Leading Edge

In This Issue

Residing in the physical heart of the cell, the nucleus has now fullyshed its once one-dimensional reputation as the repository forgenetic information and steady supplier of messages to the cyto-plasm. This sea change toward a more dynamic view of nuclearbiology has been led by a revolution in our understanding ofgenomes, regulatory elements, epigenetic modifiers, and nuclearspatial organization, which has enabled the examination of theinterplay of these factors both genome wide and at individualloci. From this vast and rapidly expanding body of work, we can

now see that, in the cellular orchestra, the nucleus is not simply the sheet music from which thenotes are played but is instead the conductor and composer, both writing the songs and leadingthe performance. This special issue, The Dynamic Nucleus, explores the meaning behind themaestro’s motions—both dramatic and nuanced—in directing the cell symphony.

Please Put on Your 3D GlassesWe open the issue by posing the question of what has been the most surprising revelation about the nucleus in recentyears. The responses of Job Dekker, Joanna Wysocka, Iain Mattaj, Erez Lieberman Aiden, and Craig Pikaard rangefrom the pervasive roles of RNA to mapping genomic ‘‘connectomes’’ and appreciating the marvel of nucleocytoplas-

mic transport (Voices, page 1207). Many of the major themes of The DynamicNucleus are then framed in an Essay from Tom Misteli (page 1209), whoplaces genome function in the context of cell biology, discussing the impactof stochasticity, epigenetics, and nuclear organization while juxtaposing thisemerging complexity with the simple elegance of the structure of DNA on the60th anniversary of its discovery.If the intervening years have taught us anything, it is that knowing the sequence

of DNA is just the starting point for understanding gene regulation, with the crit-ical importance of epigenetic modification and higher-order chromatin structurebecoming ever more apparent. In their Review, Wendy Bickmore and Bas vanSteensel (page 1270) provide an introduction to how local and long-rangecontacts between genes shape the three-dimensional organization of interphasechromosomes. Amajor driver of nuclear architecture, the formation of chromatinloops, is examined in depth by Duncan Odom and Matthias Merkenschlager(Review, page 1285), who focus on two proteins, cohesin and CTCF, whichare frequently involved in bringing together gene regulatory elements with theirtargets. Pedro Batista and Howard Chang grapple with an emerging facet of

nuclear organization in their Review (page 1298), proposing that long noncoding RNAs act as address codes directingprotein complexes, genes, and chromosomes to their appropriate locations. The formation of chromatin loopsand other modes of protein and RNA-mediated gene regulation are further featured in this issue’s Select, penned byMolecular Cell’s Brian Plosky (page 1203).

Message ControlOf course, the main nuclear preoccupation is transcription. Tong Ihn Lee and Richard Young (Review, page 1308) takeon the challenge of synthesizing the vast literature on transcriptional control, highlighting recent advances in under-standing the impact of transcription factors, cofactors, and chromatin regulators and delving into the misregulationof transcriptional control in disease states. Bernadett Papp and Kathrin Plath (Review, page 1323) similarly delveinto the evidence that chromatin regulators serve as gatekeepers of cellular reprogramming to induced pluripotency.Initiating transcription is just the first of many steps in the making of a messenger RNA fated for cytoplasmic ribo-

somes. In their Review, Blencowe and colleagues (page 1252) place RNA-splicing events within the complex spatialand temporal dynamics of the nucleus and examine the layers of crosstalk between splicing and chromatin contextand transcriptional regulation. Gene regulation by RNA also lies at the heart of the allelic control of large gene clusters,as observed in imprinting and X chromosome inactivation, the processes highlighted in a Review by Jeannie Lee andMarisa Bartolomei (page 1308).

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1199

When Bad Things Happen to Good NucleiLike transcription, maintaining genome integrity is a particular nuclear obsession.How the chromatin landscape changes in response to DNA damage and how thisthen impacts DNA repair processes is the focus of the Review by Brendan Priceand Alan D’Andrea (page 1343). On a different spatial scale, Vincent Dion andSusan Gasser (Review, page 1354) discuss how the chromatin environmentaffects the long-range mobility of DNA, with double-strand breaks increasingmovement and thus impacting genome stability and the efficiency of repair.Yet despite a cell’s best efforts, things sometimes go wrong, with a dramatic

example being chromothripsis, in which many chromosomal rearrangementsoccur in a single catastrophic event. In their Primer, Jan Korbel and Peter

Campbell (page 1226) enumerate the criteria for inferring the occurrence of chromothripsis from cancer genome-sequencing data. Luckily, however, DNA repair and replication proceed most of the time with exquisite precision,and for a view on the elegant spatial and temporal dance of DNA replication, you should check out the SnapShot byDavid Gilbert and colleagues (page 1390).

The Outer Limits and BeyondThe boundary of the nucleoplasm is the nuclear lamina, which is a unique andcomplex nexus for gene regulation. Recent studies reveal that associationsbetween chromatin and the nuclear lamina differ between cell types, a topictackled by Kevin Van Bortle and Victor Corces (Minireview, page 1213) who arguethat these interactions are critical determinants of cell fate and identity. The func-tion of the lamin proteins at the nuclear envelope likely reaches far beyond chro-matin organization, as mutations in these proteins give rise to a group of diversesyndromeswith symptoms ranging frommuscular dystrophy to progeria. Kather-ine Schreiber and Brian Kennedy review these laminopathies (Review, page1364), examining the connections between nuclear envelope dysfunction andaltered nuclear activity and commenting on how emerging molecular insightsmight be translated into new treatments.In vertebrate cells, the nuclear envelope retracts into the endoplasmic retic-

ulum at the onset of mitosis and subsequently reforms in daughter cells. CorneliaWandke undUlrike Kutay (Minireview,page 1222) parse the mechanisms behind this dramatic breakdown and reassembly. However, the nuclear envelope ismore than just a barrier that holds in nuclear contents. It also serves as a highly selective yet very rapid and efficient filterfor the trafficking of RNA and protein via nuclear pores that are essential for getting cellular components where theyneed to be, thereby maintaining the distinct properties of the nucleoplasm and cytoplasm. Understanding how thisis accomplished is examined by Rebecca Adams and Susan Wente (Minireview, page 1218), who guide us throughthis remarkable feat of selective transport, opening our eyes to nuclear pore complexity as revealed by high-resolutionstructural analysis and imaging.Although often depicted as the center of a cell, the position of the nucleus can in fact vary extensively, with important

functional consequences for cell signaling andmigration. Gregg Gundersen and HowardWorman (Review, page 1375)describe the cytoskeletal forces that position the nucleus and present evidence linking disruptions in these forces tonumerous diseases.

Creating this special issue has relied upon involvement frommany dedicated authors and reviewers, and we would liketo thank them for their time, effort, and insights. We hope that, in reading The Dynamic Nucleus, you will come awaywith an appreciation for the processes that contribute to nuclear complexity, inspiration from how much has beenlearned in recent years, and motivation to pursue answers to the many remaining mysteries of nuclear biology.

Robert P. Kruger

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1201

Leading Edge

Select

The Nucleus: Express Yourself

As the home for the bulk of eukaryotic genomes, the nucleus is a perfect venue for observingintricate regulatory mechanisms involving interactions between DNA, RNA, and protein. Theseinteractions have evolved beyond the original central dogma concept, and the papers featured inthis Select highlight some recent and exciting examples of these mechanisms.

The negative regulation of geneexpression by CpG sequences can

be through cytosine methylation (left)

or through the recruitment of a

polycomb-repressive complex to un-

methylated CGIs. Image courtesy of

X. Wu and K. Helin.

WhyDoes Polycomb Like toHangOut in the Islands?Methylation of cytosine in CpG sequences is an epigenetic modification of DNA that is asso-ciated with a repressive chromatin state. Large clusters of CpG sequences known as CpGislands (CGIs) tend to be unmethylated and are found near active transcription start sites.However, a subset of CGIs corresponds to binding regions for the repressive polycombcomplex in mammals. In the past couple of years, it has become clear that certainchromatin-modifying activities are directed to nonmethylated CGIs. So far, each of theseactivities has been associated with active transcription. The Set1 histone H3 lysine 4 meth-yltransferase complex subunit CFP1 and KDM2A, a histone H3 lysine 36 demethylase, bothcontain a zinc finger CxxC DNA-binding domain (ZF-CxxC) that preferentially bindsunmethylated CGIs. In two separate studies, Robert Klose and Kristian Helin and theircolleagues looked at the binding and function of KDM2B (a paralog of KDM2A also knownas FBXL10) to see whether it binds CGIs like KDM2A. The ChIP-seq results from both groupsshow that, like KDM2A, KDM2B binds to CGIs across the genome, and Klose and colleaguessee that KDM2B is more robustly represented at CGIs that are bound by the polycomb-repressive complex, PRC1, which ubiquitinates histone H2A at lysine 119. It was previouslyshown that KDM2B can be part of an alternative PRC1 complex, and both groups find that

KDM2B binds a specific variant of PRC1 via a subunit called NSPc1 (or PCGF1) and that the ZF-CxxCmotif of KDM2B is responsiblefor targeting these complexes to unmethylated CGIs. The ZF-CxxC motif of KDM2B is important for histone H2A ubiquitination, andloss of KDM2B reactivates a subset of polycomb targets. Helin and colleagues go on to show that KDM2B, like other PRC1 compo-nents, is important for proper differentiation of mouse embryonic stem cells. These findings add to the complexity of polycombtargeting in mammals, which lack a defined polycomb response element, and support an emerging concept that PRC1 targetingis not always dependent on the histone H3 lysine 27 trimethylation activity of PRC2.Farcas, A.M., et al. (2012). eLIFE. Published online December 18, 2012. http://dx.doi.org/10.7554/eLife.00205.Wu, X., et al. (2013). Mol. Cell. Published online February 7, 2013. http://dx.doi.org/10.1016/j.molcel.2013.01.016.

In female human embryonic stem cells,

XIST RNA (green) covers the inactive X

and XACT (red) coats the active one.

Image courtesy of C. Vallot.

