremore likely to beM30 positive than bridgessuggesting that these syncytial nuclear aggregate structures mayhave different origins and functions.

Olivia Holland discussed the morphology of syncytial nuclearaggregates released from first trimester syncytiotrophoblast.Syncytial nuclear aggregates shed from first trimester placentaein vitro were analysed by three-dimensional reconstruction ofconfocal Z-stack images. Morphologies, dimensions and nuclearcontent broadly agreed with the limited published in vivo esti-mates. Fifty-five percent of the syncytial nuclear aggregatesexhibited heterochromatic nuclei and/or blebbing of the plasmamembrane, suggesting they were undergoing programmed death,but almost half of the syncytial nuclear aggregates had euchromaticnuclei and some possessed nucleoli. The observed differing grossand nuclear morphologies of syncytial nuclear aggregates mayreflect different points along a continuum of programmed celldeath. Alternatively, they may represent structures of entirelydifferent origins.

4.3. Conclusions

This workshop highlighted the whole spectrum of villoustrophoblast differentiation. Within the cytotrophoblast factors

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22 S21

inducing differentiation such as stromal cell derived factor-2(SDF-2) and lysophosphatidylinositol were described as potentialnewplayers, while the Notch signaling pathwaywas shown to forcedifferentiation. Prior to fusion activity of caspase 8 still remains animportant step that may be accelerated in cases suffering from pre-eclampsia. During fusion ezrin, gap junctions and RhoE play crucialroles while at the same time the organelles of fusing cells undergodramatic morphological changes. On the level of the syncytio-trophoblast from term placenta, an important step towardsunderstanding of morphological structures has been made. It isbecoming more accepted that syncytial sprouts are means for thegeneration of new villi; this is why theymay contain nuclei that arestill PCNA positive. On the other hand, syncytial knots aremeans forthe accumulation of aged and late-apoptotic nuclei, which may bereleased into the maternal circulation. Interestingly, syncytialstructures released from first trimester placenta show a newphenotype that needs to be further elucidated. In conclusion,expanding our knowledge of factors, pathways, players and struc-tures in combination with relating findings from cell lines toprimary cells may provide a way to understand trophoblast turn-over in health and disease.

5. Stem cells

Chairs: Bill Kalionis and Ornella Parolini.Speakers: Chie-Pein Chen, Edward Chuong, Bill Kalionis, Julia

König, Anna Maria Nuzzo, Ornella Parolini, Alessandro Rolfo.

5.1. Outline

The workshop on placental stem cells was conceived as anopportunity to consider the placenta not only for its key role insupporting life in the womb, but also as a possible source of stem/progenitor cells with the potential to be exploited for regenerativeand reparative medicine. In recent years great strides have beenmade in understanding the basic biology of placenta-derived stemcells, as well as their differentiation potential and immunologicalcharacteristics. Encouraging results have been obtained by inves-tigating placental cell-based therapy in pre-clinical animal modelsbut much more work needs to be done in this area. This workshoptherefore represented an opportunity for interaction and discus-sion of the many unanswered questions regarding placental stemcells. These include the need for better definition of the cells beingused in terms of their exact location in the placental tissues,phenotype, stem cell potential, as well as the need to havecomparative studies to verify whether placental cells are betterthan those isolated from other sources in terms of therapeuticapplicability. Specific discussion points included: 1) which cells arewe identifying in placenta?; 2) do these cells all have the samepotential, can we call them stem or progenitor cells?; 3) what istheir differentiation capability?; 4) are immunomodulatoryfeatures the same for stem/stromal cells from different parts of theplacenta?

5.2. Summary

Ornella Parolini discussed paracrine effects of amnioticmembrane and amnion-derived cells. Transplantation of fetalmembrane-derived cells, or conditioned medium from these cells,reduces lung fibrosis in bleomycin-challenged mice. Amnioticmembrane patching promotes ischemic rat heart repair andreduces liver fibrosis in the BDL rat model. These positive outcomeswere observed despite low or absent levels of donor cells in hosttissues, suggesting that amniotic membrane-derived cells act bysecreting paracrine soluble factors. An emerging concept from the

study of transplanted placental cells is that important therapeuticactions result from the immunomodulatory effects they exerton the surrounding host environment, rather than theirdifferentiation.

Bill Kalionis discussed migration of placental mesenchymalstem cells in a placental vessel perfusion model. Human placentalmesenchymal stem/stromal cells (pMSCs) are migratory cells butpMSC migration in the placenta is poorly understood. A trans-formed stem cell line (called PRCFP) was generated from primarypMSCs. PRCFP retained important stem cell properties includingmultipotent differentiation and migration. PRCFP cells werelabelled with a live-cell fluorescent stain, perfused into placentalvessels and their migration across the endothelial cell barrierassessed. PRCFP cells crossed the endothelium of the vessel butremained within the perivascular stem cell niche and did notmigrate into the stroma. It was concluded that pMSC migration istightly regulated by the perivascular stem cell niche.