How Does Human X Inactivation Work? I’m NoteXACTly SureIn addition to the protein complexes that regulate gene expression, we know that non-coding RNAs can be involved in targeting regulatory complexes to chromatin. Whileanalyzing RNA-seq data from a female-derived human embryonic stem cell line (hESC),Claire Rougeulle and colleagues found a previously unannotated region of the X chromo-some that was actively transcribed. The resulting 251.8 kb, unspliced, polyadenylatedtranscript is primarily nuclear, bearing the hallmarks of a long noncoding RNA (lncRNA).The X chromosome is not a stranger to lncRNAs, as two such RNAs, XIST and TSIX, arewell known to have a role in the mammalian version of sex chromosome dosage compen-sation, X chromosome inactivation (XCI). So what might be special about another lncRNAexpressed from the X chromosome? Could it have a role in XCI? XIST is known to coat theinactive X and to maintain it in an inactive state. However, this newly identified lncRNA isfound to be associated with the active X chromosome, which led to the name of XACT(X active coating transcript). Interestingly, XACT is downregulated upon differentiation ofhESCs—nearly completely silent after 10 days—and is either weakly expressed or not

detectable in adult tissues. However, upon conversion of mesenchymal stem cells to induced pluripotent stem cells, there is strongre-expression of XACT. Also, XIST appears to prevent XACT expression from the inactive X. So what does this lncRNA do? Thatremains to be determined. It is not the first case of a lncRNA being associated with an active chromosome in a dosage compensationprocess. In fruit flies (where females do not inactivate either X chromosome), the lncRNAs rOX1 and rOX2 are critical components ofa histone acetyltransferase complex that is responsible for upregulating the single X chromosome inmales. Whether XACT is respon-sible for targeting a chromatin-modifying activity or has some other function remains to be seen. If you are thinking about checkingthis out in somemouse embryonic stem cells, it might not beworth the effort because it looks like expression ofXACTmay be a recentevolutionary event and is seen only in human cells.Vallot, C., et al. (2013). Nat. Genetics 45, 239–241.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1203

Isw2 is targeted to canonical sites

via physical interactions with Ume6. In

contrast, Isw2 is targeted to ectopic sites

via Ume6- and TFIIB-dependent DNA

looping. A light-blue oval representsthe transcription preinitiation complex.

Image courtesy of T. Tsukyama.

A Remodeler Gets Drawn into a LoopIn addition to the interactions of regulatory molecules with arrays of nucleosomal DNA, it isimportant to consider the impact of higher-order organization such as looping. Typically,ATP-dependent chromatin-remodeling complexes are targeted through either specificinteractions with modified histone tails or interactions with specific transcription factors(TFs). When asking whether the targeting of yeast remodeler Isw2 is transcription factordependent, Toshio Tsukiyama and colleagues stumbled upon something unusual.Although they did see that Isw2 targeting overlapped and depended on transcriptionfactors, curiously, many of these sites seemed to lack annotated TF-binding sites. In thecase of the yeast transcription factor Ume6, nearly 90% of the 563 Isw2 binding sitesthat are Ume6 dependent were devoid of Ume6-specific binding sites. To examine themechanism by which this targeting could happen, they turned back to their Isw2 ChIP-chipdata and noticed that the remodeler often bound at 5’ and 3’ ends of the same gene,reminiscent of the binding pattern of TFIIB when it forms gene loops. Using chromosomeconfirmation capture (3C) and yeast genetics, Tsukiyama and colleagues were able toshow that DNA looping targets Isw2 to specific loci and that Ume6 and TFIIB are necessaryfor formation of these loops and also for the repression of Ume6 target genes. Interestingly,all previously known examples of gene looping in yeast are associated with transcriptional

activators. This is the first time that a nucleosome remodeler has been shown to be targeted by looping, and it will be interesting tosee whether other related enzymes might be targeted similarly.Yadon, A.N., et al. (2013). Mol. Cell. Published online March 7, 2013. http://dx.doi.org/10.1016/j.molcel.2013/02.005.

Looping, kissing, and recombining. On the left, a balloon-

based model of the image on the right, which shows monoal-

lelic looping and pairing that faciliate targeted cleavage on

one Tcra allele (red) in recombination centers away from therest of chromosome 14 (green). Image courtesy of J. Skok.

Two Kinds of Loops for Proper T CellReceptor Locus RearrangementsLooping and higher-order structures have implications in genomicstability as well. The recombinases RAG1 and RAG2 generate diversityin immunoglobin (Ig) loci in B cell and in the T cell receptor loci inT cells through a process known as V(D)J recombination. The processrearranges the variable (V), diversity (D), and joining (J) segments withineach locus to create a broad repertoire of receptors to recognizeforeign antigens. The rearrangement of the T cell receptor a locusTcra is one of the last steps in the complex process of diversifyingT cell receptors. There are a large number of possible outcomes foreach recombination event, and tight control is essential. One level ofcontrol is to limit recombination to a single allele. Jane Skok andcolleagues have found that, like other loci, Tcra only cleaves one alleleat a time, and they have added some key insights into how thishappens. First, they show that higher-order intrachromosomal looping

is linked to RAG-mediated cleavage of each allele. The association of RAG proteins coincides with transcription from the soon to becleaved alleles along with decondensation of the chromatin. So what prevents the cleavage of the other allele? Previous work hadshown that a key DNA-damage-signaling kinase, ATM, was involved in repositioning the second allele at Ig loci to heterochromatin.For Tcra, they find that ATM not only controls the positioning to heterochromatin, but also prevents looping. This suggests a negativefeedback loop that acts in trans, wherein ATM is recruited to a break on the first allele and acts to suppress higher-order looping onthe second allele.Chaumell, J., et al. (2013). Cell Rep. 3, 359–370.

Brian PloskySenior Editor, Molecular Cell

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1205

Leading Edge

Voices

Nuclear Biology: What’s Been Most Surprising?

Restricting Genomic Partners

Job DekkerUniversity of Massachusetts Medical School

The regulatory potential of the human

genome is much richer than some had antic-

ipated. With greatly refined annotations,

we now realize that each gene finds itself

surrounded by a huge number of potentially

regulatory elements in a very crowded

nucleus. Given that many regulatory

elements control genes through direct phys-

ical interaction, one can imagine that this

could create a potentially risky situation in

which genes get misregulated by chance

encounters with inappropriate elements.

So, a major question in the field of nuclear

organization is how do cells ensure that

genes only respond to the right regulatory

elements while ignoring the hundreds of

thousands of others?

Recent work has revealed a surprisingly

simple strategy for matching genes to only

some regulatory elements, which involves

the spatial organization and folding of chro-

mosomes inside the nucleus. In Drosophila,

mouse, and human nuclei, chromosomes

are spatially compartmentalized. Using 5C

and Hi-C technologies, it has been shown

that chromosomes form strings of topologi-

cally associating domains (TADs) that are

each hundreds of Kb in size but are spatially

insulated fromneighboring TADs. As a result,

a given gene lives in a relatively small neigh-

borhood where it encounters only a small

section of the genome and thus can partner

with only a small number of regulatory

elements. Future studies will no doubt unveil

how TADs are established and how they

insulate genes from the wrong crowd.

Plasticity of Interpretation

Joanna WysockaStanford University

Advances in genomic profiling technologies

combined with the realization that certain

chromatin features can be effectively used

to annotate cis-regulatory elements enabled

a large number of recent epigenome

mapping efforts across a myriad of cell

types and organisms. The picture that

emerges from these studies elucidates the

astounding degree to which our genome,

including the repetitive regions derived

from transposon elements, appears to be

dynamically utilized for the purposes of

gene regulation. The human ENCODE

project alone mapped nearly 400,000

distinct transcriptional enhancers, most of

which showed high cell type specificity of

the chromatin-marking patterns. Other

studies have demonstrated that thousands

of regulatory regions undergo activation or

decommissioning even during transitions

between closely developmentally related

cell types. It seems highly likely that the infor-

mation content within regulatory parts of

the genome substantially exceeds that of

protein-coding regions, suggesting the

enormous potential for combinatorial com-

plexity of gene expression regulation during

embryogenesis.

Dynamic changes between distinct chro-

matin states have proven to be remarkably

commonplace during differentiation. More-

over, discoveries of enzymatic activities

that are responsible for removal or alteration

of chromatin modifications previously

thought of as relatively stable, such asmeth-

ylation of histone proteins and DNA,

contribute to the mechanistic explanation

of the observed chromatin dynamics.

Taken together, emerging views change our

thinking about both the content of our

genomeand the plasticity of its interpretation

through chromatin-mediated mechanisms.

Cell 152

Surprises at the Membrane

Iain MattajEuropean Molecular Biology Laboratory

Nuclear biology is full of surprises because,

like all biology, the underlying mechanisms

result from evolution and have been selected

to work, independent of how. Studying

nucleocytoplasmic transport, an early eye-

popping moment resulted from calculating

the flux of macromolecules transported

between cytoplasmic and nuclear compart-

ments. Who would guess that transport

through nuclear pore complexes (NPCs)

affects millions of macromolecules every

minute in mammalian cells? Transport

substrates vary, but some are huge, like

viruses or RNPs of up to 50 MDa, whereas

passive macromolecules of >60 KDa are

excluded from NPC transit. It is therefore

remarkable that NPC passage per se occurs

independent of energy input. Instead, import

or export is driven indirectly via energy-

dependent assembly or disassembly of

transport-competent complexes on one

side or the other of the NPC. NPC selectivity

results from the properties of intrinsically

disordered segments of the NPC proteins

that line and occupy the transport channel

and the transport receptors with which they

interact. But the most dynamic aspect of

nuclear biology is the complete disassembly

and reassembly of the nucleus during each

metazoan cell division. Here, the finding is

so old that it is no longer a surprise, but we

have no idea why this should happen, espe-

cially as many single-celled eukaryotes

undergo mitosis with an intact nucleus.

, March 14, 2013 ª2013 Elsevier Inc. 1207

A Blueprint for Spatial Sequencing

Erez Lieberman AidenHarvard University

From an information theoretic standpoint,

the throughput of today’s sequencers

dwarfs almost any other means of interro-

gating a biological sample. This makes it

increasingly tempting to try to ‘‘translate’’

far-flung biological questions into the

language of DNA sequence. But how well

can this sort of experimental shoehorning

work?

If the recent experience of nuclear biology

is any guide, the answer is: better than we

might have guessed. New proximity ligation

methods based on the nuclear ligation

assay and its intellectual descendants have

made DNA sequencers the platform of

choice for rapidly estimating the physical

distance between genomic loci in the

nucleus of a cell. As a result, ‘‘three-dimen-

sional’’ DNA sequencing has begun to have

a marked impact on our understanding of

chromatin structure, playing a role that is

highly complementary to microscopy.

Because ligation-based methods can be

used to probe the distance between other

cellular actors, such as RNAs and proteins,

this development suggests a broader

template for translating cell biology’s spatial

puzzles. And why limit ourselves to the cell’s

interior? Recent proposals have suggested

mapping ‘‘connectomes’’ by tagging indi-

vidual neurons with DNA barcodes and

then ligating the tags. Today’s nuclear

biologymight prove to be tomorrow’s neuro-

science.

1208 Cell 152, March 14, 2013 ª2013 Elsevie

RNA Rewrites Central Dogma

Craig PikaardIndiana University

The pervasive role of RNA in nearly all

aspects of nuclear biology is a continuing

revelation. The eukaryotic nucleus is

commonly perceived to be a realm in which

DNA reigns supreme. Elucidation of the

genetic code showed that messenger

RNAs, transfer RNAs, and ribosomal RNAs

transcribed in the nucleus are exported to

the cytoplasm for protein synthesis. These

early studies suggested that DNA gave the

orders and RNA carried out themission else-

where. Fast forward to today and compelling

evidence that an RNA-based biology preda-

ted the evolution of DNA for information

storage, and one sees the nucleus in a

new light: as a hotbed of RNA-mediated

information management. DNA replication

is initiated by RNA primers. Chromosome

ends aremaintained by RNA-templated telo-

mere addition. Multiple classes of small

regulatory RNAs (e.g., snRNAs, snoRNAs,

and scaRNAs) are critical for messenger

RNA splicing, transfer RNA maturation, ribo-

somal RNA processing, and RNA chemical

modification by methylation or pseudouridy-

lation. More recently, long noncoding RNAs

and short RNAs (siRNAs, miRNAs, and

piRNAs) have been shown to act within the

nucleus to regulate cytosine methylation

(e.g., plants and mammals) or histone modi-

fication (most eukaryotes). These epigenetic

modifications regulate genes during devel-

opment, silence transposons and retrovi-

ruses, and contribute to centromere function

and accurate chromosome segregation. An

emerging RNA-centric view of the nucleus

represents a major paradigm shift.

r Inc.

Leading Edge

Essay

The Cell Biology of Genomes:Bringing the Double Helix to Life

Tom Misteli1,*1National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA

*Correspondence: mistelit@mail.nih.govhttp://dx.doi.org/10.1016/j.cell.2013.02.048

The recent ability to routinely probe genome function at a global scale has revolutionized our view ofgenomes. One of the most important realizations from these approaches is that the functionaloutput of genomes is affected by the nuclear environment in which they exist. Integration ofsequence information with molecular and cellular features of the genome promises a fuller under-standing of genome function.

It was a moment of scientific amaze-

ment in 1952 when Watson and Crick

revealed the structure of DNA. The

magnificence of the double helix and its

elegant simplicity were awe inspiring.

But more than just being beautiful, the

double helix immediately paved the way

forward; its structure implied fundamental

biological processes such as semiconser-

vative replication and the notion that

chemical changes in its composition

may alter heritable traits. The linear struc-

ture of DNA laid the foundation for the

concept that a string of chemical entities

could encode the information that deter-

mines the very essence of every living

organism. The beauty of the double helix

was the promise that, if the sequence of

bases in the genome could be mapped

and decoded, the genetic information

that underlies all living organisms would

be revealed and the secret of biological

systems would be unlocked.

The idea of linearly encoded genetic

information has been spectacularly

successful, culminating in the recent

development of powerful high-throughput

sequencing methods that now allow the

routine reading of entire genomes. The

conceptual elegance of the genome is

that the information contained in the

DNA sequence is absolute. The order of

bases can be determined by sequencing,

and the result is always unequivocal. The

ability to decipher and accurately predict

the behavior of genome sequences was

appealing to the early molecular biolo-

gists, has given rise to the discipline of

molecular genetics, and has catalyzed

the reductionist thinking that has driven

and dominated the field of molecular

biology since its inception.

But the apparent simplicity and deter-

ministic nature of genomes can be decep-

tive. One of the most important lessons

learned from our ability to exhaustively

sequence DNA and to probe genome

behavior at a global scale by mapping

chromatin properties and expression

profiling is that the sequence is only the

first step in genome function. In intact

living cells and organisms, the functional

output of genomes is modulated, and

the hard-wired information contained in

the sequence is often amplified or sup-

pressed. While mutations are an extreme

case of genome modulation, most

commonly occurring changes in genome

function are more subtle and consist of

fluctuations in gene expression, tempo-

rary silencing, or temporary activation of

genes. Although not caused by muta-

tions, these genome activity changes are

functionally important.

Several mechanismsmodulate genome

function (Figure 1). At the transcription

level, the limited availability of com-

ponents of the transcription machinery

at specific sites in the genome influences

the short-term behavior of genes and

may make their expression stochastic.

Epigenetic modifications are capable of

overriding genetically encoded informa-

tion via chemical modification of chro-

matin. Similarly, changes in higher-order

chromatin organization and gene posi-

tioning within the nucleus alter functional

properties of genome regions.

The existence of mechanisms that

modulate the output of genomes makes

Cell 152

it clear that a true understanding of

genome function requires integration of

what we have learned about genome

sequence with what we are still discov-

ering about how genomes are modified

and how they are organized in vivo in the

cell nucleus.

The Stochastic GenomeThe genome is what defines an or-

ganism and an individual cell. It is there-

fore tempting to assume that identical

genomes behave identically in a popula-

tion of cells. We now know that this is

not the case. Individual, genetically iden-

tical cells can behave very differently

even in the same physiological en-

vironment. It is rare to find a truly homo-

geneous population of cells even under

controlled laboratory conditions, as any-

one who has tried to make a cell line

stably expressing a transgene knows.

Much of the variability in biological

behavior between individual cells comes

from stochastic activity of genes (Raj

and van Oudenaarden, 2008).

Genes are by definition low-copy-

number entities, as each typically only

exists in two copies in the cell. Similarly,

many transcription factors are present in

relatively low numbers in the cell nucleus.

The low copy number of genes and

transcription factors makes gene ex-

pression inherently prone to stochastic

effects (Raj and van Oudenaarden,

2008). Numerous observations make it

clear that gene expression is stochastic

in vivo. For example, dose-dependent

increases in gene expression after treat-

ment of cell populations with stimulating

, March 14, 2013 ª2013 Elsevier Inc. 1209

Figure 1. From Primary Sequence to Genome OutputThe hard-wired primary information contained in the genome sequence ismodulated at short or long timescales by severalmolecular and cellular events.Modulation may lead to activation (green) or silencing (red) of genome regions.

ligands, such as hormones,

are often brought about by

high expression of target

genes in a relatively small

number of cells in the

population rather than by

a uniform increase in the

activity in all cells. Stochastic

gene behavior is most evi-

dent in single-cell imaging

approaches, and mapping by

fluorescence in situ hybridiza-

tion of multiple genes, which

according to population-

based PCR analysis are

active in a given cell popula-

tion, shows that only a few

cells transcribe all ‘‘constitu-

tively active’’ genes at any

given time. Most cells only

express a subset of genes,

and the combinations vary

considerably between indi-

vidual cells. These observa-

tions suggest that many

genes blink on and off and

are expressed in bursts rather

than in a continuous fashion

(Larson et al., 2009).

The molecular basis for

stochastic gene expression

is unknown. There are several

candidate mechanisms, all of

which are related to genome

or nuclear organization. Most genes

require some degree of chromatin remod-

eling for activity, which is thought to make

regulatory regions accessible to the tran-

scription machinery. Several observa-

tions suggest that chromatin remodeling

contributes to the stochastic bursting of

gene expression. Maybemost compelling

is the finding that genes located near each

other on the same chromosome show

correlated blinking behavior, indicating

that a local chromosome property, such

as chromatin structure, drives stochastic

behavior (Becskei et al., 2005). Further-

more, altering chromatin, for example

by deletion of chromatin remodeling

machinery, affects stochastic variability

in yeast. It can be envisioned that

the stochastic behavior of genes is

caused by the requirement for cyclical

opening of chromatin regions. Open chro-

matin has a limited persistence time, and

maintaining chromatin in an open state

requires the cyclical action of chromatin

1210 Cell 152, March 14, 2013 ª2013 Elsevie

remodelers. Whether an ‘‘active’’ gene is

transcribed at any given time may thus

depend on the transient condensation

status of its chromatin at a particular

moment.

A second mechanism to impose non-

uniform stochastic genome activity may

be the local availability of the transcription

machinery at a gene. Although transcrip-

tion factors are able to relatively freely

diffuse through the nuclear space, and in

this way effectively scan the genome for

binding sites, their availability and func-

tionality at a given local site may undergo

significant temporal fluctuations (Misteli,

2001). The local availability of transcrip-

tion complexes may affect transcription

frequency positive or negatively. On the

one hand, it is possible that relatively

stable preinitiation complexes persist on

a given gene, where they may support

multiple rounds of transcription and in

this way boost initiation frequency. On

the other hand, assembly of the full poly-

r Inc.

merase is a stochastic and

relatively inefficient event it-

self. In order for a functional

polymerase complex to as-

semble, individual transcrip-

tion machinery components

associate with chromatin in

a step-wise fashion, and

formation of the mature po-

lymerase complex involves

multiple partially assembled

intermediates, many of which

are unstable and disintegrate

before a functionally compe-

tent complex is formed (Mis-

teli, 2001). The inefficiency of

polymerase assembly may

create stochasticity at an indi-

vidual locus.

A further contributor to

stochastic gene expression

may be the organization of

transcription events in tran-

scription factories. These

hubs of transcription consist

of accumulations of transcrip-

tion factors to which multiple

genes, often located on

distinct chromosomes, are

recruited (Edelman and

Fraser, 2012). Typically only

a few hundred such transcrip-

tion factories are observed in

a mammalian cell nucleus. It

is possible that some genes need to phys-

ically relocate from nucleoplasmic loca-

tions to transcription factories. A nomi-

nally ‘‘active’’ gene locus that is not

associated with a transcription factory

may thus be stochastically silent. The

relatively low number of transcription sites

makes them a limiting factor in

the transcription process and thus

a potential mediator of stochastic gene

expression.

Epigenetics—And WhenEpigenetics Is Not EpigeneticsStochastic effects modulate genome

output on short timescales. A mechanism

to modulate the hardwired information of

genomes on longer timescales is via

epigenetics. The Greek-derived ‘‘Epi’’

means ‘‘over’’ or ‘‘above,’’ and epigenetic

effects are defined as heritable changes in

genome activity caused by mechanisms

other than changes in DNA sequence.