Chie-Pein Chen discussed how human placental multipotentmesenchymal stromal cells may modulate trophoblast migrationvia Rap1 activation. The Rap1 regulatory pathway was examined inhuman trophoblast cells after stimulation with hepatocyte growthfactor (HGF) secreted by human placental multipotent mesen-chymal stromal cells (hPMSCs). HGF induced trophoblast cAMPproduction with the downstream pathways mediated through PKAand Epac1 to trigger trophoblast Rap1 and integrin b1 activation.The total trophoblast cell integrin b1 protein did not change, but thecell surface active form of integrin b1 was significantly increased.hPMSC conditioned medium enhanced trophoblast migration,which was decreased by c-Met blocking antibody or conditionedmedium containing lower levels of HGF from hPMSCs cultured inhypoxia. Thus, trophoblast cells may acquire signals for migrationfrom villous hPMSCs.

Alessandro Rolfo discussed increased production of macrophagemigration inhibitor factor and other pro-inflammatory cytokines bypre-eclamptic placental mesenchymal stem cells. Placenta-derivedmesenchymal stem cells (PDMSCs) are well renowned for theirunique anti-phlogistic activity. Since pre-eclamptic placentae arecharacterized by exacerbated inflammation and increased cytokineproduction, it was hypothesized that pre-eclamptic-PDMSCspossess altered anti-inflammatory properties. By cytokine array,it was demonstrated that there was increased release of pro-inflammatory mediators by pre-eclamptic-PDMSCs versuscontrols. In particular, significantly increased levels of pro-inflammatory MIF (macrophage migration inhibitory factor) andTNF-a (tumor necrosis factor a) were reported. Moreover, it wasfound that explants treated with pre-eclamptic-PDMSC condi-tioned media showed significantly increased MIF expression rela-tive to controls. These data suggest that pre-eclamptic-PDMSCspossess an aberrant pro-inflammatory action.

Anna Maria Nuzzo discussed characterization of activatorprotein-1 (AP-1) family gene expression in PDMSCs derived fromnormal and pre-eclamptic pregnancies. Mesenchymal stromal cellscontrol proliferation and apoptosis in neighbouring cells. Less isknown about PDMSCs and their contribution to proper trophoblastdevelopment. It was found that there is differential expression ofAP-1 family genes, key cell fate regulators, in PDMSCs and placentalchorionic tissue from normal and pre-eclamptic pregnancies,characterized by excessive trophoblast apoptosis. Pre-eclamptic-PDMSCs presented increased pro-apoptotic JunB expressionversus controls, while pre-eclamptic placentae showed decreasedanti-proliferative JunD and increased pro-inflammatory Fra1 levels.These data suggest native differences in AP-1 gene expressionamong healthy, pre-eclamptic-PDMSCs and neighbouring chorionictissue, thus indicating a contribution of PDMSCs to physiologicaland pathological placenta development.

W.E. Ackerman IV et al. / Placenta 33, Supplement A, Trophoblast Research, Vol. 26 (2012) S15eS22S22

Julia König discussed secretion of angiogenic growth factorsby amnion-derived mesenchymal stromal cells. To investigatewhether mesenchymal stromal cells (MSC) from the amnion ofhuman term placenta (hAMSC) have angiogenic properties similarto bone marrow-derived MSC (BMSC), the paracrine effects ofhAMSC on network formation and viability of endothelial cellswere examined. hAMSC-conditioned medium stabilized endothe-lial networks in the Matrigel assay and enhanced endothelialviability as shown by a lactate dehydrogenase (LDH) assay. Proteinanalysis of hAMSC-conditionedmediumwith an angiogenesis arraykit revealed secretion of pro-angiogenic factors such as angiogeninand IL-8. This suggests that hAMSC might be a useful alternative toBMSC for therapeutic applications in vascular biology.

Edward Chuong discussed evolutionary aspects of trophoblaststem cell biology. The placenta has undergone significantmorphological diversification during eutherian evolution. Asa consequence, it has proven difficult to elucidate a conserved set ofgenes that establish trophoblast identity in all mammals. RNA-Seqwas performed on the earliest trophoblast population, trophoblaststem cells, from mouse, rat, rabbit and rhesus macaque. Bycomparing the trophoblast stem cell transcriptomes of multiplemammals, it is possible to focus on which genes may have beenexpressed in the ancestral trophoblast, and have remainedconserved throughout eutherian evolution.

5.3. Conclusions

This workshop highlighted the varied research into placentalstem cells of varying origin. In particular human placental mesen-chymal stem cells were shown to be a rich source of cytokines andangiogenic growth factors that may play roles in placental vascularstability and in regulating trophoblast invasion. In addition, theirproperties may be altered in pre-eclampsia, although whether thisis a cause or effect of the condition is not clear. In animal modelsmembrane-derived stem cells appear to play roles in immunomo-dulation of surrounding host tissue after transplantation.Trophoblast-stem cells also provide a species comparator forevolutionary studies. The encouraging data provided in this work-shop, although preliminary, prompts further investigations fromboth stem cell researchers and placental biologists. Stem cellresearchers need a better understanding of the basic biology ofplacental stem cells in order to assess, and improve, their potentialutility in future clinical settings. Placenta biologists need to furtherexplore these cells because abnormalities in their functions arelikely to contribute to clinically significant placental pathologies.

Conflict of interest statement

None of the authors have any conflict of interest to declare.