Epigenetic events are mediated by

chemical modifications of DNA or core

histones in complex patterns by methyla-

tion, acetylation, ubiquitination, phos-

phorylation, etc. Thesemodifications alter

gene expression by changing the chro-

matin surface and in this way affect the

binding of regulatory factors. Well-estab-

lished examples of such effects include

binding of the DNA-methylation-depen-

dent binding of the MeCP2 protein or the

binding of PHD-domain-containing

proteins to trimethylated histone H3 tails.

Prominent biological effects based on

epigenetic regulation are phenotypic

differences between homozygous twins

or imprinted genes that are expressed

from only one allele in a diploid organism.

A central tenet in the definition of

epigenetic regulation is that its effects

are heritable, i.e., transmittable over

generations. In fact, the concept of epige-

netics was inspired by epidemiological

findings that nutrient availability in pre-

adolescents during the 19th century

Swedish famine determined life expec-

tance of their grandchildren. The epidemi-

ological studies have recently been com-

plemented by controlled laboratory

studies in mice (Rando, 2012), and they

have been extended to themolecular level

by the findings that loss of the histone

H3K4-trimethylation prolongs lifespan

in C. elegans in a heritable fashion for

several generations (Greer et al., 2011).

A complicating aspect of epigenetics is

that the same modifications that mediate

heritable epigenetic regulation may also

bring about nonheritable transient modu-

lations of the genome. In fact, the term

‘‘epigenetic’’ is nowadays often used in

a very cavalier manner to refer to any bio-

logical effect, heritable or not, that is

affected by histone modifications. Even

if they are not heritable, histone modifi-

cations are biologically relevant modula-

tors of genome function. The system of

histone modifications is in many ways

akin to the mechanisms by which signal

transduction pathways work (Schreiber

and Bernstein, 2002). Just as in signal

transduction pathways, posttranslational

modifications on histone tails create

binding sites that are then recognized

by adaptor or reader proteins, which in

turn elicit downstream effects such

as activation of kinases in the case

of signaling cascades or recruitment

of transcription factors in the case of

histone modifications. In further analogy

to the reversible events in signaling

pathways, histone modifications can be

altered or erased by modifying enzymes.

Such transient and reversible modulatory

effects of histone modifications have

been implicated in every step of gene

expression, starting from chromatin re-

modeling to recruitment of transcription

machinery and even to downstream

events that were thought to be chromatin

independent, such as alternative pre-

mRNA splicing (Luco et al., 2011). It is

often difficult to determine heritability of

these histone modification effects, and

it therefore remains unclear how many

of them are truly epigenetic. Regardless,

DNA and histone modifications are

an obvious source of modulation of the

information contained in the genome

sequence.

Genome Organization asa Modulator of Genome FunctionGenomes of course do not exist as linear,

naked DNA in the cell nucleus but are

organized into higher-order chromatin

fibers, chromatin domains, and chro-

mosomes. Many correlations between

genome organization and activity have

been made—most prominently, the find-

ings that transcriptionally active genes

are generally located in decondensed

chromatin and that transcriptionally re-

pressed genome regions are often found

at the nuclear periphery. These observa-

tions point to the possibility that the

spatial organization of the genomemodu-

lates its functional output.

But in considering the relationship of

genome structure with its function, we

are faced with a perpetual chicken-and-

egg problem. Does structure drive func-

tion, or is structure merely a reflection of

function? Much of the thinking on this

topic has been guided by observations

on individual genes. How representative

these were for the genome as a whole

has been a confounding concern. Recent

unbiased genome-wide analysis of struc-

ture/function relationships has validated

the tight link between structure and

function. Large-scale analysis of chro-

matin structure, histone modifications,

and expression profiles shows that

genomes are portioned into well-defined

domains that closely correlate with their

activity status and the presence of active

Cell 152

or repressive histone marks (Sexton

et al., 2012). The domains are separated

by sharp boundaries marked by particular

histone modification patterns and binding

sites for chromatin insulator proteins such

as CTCF. Even stronger evidence comes

from the analysis of physical interactions

between chromatin domains. At least

in fruit flies, functionally equivalent

domains tend to preferentially interact;

that is, domains containing silent regions

cluster in three-dimensional space, as

do domains containing active regions

(Sexton et al., 2012).

But can genome structure drive its func-

tion? The best example for structure-

mediated gene expression effects is the

silencing of genes when they become

juxtaposed to heterochromatin domains,

be it in the nuclear interior or at the nuclear

periphery (Beisel and Paro, 2011). Gene

activity has also frequently been linked to

the position of a gene within the cell

nucleus. The strongest evidence for such

a relationship is experiments in which

genes are transplanted from the nuclear

interior to the lamina, leading to their

repression or making them refractory to

activation (Geyer et al., 2011). Based on

these and similar experiments, it is often

quite categorically stated that active

genome regions are found in the interior

of the nucleus and inactive ones at the

periphery. This is a somewhat misleading

oversimplification. Although lamina-asso-

ciated genome regions are generally gene

poor andare not transcribed, transcription

labeling experiments reveal numerous

active transcription sites at the periphery,

and genes that are near the periphery,

but not physically associated with it, are

often active. On the other hand, inactive

genes are frequently found in the interior.

As far as we can tell, nuclear position per

se does not determine activity, but associ-

ation with repressive regions of the

nucleus, be it at the periphery or the inte-

rior, does.

So, how then should we think about the

chicken-and-egg problem of nuclear

structure and function? How can it be

that clear evidence exists for both ‘‘func-

tion-driving-structure’’ as well for ‘‘struc-

ture-driving-function’’? The likely answer

is that both effects are at play and are

part of an overarching principle in which

the mutual interplay of structure and func-

tion at multiple levels influences gene

, March 14, 2013 ª2013 Elsevier Inc. 1211

expression. The fact that there are very

few known heterochromatic active genes

suggests that a structural change in the

form of chromatin decondensation is

a crucial early step in gene activation.

However, because chromatin states are

generally unstable, mechanisms that rein-

force a decondensed chromatin state

must be in force for a gene to remain

active. Such reinforcing mechanisms are

dependent on gene activity and represent

the ‘‘activity-drives-function’’ aspect of

gene expression. Reinforcement mecha-

nisms might be mediated by what we

consider ‘‘active’’ histone modifications,

some of which are known to be deposited

during transcription as the polymerases

traverse genes. On the flipside, a chro-

matin domain may also impose its effect

on neighboring regions, either in cis on

the same chromosome by spreading or

in trans on distinct chromosomes. This

effect represents the ‘‘structure-drives-

function’’ aspect of genome function.

Such a bidirectional, self-enforcing func-

tion-structure-function model accounts

for most experimental observations on

structure-function relationships in gene

expression.

Facing the ComplexitySince the discovery of the double helix,

we have come to realize that under-

standing genomes requires more than

reading their sequence and that the in-

formation contained in the sequence is

modulated by the cellular environment.

How then do we gain full knowledge of

the functional information encoded in

genomes?

To get a comprehensive picture of

the functional output of genomes, the

sequence information needs to be inte-

grated with other information parameters

such as epigenetic patterns, higher-order

chromatin landscapes, and noncoding

RNA profiles. The technology to do this

is now available, and intense efforts are

currently underway to comprehensively

gather these data sets in various biolog-

ical systems. The first examples of

such multilevel mapping analyses are

emerging, such as the recent flurry of

reports from the ENCODE consortium,

1212 Cell 152, March 14, 2013 ª2013 Elsevie

which has systematically mapped

genome properties ranging from histone

modification profiles to regulatory

elements and chromatin structure (Ecker

et al., 2012). Given the scale and

complexity of the generated data, not to

mention the technical difficulties in gath-

ering it, this is a challenging undertaking

that will require a series of progressively

larger studies. Ideally, future studies

should be designed to systematically

map multiple genome properties for

focused biological systems such as

specific human diseases.

Large-scale mapping of genome-

related parameters and their comparison

is a logical and necessary next step in

the exploration of genomes and their

function. These efforts will create invalu-

able catalogs of genome properties, and

the hope is that, by cross-comparing

data sets, insight into the rules that govern

genome regulation will be gleaned. One

can go one step further and advocate for

an even more comprehensive approach

in which genome expression data are

then compared to other cellular charac-

teristics such as proteomic, metabolomic,

morphological, and physiological data to

systematically link genome activity to bio-

logical behavior. The ultimate version of

such an approach was recently described

in a report by the US National Academies

of Sciences entitled ‘‘Toward Precision

Medicine,’’ which envisioned a fully

minable biomedical data repository that

would include information ranging from

genomic and epigenetic parameters

to physiological features and clinical

symptoms.

The elegant simplicity of the DNA struc-

ture revealed by Watson and Crick is still

stunning. True to its promise when it was

first discovered, it opened up the flood-

gates to understanding heredity. But one

of the most profound lesions from the

ensuing decades of genome exploration

must be that the linear arrangement of

bases in the DNA is not an absolute set

of instructions but is malleable by the

cellular environment. We are just begin-

ning to uncover some of the mechanisms

that are responsible for these effects. As

is the rule in biology, wherein the whole

r Inc.

is often greater than the sum of its parts,

we are realizing that the genome is far

more complex than the sequence of

its DNA.

ACKNOWLEDGMENTS

Due to space limitations, mostly review articles

were cited. Work in the author’s laboratory is sup-

ported by the Intramural Research Program of the

National Institutes of Health (NIH), NCI, Center for

Cancer Research.

REFERENCES

Becskei, A., Kaufmann, B.B., and van Oudenaar-

den, A. (2005). Contributions of low molecule

numberandchromosomal positioning tostochastic

gene expression. Nat. Genet. 37, 937–944.

Beisel, C., and Paro, R. (2011). Silencing chro-

matin: comparing modes and mechanisms. Nat.

Rev. Genet. 12, 123–135.

Ecker, J.R., Bickmore, W.A., Barroso, I., Pritchard,

J.K., Gilad, Y., and Segal, E. (2012). Genomics:

ENCODE explained. Nature 489, 52–55.

Edelman, L.B., and Fraser, P. (2012). Transcription

factories: genetic programming in three dimen-

sions. Curr. Opin. Genet. Dev. 22, 110–114.

Geyer, P.K., Vitalini, M.W., and Wallrath, L.L.

(2011). Nuclear organization: taking a position

on gene expression. Curr. Opin. Cell Biol. 23,

354–359.

Greer, E.L., Maures, T.J., Ucar, D., Hauswirth,

A.G., Mancini, E., Lim, J.P., Benayoun, B.A., Shi,

Y., and Brunet, A. (2011). Transgenerational epige-

netic inheritance of longevity in Caenorhabditis

elegans. Nature 479, 365–371.

Larson, D.R., Singer, R.H., and Zenklusen, D.

(2009). A single molecule view of gene expression.

Trends Cell Biol. 19, 630–637.

Luco, R.F., Allo, M., Schor, I.E., Kornblihtt, A.R.,

and Misteli, T. (2011). Epigenetics in alternative

pre-mRNA splicing. Cell 144, 16–26.

Misteli, T. (2001). Protein dynamics: implications

for nuclear architecture and gene expression.

Science 291, 843–847.

Raj, A., and van Oudenaarden, A. (2008). Nature,

nurture, or chance: stochastic gene expression

and its consequences. Cell 135, 216–226.

Rando, O.J. (2012). Daddy issues: paternal effects

on phenotype. Cell 151, 702–708.

Schreiber, S.L., and Bernstein, B.E. (2002).

Signaling network model of chromatin. Cell 111,

771–778.

Sexton, T., Yaffe, E., Kenigsberg, E., Bantignies,

F., Leblanc, B., Hoichman, M., Parrinello, H.,

Tanay, A., and Cavalli, G. (2012). Three-dimen-

sional folding and functional organization princi-

ples of the Drosophila genome. Cell 148, 458–472.

Leading Edge

Minireview

Spinning the Web of Cell Fate

Kevin Van Bortle1 and Victor G. Corces1,*1Department of Biology, Emory University, Atlanta, GA 30322, USA

*Correspondence: vcorces@emory.edu

http://dx.doi.org/10.1016/j.cell.2013.02.052

Spatiotemporal changes in nuclear lamina composition underlie cell-type-specific chromatin orga-nization and cell fate, suggesting that the lamina forms a dynamic framework critical for genomefunction, cellular identity, and developmental potential.

IntroductionThe incredible complexity and plasticity of eukaryotic genome

organization underlies the transformational ability of stem cells

to become an array of diverse tissues and differentiated cell

types. Interphase chromosomes are spatially arranged into

dynamic structures and subcompartments that significantly

influence gene activity. The nuclear lamina (NL), for example,

preferentially interacts with transcriptionally silent chromatin

characterized by low gene density and the absence of active

histone modifications. Lamina-associated chromatin domains

(LADs) are sharply defined and vary between cell types, sug-

gesting interactions between chromatin and the NL are actively

established and dynamically modified during cellular differen-

tiation and development. Nevertheless, to what degree this

nonchromatin nuclear structure actively participates in gene

regulation and differentiation remains an active area of

research. Recent studies by Clowney et al. (2012), Kohwi

et al. (2013), and Solovei et al. (2013) provide evidence that

spatiotemporal differences in lamina composition and genome

architecture underlie developmental competence and differen-

tiation, suggesting the nuclear lamina is directly involved in

spinning the web of cell fate.

Chromatin at the Nuclear LaminaThe nuclear lamina is a thin proteinacious layer of highly

conserved intermediate filament proteins, called lamins, which

lie at the interface between interphase chromatin and the inner

nuclear membrane. Lamins maintain the mechanical integrity

and shape of the nucleus and serve as a platform for chromatin

organization and gene regulation. Lamins are encoded by three

genes in mammals and categorized as either A type (lamin A/C),

or B type (lamins B1 and B2). Whereas B type lamins are ex-

pressed in essentially all mammalian cell types, A type lamins

appear only in a subset of differentiated cell types and at low

levels in embryonic stem cells (ESCs) (Eckersley-Maslin et al.,

2013).The requirement of A type and/or B type lamins for

appropriate nuclear architecture is often cell type specific, sug-

gesting the function of lamins can be differentially utilized in

a cell-type- and tissue-specific manner. The flexibility in lamin

dependency may also be contingent on the presence of lamin-

associated proteins, such as the Lamin-B Receptor (LBR),

a nuclear envelope protein that can also anchor heterochromatin

to the nuclear periphery (Solovei et al., 2013). Nevertheless,

B type lamins are essential for tissue differentiation and organ

development, and mutations in A type lamins and lamin-associ-

ated proteins cause a wide range of human diseases referred to

as laminopathies.

Nuclear Periphery and Gene RepressionIn most cell types, the nuclear periphery is associated with tran-

scriptionally silent and late replicating chromatin, a feature that

appears to be conserved from yeast to humans. Movement

of genes to the nuclear periphery often coincides with gene

repression, yet artificial tethering experiments suggest that

perinuclear localization is sufficient for downregulation of

some, but not all, genes (Burke and Stewart, 2013). The mecha-

nisms responsible for perinuclear gene silencing and the role of

lamins remain poorly defined. However, mapping of interactions

between chromatin and lamins in vivo using a microarray-based

approach indicates that the NL associates with large, sharply

defined domains characterized by low gene expression levels

(Guelen et al., 2008; Pickersgill et al., 2006). LADs identified in

both Drosophila melanogaster and human fibroblasts contain

widely spaced, coordinately expressed gene clusters, confirm-

ing earlier microscopy-based evidence that the nuclear

periphery preferentially interacts with gene-poor regions. LADs

are also partially enriched for repressive H3K9 and H3K27 meth-

ylation, and recent genetic screens in Caenorhabditis elegans

demonstrate that enzymes involved in H3K9 methylation are

essential for sequestering heterochromatin at the nuclear

periphery (Towbin et al., 2012). Whether the formation of hetero-

chromatin itself is sufficient to drive perinuclear anchoring

is unknown. However, many genes are devoid of repressive

histone modifications in human LADs (Guelen et al., 2008),

suggesting that mammalian chromatin-lamina interactions are

not solely dependent on H3K9 methylation. LAD organization

also requires the transcriptional repressor HDAC3, a histone de-

acetylase targeted to the nuclear periphery by lamin-associated

protein Emerin (Demmerle et al., 2012; Zullo et al., 2012), sug-

gesting the removal and absence of active histone marks are

the defining features of peripheral localization.

Understanding the mechanisms by which LADs are estab-

lished and the molecular link between perinuclear localization

and heterochromatin remains a priority for future research.

Nevertheless, mapping of LADs in both Drosophila and humans

provides preliminary evidence that chromatin insulators, which

correlate with physical domain borders and mediate long-range

interactions, are involved in peripheral compartmentalization.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1213

For example, insulator protein CTCF delineates sharply defined

lamina domain borders in human fibroblasts and mouse embry-

onic stem cells (Guelen et al., 2008; Handoko et al., 2011) and is

essential for perinuclear positioning of the cystic fibrosis-rele-

vant CFTR gene (Muck et al., 2012). Functional analysis of

LAD-derived DNA sequences in murine fibroblasts further

reveals an enrichment for GAGA sequences, which are bound

by a transcriptional repressor, cKrox, in complex with HDAC3

and lamina-interacting protein Lap2b (Zullo et al., 2012). cKrox

mediates chromatin-lamina interactions in a cell-type-specific

manner, suggesting LADs may be developmentally regulated

by differential recruitment of cKrox and other factors. In addition

to insulator proteins, a subset of lamina-associated domain

borders are enriched for H3K4me3 in the absence of CTCF (Zullo

et al., 2012) and delineated by promoters oriented away from

LADs (Guelen et al., 2008), suggesting genome-NL domain

organization is likely specified by a complex combination of

nuclear factors.

Genome-wide mapping studies of chromatin-lamina interac-

tions cannot discriminate between perinuclear associations

and interactions that occur within the nucleoplasm, which are

also involved in cell proliferation and differentiation (Burke and

Stewart, 2013). Interactions between chromatin and nuclear

pore proteins, which are often associated with gene activation,

similarly take place both in and away from the nuclear periphery,

suggesting dynamic movement of perinuclear lamins and pore

proteins is an important process in gene regulation. Targeting

of lamins to the nucleoplasm appears cell type specific and

depends on the expression of different lamin-interacting

proteins, yet the dynamics of chromatin-lamina interactions in

the nucleoplasm remain ill-defined.

Genome-NL Dynamics through Differentiationand DiseaseGene expression patterns underlying cellular identity must be

reprogrammed in order for pluripotent stem cells to give rise

to a complex system of tissues and differentiated cell types,

a feat accomplished collectively by transcription factors, chro-

matin, and DNA modifications, and by 3D rearrangement of

chromatin organization. A series of genome-wide mapping

experiments carried out in mouse ESCs, sequentially derived

neural precursor cells (NPCs), and differentiated astrocytes

(ACs) reveal how genome-NL interactions are reorganized

during lineage commitment and terminal differentiation,

(Peric-Hupkes et al., 2010). LADs are surprisingly congruent

across cell types, with overlap ranging between 73%–87%. In

a follow-up study, cell-type-independent LADs are shown to

be highly conserved between mouse and humans ESCs

and characterized by high A/T content (Meuleman et al.,

2013), suggesting constitutive LADs are specified by interac-

tions between A/T sequence elements and the nuclear lamina.

Cell-type invariant NL-interacting sequences are also A/T rich

in ESCs, further suggesting that conserved LADs represent

an inherited backbone structure for peripheral chromatin

contacts. Nevertheless, localized, cell-type-specific differences

in chromatin interactions indicate that some degree of LAD

reorganization occurs concomitant to differentiation (Peric-

Hupkes et al., 2010). Reorganization of NL interactions from

1214 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

ESC/NPCs/ACs is largely cumulative, i.e., gene relocation

during lineage commitment is maintained during subsequent

cell-type transitions. Genes that undergo repositioning are

substantially different across differentiation lineages and impor-

tant for cellular identity, suggesting genome-NL dynamics

reflect a progressive, lineage-specific process in which factors

important for maintaining pluripotency or involved in cell fate

decisions are regulated by ‘‘locking,’’ or unlocking, genes at

the nuclear periphery.

Kohwi et al. (2013) provide supporting evidence that lamins

indeed contribute to cell fate decisions through gene reposition-

ing and repression at the nuclear periphery. InDrosophila embry-

onic neuroblasts, progenitor competence is lost over time,

wherein sequential expression of temporal identity genes deter-

mines the cell fate of neuronal progeny. The first transcription

factor expressed, Hunchback (Hb), specifies early-born U1/U2

neuronal identity within a limited early competence window.

Tracking of the hunchback (hb) genomic locus in vivo reveals

that the hb gene is gradually and synchronously repositioned

to the nuclear lamina coincidentally with the end of the neuro-

blast early competence window (Kohwi et al., 2013). Depletion

of lamin extends neuroblast competence by reducing both hb

positioning and gene silencing, suggesting peripheral compart-

mentalization and repression of hb is an important determinant

of neuronal fate specification and progenitor competence. To

what extent lamins are required for developmentally regulated

reorganization of other competence-relevant loci will require

future exploration. However, disruption of Drosophila lamin

also prevents peripheral compartmentalization and repression

of testis-specific gene clusters in somatic cells (Shevelyov

et al., 2009), supporting a general model in which the nuclear

lamina imprisons developmental loci for tissue-specific gene

repression.

Independent studies of laminopathies also provide insight into

the function of chromatin interactions at the nuclear lamina. For

example, Hutchinson-Gilford progeria syndrome (HGPS) is

a premature-aging disease caused by progerin, an incompletely

processedmutant form of lamin A that promotes abnormal chro-

matin structure and increased DNA damage. Expression of

a GFP-progerin transgene in human mesenchymal stem cells

(hMSCs) causes aberrant expression of general and tissue-

specific differentiation markers and disrupts the cellular identity,

function, and differentiation potential of hMSCs in a manner

consistent with phenotypes of HGPS patients (Scaffidi and

Misteli, 2008). Examination of cells from HGPS patients further

revealed that SKIP, a downstream coactivator of Notch target

genes normally sequestered and repressed by the nuclear

lamina, loses association with this structure, suggesting aberrant

Notch signaling and differentiation abnormalities result from

disrupted genome-NL interactions. Indeed, recent sequencing-

based mapping of lamin A/C associations reveals that hetero-

chromatin interactions at the NL are reduced genome-wide in

HGPS cells, in accordance with microscopy-based evidence

(McCord et al., 2013). By integrating profiles for lamin A/C and

H3K27me3 with 3D organization changes, McCord et al. (2013)

also demonstrate global changes in chromatin compartmen-

talization in HGPS cells. Changes observed in spatial genome

organization correlate with and are preceded by changes in

Figure 1. Retinal Cells Differentiation and

Chromatin OrganizationSpatiotemporal differences in the nuclear laminacomposition of differentiating retinal cells (left toright) underlie tissue-specific chromatin organi-zation and genome function. Comparison oflamina composition and genome-NL interactionsin the nuclei of embryonic stem cells (ESC),progenitor cells, and differentiated retinal rodcells, bipolar neurons, and ganglion cells. Periph-eral compartmentalization of heterochromatin ismediated by lamin proteins; euchromatin is largelynucleoplasmic. Restructuring of chromatin-laminainteractions is a gradual and cumulative process,wherein changes that occur during lineagecommitment are often maintained in terminallydifferentiated cells (Peric-Hupkes et al., 2010). Inmost cell types, lamin A/C and the inner nuclearmembrane protein LBR are consecutively tran-scribed, with LBR expressed early and A typelamins developmentally regulated for expressionin differentiated cells (see bipolar neurons andganglion cells; Solovei et al., 2013). However,neither LBR nor lamin A/C are transcribed in thedifferentiated rod photoreceptor cells of nocturnalmammals, causing inversion of nuclear architec-ture with implications for night vision (Soloveiet al., 2009).

lamin A/C and heterochromatin, providing additional evidence

that reduction of H3K27me3 and loss of heterochromatin-lamina

interactions underlie changes in chromatin structure and

genome function.

Additional disease-related mutations in lamins and lamin-

associated proteins provide insight into the functional relevance

of dynamic genome-NL interactions for tissue differentiation.

Emery-Dreifuss muscular dystrophy (EDMD) is a slow progress-

ing degenerative muscle disease caused by autosomal-domi-

nant or X-linked mutations in LMNA or in the lamin-interacting

protein Emerin, respectively. Recapitulation of a severe late-

onset EDMD-linked lamin mutation in C. elegans leads to

muscle-specific perinuclear retention and repression of trans-

gene-generated heterochromatin carrying a strong muscle-

specific promoter (Mattout et al., 2011). The dominant single

point mutation in lamin also disrupts tissue-specific expression

patterns and leads to defective muscle organization. Together,

integration of basic and clinical research suggests that

genome-NL interactions are an important regulatory mechanism

for controlling cellular identity, differentiation potential, and

maintenance of tissue integrity.

Inverted Nuclear Architecture: Learning from‘‘Inside Out’’In an extreme twist on the relationship between nuclear organi-

zation and genome function, specific cell types exhibit an

‘‘inside-out’’ architecture in which genes and markers of active

chromatin are found exclusively at the nuclear periphery and

heterochromatin centrally positioned. Nuclear inversion occurs

in the nuclei of mouse retinal rod cells (Solovei et al., 2009),

wherein rearrangement of chromatin takes place during terminal

differentiation of rod nuclei (Figure 1) and affects the optical

properties of the retina by reducing light scattering in the outer

nuclear layer. This unusual pattern of nuclear inversion also

develops in the rod nuclei of several other nocturnal mammals,

suggesting rearrangement of chromatin represents an adapta-

tion for night vision. Nuclear inversion is gradually established

over several weeks, and a recent follow up study (Solovei

et al., 2013) suggests that changes in NL composition underlies

the dynamic arrangement andmaintenance of chromatin organi-

zation in the differentiating rod cells.

In most cells, lamin A/C and the inner nuclear membrane

protein LBR are consecutively transcribed, with LBR expressed

early and A type lamins developmentally regulated for expres-

sion in differentiated cell types. Sequential expression of LBR

and lamin A/C is common across diverse cell types, and differen-

tiated cells that do not express A type lamins often persistently

express LBR. Strikingly, inverted rod nuclei express neither

lamin A/C nor LBR, and transgenic expression of LBR preserves

establishment of the conventional nuclear architecture in differ-

entiated rod cells (Solovei et al., 2013). Moreover, nonrod cells

that do not express lamin A/C undergo inversion in LBR null

mice, and all postmitotic cells undergo inversion in double-null

LBR�/� LMNA�/� mice, indicating that nuclear inversion is

caused by the loss of both LBR and/or lamin A/C. Transgenic

expression of lamin C alone does not prevent inversion in rod

nuclei, suggesting that in contrast to LBR, A type lamins require

additional lamin-associated factors for establishing heterochro-

matin tethers. In myoblasts, deletion of A type lamins reduced

expression of muscle-related genes, whereas deletion of LBR

had a slightly opposite effect, indicating that lamin A/C and

LBR inversely regulate tissue-specific transcription patterns.

Loss of lamin A/C or LBR had comparatively smaller effects on

muscle-specific transcription in differentiated muscle, suggest-

ing lamin dynamics are most critical during the early stages of

myotube differentiation.

Similar evidence for the role of LBR and NL composition in

development and tissue-specific gene expression comes from

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1215

recent studies in mouse olfactory neurons. Murine olfactory

sensory neurons (OSN), which choose and monoallelically

expresses one out of �1,400 olfactory receptor (OR) genes,

are organized into subregions of the olfactory epithelium called

zones. OR genes in mice are highly similar, and previous findings

suggest OR choice might be determined by dynamic reversal of

repressive H3K9 and H4K20 methylation marks along OR

clusters (Magklara et al., 2011). Repressed OR loci colocalize

with H3K9 and H4K20 marks in differentiation-dependent and

OSN-specific nuclear aggregates, which may function to main-

tain silencing and conceal transcription factor binding sites

that might otherwise disrupt transcription of the active OR allele

(Clowney et al., 2012). Silenced OR foci are established near the

center of OSN nuclei and requires the downregulation and

removal of LBR, reminiscent of heterochromatin remodeling in

differentiating rod photoreceptor cells. Similarly, loss of LBR

leads to OR aggregation in non-OSN cells, and OR foci in

OSNs are disrupted by ectopic expression of LBR, which causes

decompaction of OSN heterochromatin and coexpression of

many OR genes. Dynamics in NL composition are therefore

critical for remodeling and effective silencing of nonchosen OR

alleles in olfactory neurons.

Implications for Reprogramming?The dynamics of nuclear lamina composition during differentia-

tion and the importance of lamins in human health have influ-

enced our evolving view of the nuclear periphery, from a simple

framework for nuclear structure to a complex system under-

lying genome function and development. The spatiotemporal

differences in NL composition and nuclear organization also

suggest that genome-NL interactions are likely to be an impor-

tant and understudied feature of ‘‘dedifferentiation.’’ Reprog-

ramming of somatic cells into induced pluripotent stem cells

remains an inefficient process, and cells that successfully

acquire pluripotency do so gradually, through multiple waves

of transcription and changes in chromatin and DNA modifica-

tions. The similarly slow progression in restructuring of nuclear

architecture in differentiating tissues, including repositioning of

the hb locus in differentiating neuroblasts (Kohwi et al., 2013),

remodeling of heterochromatin compartmentalization in rod

photoreceptor cells and olfactory sensory neurons (Clowney

et al., 2012; Solovei et al., 2013), and the gradual loss of lamin

A/C interactions and compartmentalization in HGPS cells

(McCord et al., 2013), suggests that changes in genome-NL

interactions may be the rate-limiting step for both cellular differ-

entiation and reprogramming. The progressive, lineage-specific

nature of remodeling also indicates that peripheral compart-

mentalization is altered in intermediate steps, perhaps relying

on multiple rounds of cell division. The important role of NL

composition and dynamic genome-NL interactions in early

differentiation suggests that the nuclear architecture estab-

lished in somatic cells may also represent a barrier to reprog-

ramming, where factors important for maintaining pluripotency

are ‘‘locked’’ away. It is therefore conceivable that under-

standing the step-wise progression of chromatin-lamina alter-

ations and NL composition differences concomitant to lineage

commitment and terminal differentiation might serve as a guide

for how to find our way back.

1216 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

ACKNOWLEDGMENTS

Work in the authors’ laboratory is supported by NIH award R01GM035463.

The content is solely the responsibility of the authors and does not necessarily

represent the official views of the NIH.

REFERENCES

Burke, B., and Stewart, C.L. (2013). The nuclear lamins: flexibility in function.

Nat. Rev. Mol. Cell Biol. 14, 13–24.

Clowney, E.J., LeGros, M.A., Mosley, C.P., Clowney, F.G., Markenskoff-

Papadimitriou, E.C., Myllys, M., Barnea, G., Larabell, C.A., and Lomvardas,

S. (2012). Nuclear aggregation of olfactory receptor genes governs their mono-

genic expression. Cell 151, 724–737.

Demmerle, J., Koch, A.J., and Holaska, J.M. (2012). The nuclear envelope

protein emerin binds directly to histone deacetylase 3 (HDAC3) and activates

HDAC3 activity. J. Biol. Chem. 287, 22080–22088.

Eckersley-Maslin, M.A., Bergmann, J.H., Lazar, Z., and Spector, D.L. (2013).

Landes highlights. Nucleus 4, 1–2.

Guelen, L., Pagie, L., Brasset, E., Meuleman, W., Faza, M.B., Talhout, W.,

Eussen, B.H., de Klein, A., Wessels, L., de Laat, W., and van Steensel, B.

(2008). Domain organization of human chromosomes revealed by mapping

of nuclear lamina interactions. Nature 453, 948–951.

Handoko, L., Xu, H., Li, G., Ngan, C.Y., Chew, E., Schnapp, M., Lee, C.W., Ye,

C., Ping, J.L., Mulawadi, F., et al. (2011). CTCF-mediated functional chromatin

interactome in pluripotent cells. Nat. Genet. 43, 630–638.

Kohwi, M., Lupton, J.R., Lai, S.L., Miller, M.R., and Doe, C.Q. (2013).

Developmentally regulated subnuclear genome reorganization restricts

neural progenitor competence in Drosophila. Cell 152, 97–108.

Magklara, A., Yen, A., Colquitt, B.M., Clowney, E.J., Allen, W., Markenscoff-

Papadimitriou, E., Evans, Z.A., Kheradpour, P., Mountoufaris, G., Carey, C.,

et al. (2011). An epigenetic signature for monoallelic olfactory receptor ex-

pression. Cell 145, 555–570.

Mattout, A., Pike, B.L., Towbin, B.D., Bank, E.M., Gonzalez-Sandoval, A.,

Stadler, M.B., Meister, P., Gruenbaum, Y., and Gasser, S.M. (2011).

An EDMDmutation in C. elegans lamin blocksmuscle-specific gene relocation

and compromises muscle integrity. Curr. Biol. 21, 1603–1614.

McCord, R.P., Nazario-Toole, A., Zhang, H., Chines, P.S., Zhan, Y., Erdos,

M.R., Collins, F.S., Dekker, J., and Cao, K. (2013). Correlated alterations in

genome organization, histone methylation, and DNA-lamin A/C interactions

in Hutchinson-Gilford progeria syndrome. Genome Res. 23, 260–269.

Meuleman, W., Peric-Hupkes, D., Kind, J., Beaudry, J.B., Pagie, L., Kellis, M.,

Reinders, M., Wessels, L., and van Steensel, B. (2013). Constitutive nuclear

lamina-genome interactions are highly conserved and associated with A/T-

rich sequence. Genome Res. 23, 270–280.

Muck, J.S., Kandasamy, K., Englmann, A., Gunther, M., and Zink, D. (2012).

Perinuclear positioning of the inactive human cystic fibrosis gene depends

on CTCF, A-type lamins and an active histone deacetylase. J. Cell. Biochem.

113, 2607–2621.

Peric-Hupkes, D., Meuleman, W., Pagie, L., Bruggeman, S.W., Solovei, I.,

Brugman, W., Graf, S., Flicek, P., Kerkhoven, R.M., van Lohuizen, M., et al.

(2010). Molecular maps of the reorganization of genome-nuclear lamina

interactions during differentiation. Mol. Cell 38, 603–613.

Pickersgill, H., Kalverda, B., de Wit, E., Talhout, W., Fornerod, M., and van

Steensel, B. (2006). Characterization of the Drosophila melanogaster genome

at the nuclear lamina. Nat. Genet. 38, 1005–1014.

Scaffidi, P., and Misteli, T. (2008). Lamin A-dependent misregulation of adult

stem cells associated with accelerated ageing. Nat. Cell Biol. 10, 452–459.

Shevelyov, Y.Y., Lavrov, S.A., Mikhaylova, L.M., Nurminsky, I.D., Kulathinal,

R.J., Egorova, K.S., Rozovsky, Y.M., and Nurminsky, D.I. (2009). The B-type

lamin is required for somatic repression of testis-specific gene clusters.

Proc. Natl. Acad. Sci. USA 106, 3282–3287.

Solovei, I., Kreysing, M., Lanctot, C., Kosem, S., Peichl, L., Cremer, T., Guck,

J., and Joffe, B. (2009). Nuclear architecture of rod photoreceptor cells adapts

to vision in mammalian evolution. Cell 137, 356–368.

Solovei, I., Wang, A.S., Thanisch, K., Schmidt, C.S., Krebs, S., Zwerger, M.,

Cohen, T.V., Devys, D., Foisner, R., Peichl, L., et al. (2013). LBR and Lamin

A/C Sequentially Tether Peripheral Heterochromatin and Inversely Regulate

Differentiation. Cell 152, 584–598.

Towbin, B.D., Gonzalez-Aguilera, C., Sack, R.,Gaidatzis, D., Kalck, V.,Meister,

P., Askjaer, P., and Gasser, S.M. (2012). Step-wise methylation of histone

H3K9 positions heterochromatin at the nuclear periphery. Cell 150, 934–947.

Zullo, J.M., Demarco, I.A., Pique-Regi, R., Gaffney, D.J., Epstein, C.B.,

Spooner, C.J., Luperchio, T.R., Bernstein, B.E., Pritchard, J.K., Reddy, K.L.,

and Singh, H. (2012). DNA sequence-dependent compartmentalization

and silencing of chromatin at the nuclear lamina. Cell 149, 1474–1487.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1217

Leading Edge

Minireview

Uncovering Nuclear PoreComplexity with Innovation

Rebecca L. Adams1 and Susan R. Wente1,*1Department of Cell and Developmental Biology, Vanderbilt University School of Medicine, Nashville, TN 37232, USA

*Correspondence: susan.wente@vanderbilt.eduhttp://dx.doi.org/10.1016/j.cell.2013.02.042

Advances in imaging and reductionist approaches have provided a high-resolution understandingof nuclear pore complex structure and transport, revealing unexpected mechanistic complexitiesbased on nucleoporin functions and specialized import and export pathways.

First impressions can be misleading. Pioneering transmission

electron microscopy (EM) approaches 60 years ago first re-

vealed a structure within the eukaryotic nuclear envelope (NE):

the nuclear pore complex (NPC) (Gall, 1954) (Figure 1A). The

original view is striking yet deceptively simple, with the �100

MDa proteinaceous NPC assembly spanning the NE to provide

a passageway between the nucleus and cytoplasm. Over time,

insights into NPC structure and function have revealed unex-

pected complexities.

NPC pathways for nucleocytoplasmic transport are based on

the type of cargo. Diffusion through NPCs is inhibited for mole-

cules > �40 kDa; larger macromolecules and/or accumulation

against a concentration gradient requires facilitated transport

(Aitchison and Rout, 2012). Nuclear RNAs are actively exported

for function in the cytoplasm, whereas nuclear import is required

for proteins made in the cytoplasm during interphase. Increased

eukaryotic proteome and RNA repertoires have expanded the

range and bulk of macromolecules that require facilitated trans-

port throughNPCs. Based on the plethora of physiological needs

for proper gene expression, theNPCmust be a robust and selec-

tive portal.

Do all NPCs in a given cell and all transport pathways in

a given NPC function the same? Recent work uncovers unan-

ticipated layers of complexity in NPC structure and function.

High-resolution imaging has allowed dynamic visualization

of NPC transport events, whereas reductionist approaches

pinpoint how both complex and simple components contribute

to transport pathway specialization. How such specialization

might contribute to the transport mechanism and high

cargo load capacity is intriguing. This also sets the stage for

future studies taking into account possible heterogeneity

between NPCs.

Insights Gained from High-Resolution NPC StructuresThe original EM views of the NPC documented a simple structure

with 8-fold rotational symmetry in the plane of the NE. Details of

cytoplasmic filaments and a nuclear basket structure were

defined by scanning EM (Aitchison and Rout, 2012)

(Figure 1C). Leaps in structural resolution come from a combina-

tion of X-ray crystallography studies of NPC proteins (Nups)

(Bilokapic and Schwartz, 2012) and high-resolution cryoelectron

tomography (cryo-ET) of NPCs in intact NEs, with cryo-ET work

1218 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

yielding a 6.6 nm resolution image of the human NPC (Maimon

et al., 2012). Coupling these with strategies to individually

pinpoint different Nups may allow crystal structures of compo-

nents to be modeled into the entire NPC. Tour de force analysis

of most yeast Saccharomyces cerevisiae (S. cerevisiae) Nups

(‘‘NPC-wide’’) by parallel structural and biochemical approaches

enabled in silico computational modeling, generating insights

into NPC molecular architecture (Alber et al., 2007).

Importantly, whereas previous low-resolution studies show

conservation of structure between humans and other eukary-

otes, high-resolution cryo-ET unravels subtle differences in

divergent NPCs. Variations in the cavities near the periphery of

the central transport channel suggest functional divergence in

this part of the NPC (Maimon et al., 2012). These may arise

from the few protein composition differences across species.

Innovations in super-resolution light microscopy should allow

Nup localization to be examined at an EM-level resolution. These

methods have already permitted visualization of the 8-fold

symmetry of Nups in fixed cells (Loschberger et al., 2012)

(Figure 1B) and direct live cell observations of the asymmetric

nuclear-cytoplasmic distribution of Nups in NPCs (Hayakawa

et al., 2012). Further studies employed to map Nups in NPCs

could establish how specific Nup subcomplexes are oriented

in NPCs.

Functional Complexity Revealed by NPC-wide AnalysisMost of the S. cerevisiae and human NPC-constituting proteins

were identified a decade ago. The �30 proteins are grouped

into three functional classes (Terry and Wente, 2009): trans-

membrane Nups that anchor the NPC in the NE, also called

pore membrane proteins (Poms); structural Nups that stabilize

the NE curvature at nuclear pores and provide scaffolding for

assembling other peripheral Nups; and FG Nups that

contribute to the permeability barrier for nonspecific transport

and facilitate movement as direct binding sites for transport

receptors. Nups adopt a limited variety of structural folds

such as b-propeller, a-solenoid, or FG domains (Aitchison

and Rout, 2012; Bilokapic and Schwartz, 2012). Parts of this

simple structural assembly reflect the Nups’ ancestral relation-

ship with vesicle coat complexes. Thus, this complex machine

derives its function through surprisingly simple structural

elements.

Figure 1. NPC Structure and Transport(A) Early EM image of the NPC cytoplasmic face in a salamander oocyte NE.Reprinted with permission from Gall (1954). Scale bar, 100 nm.(B) 8-fold symmetry of the NPC in the NE plane resolved by dSTORMmicroscopy. Lumenal domain of the transmembrane Nup gp210 (magenta)and the FG Nups (green) in a Xenopus oocyte NE. Reprinted with permissionfrom Loschberger et al., 2012. Scale bar, 100 nm.(C) Schematic of NPC architecture. Measurements indicate dimensions forhuman NPC from cryo-ET (Maimon et al., 2012).(D) Transport pathways through the NPC, with distinct FG Nup requirementsfor karyopherin transport versus mRNA export (Terry and Wente, 2009).Protein transport occurs in �10 ms (Yang and Musser, 2006), whereas mRNAexport takes 180 ms (Grunwald and Singer, 2010). Transport cargo sizes toscale with NPC: protein cargo as �80 kDa globular shape, mRNP sizeproportional to the transcript length and shown covered with RNA-bindingproteins, green circles. CBP: 50 cap-binding protein complex.

The complexity in NPC function comes from several elements.

First, different Nups are associated with NPCs for different time

periods. Structural Nups are among the most stable proteins in

a cell, persisting for months or years in a nondividing cell (Savas

et al., 2012); moreover, these remain stably NPC associated

once assembled into the NPC (Rabut et al., 2004). In contrast,

FG Nups are highly dynamic (Rabut et al., 2004), with seconds

to minutes of residence times in the NPC. It is unknown how

this dichotomy in association times for different components

might affect transport. Second, NPC cargo load can alter the

transport mechanism. Single-molecule microscopy studies

show that increasing concentrations of the importin-b transport

receptor alters transport time of both its cargo and molecules

that passively diffuse (Yang and Musser, 2006). It is intriguing

to consider that the environment of a given transport channel

might be temporally impacted due to either cargo load or the

specific associated FG Nups.

Third, diversity in function among the FG Nups is illuminated

by several key NPC-wide studies. FG Nups have been consid-

ered to be interchangeable and of uniform function due to their

common attributes. FG Nups contain motifs enriched in phenyl-

alanine (F) and glycine (G) repeats, such as FXFG and GLFG

(L, leucine; X, any amino acid); the spacer sequences between

FG repeats consist of �5–30 residues that are typically enriched

in polar amino acids. Analyses to date indicate that FG domains

are unstructured and occupy the central NPC channel (Terry and

Wente, 2009; Yamada et al., 2010; Aitchison and Rout, 2012).

Although these FG domains constitute �12% of the NPC

mass, they are not resolved in high-resolution structures. EM

analysis of anti-Nup immunogold-labeled NPCs indicates

that a single FG domain type occupies multiple topologies

(Fahrenkrog et al., 2002). Thus, all FG Nups may share an unex-

pected structural flexibility as a defining feature.

Several notable distinctions are also defined among the FG

domains. NPC-wide analyses of biochemical and biophysical

properties of individual FG domains or subdomains show differ-

ences in cohesive properties in terms of self- and inter-FG inter-

actions and in levels of compaction (collapsed versus random

coil) (Yamada et al., 2010). In vivo evidence reveals distinct

functions for FG domains. In an analysis of FG domain deletion

mutants, S. cerevisiae viability required only specific combina-

tions of FG domains; individual ones were dispensable, with

only a few required in higher-order mutant combinations (Terry

and Wente, 2009). Importantly, FG domain deletion mutants

were defective in specific nuclear transport pathways. For

example, an FG deletion mutant defective in Kap121 import

was competent for mRNA export and vice versa (Terry

and Wente, 2009). Recently, in a Xenopus in vitro system, the

Nup98 was shown to be necessary for generation of the

permeability barrier that inhibits diffusion of macromolecules

(Hulsmann et al., 2012). Without the Nup98 FG domain, only

substitution with another cohesive FG domain restored the

barrier. That the permeability barrier function could be attributed

to one specific FG Nup provides further evidence that all FG

Nups are neither the same nor interchangeable.

A final layer of complexity stems from Nup posttranslational

modifications. It is known that vertebrate FG Nups are modified

by O-linked glycosylation, and this may regulate the vertebrate

NPC permeability barrier (Labokha et al., 2012). Nup98 phos-

phorylation is an initial step in the breakdown of the NPC during

open mitosis (Laurell et al., 2011). Phosphorylation increases

permeability of the NPC either through altering the conformation

of the Nup98 GLFG domain or through inducing its dissociation

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1219

from NPCs (Hulsmann et al., 2012). In an NPC-wide analysis

of ubiquitylation carried out in S. cerevisiae (Hayakawa et al.,

2012), this modification was discovered on almost all Nups.

Interestingly, proper nuclear migration during mitosis requires

Nup159 ubiquitylation. Future work should reveal how these

layers of complexity impact nuclear transport function.

Dynamic and Diverse Transport Pathways Uncoveredwithin NPCsNPC translocation is defined by docking, translocation, and

release steps for cargo complexes (Aitchison and Rout, 2012).

Proteins typically display a nuclear localization sequence (NLS)

for entry or nuclear export sequence (NES) for exit. These motifs

provide binding sites for transport receptors (karyopherins,

importins, exportins, and transportins). RNA transport receptors

either recognize the RNA directly (tRNA and miRNA) or interact

with an RNA-binding adaptor protein (in the mRNA ribonucleo-

protein [mRNP] complex). In addition to cargo interactions,

transport receptors also contain hydrophobic pockets that

bind the phenylalanine residues of FG domains (Terry and

Wente, 2009).

Alternative models for how transport receptor-FG interactions

mediate NPC translocation are under investigation. However,

the understanding of how transport directionality is dictated

has reached better consensus. For karyopherins, accumulation

of cargo against its concentration gradient and recycling of the

transport receptor are based on localized control of Ran GTPase

activity (GTP state in the nucleus and GDP in the cytoplasm).

Specifically, the importin-cargo complex binding to Ran-GTP

in the nucleus causes cargo release. In contrast, a RanGTP-

exportin-cargo complex disassembles in the cytoplasm with

GTP hydrolysis (Aitchison and Rout, 2012). An analogous non-

RanGTP mechanism exists for mRNA export by the NXF1 re-

ceptors (S. cerevisiae Mex67), wherein ATP/ADP cycling of an

RNA-dependent DEAD box ATPase (Dbp, or DDX) localized on

the NPC cytoplasmic filaments drives directional transport

(Folkmann et al., 2011). Overall, directional facilitated transloca-

tion is dictated by spatially controlled, nucleotide-dependent

switches at exit sites.

The requirements of different FG Nups for specific transport

receptors underscore the potential for multiple preferential path-

ways existing in an NPC (Figure 1D) (Terry and Wente, 2009).

Whether the active and passive transport pathways are both

functionally and spatially distinct in the NPC central channel

has been debated. Recent microscopy technologies have

documented real-time single translocation events (Yang and

Musser, 2006; Grunwald and Singer, 2010; Lowe et al., 2010;

Mor et al., 2010; Ma et al., 2012) based on both high spatial

and temporal resolution coupled with single-molecule innova-

tions for specific protein cargo labeling such as large quantum

dots (Lowe et al., 2010). NPC interaction times during facilitated

protein transport were measured as �10 ms, with a reported

range of 2–34 ms (Yang andMusser, 2006), with RanGTP driving

release of large cargo from the NPC (Lowe et al., 2010). These

approaches have also allowed mapping of NPC transport path-

ways, and recent studies suggest that importin-b cargo moves

more peripherally to the central NPC channel, as compared to

diffusive cargo (Figure 1D) (Ma et al., 2012).

1220 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Single mRNAs have also been observed moving across the

NPC by engineering sequence-specific RNA stem loops into

endogenous or inducible transcripts and by coexpressing

fluorescently tagged MS2 RNA stem-loop-binding proteins

(Grunwald and Singer, 2010; Mor et al., 2010). Here, the

observed time frame for mRNA transport through the pore is

180 ms (Grunwald and Singer, 2010) to 500 ms (Mor et al.,

2010), with nuclear and cytosolic rate-limiting steps (Grunwald

and Singer, 2010). The rate-limiting interval at the cytoplasmic

face is likely due to mRNP remodeling to promote directionality.

Although both fast and slow (>800 ms) transport rates are

observed for a single mRNA type (Grunwald and Singer, 2010),

mRNP translocation through the NPC occurred 15-fold faster

than diffusion through the nucleus (Mor et al., 2010).

Comparing the transport of protein and mRNA reveals differ-

ences, with a longer duration for mRNA transport across the

NPC that is possibly due to the size differences in the respective

protein versus mRNP cargos (Figure 1D). mRNA export also has

a rate-limiting step at the NPC entry site that might be attributed

to the mRNA quality control and surveillance mechanisms prior

to export. For protein and mRNA transport single-molecule

experiments, a striking common conclusion is that cargo enters

the NPC and explores the channel in a diffusive/ subdiffusive

manner with observed back and forthmovements. This suggests

the lack of a straight path through the NPC and that movement

itself is not inherently directional. It is remarkable that the

transport events are most often unsuccessful (Grunwald and

Singer, 2010; Yang and Musser, 2006), raising the question

of how the NPC accommodates not only a large amount of

successful transport events, but also an even larger number of

unsuccessful events.

Models Impacted by Nuclear Pore Complexity andHeterogeneityThe NPC’s inherent complexity has favored reductionist

approaches to gain molecular insights into transport mecha-

nisms. Innovations include the development of in vitro nano-

pores and hydrogels for testing the selective barrier properties

with transport receptors and cargo. In a nanopore approach,

recombinant FG domains were coupled to a small nanopore

(30 nm holes) (Jovanovic-Talisman et al., 2009). In contrast, the

hydrogels self-formed under experimentally determined condi-

tions with recombinant FG domains (Labokha et al., 2012). These

strategies demonstrated that FG domains are sufficient for

allowing selective passage of transport receptors. A recent

hydrogel study characterized individual FG domains of Xenopus

laevis on an NPC-wide level, finding that resulting hydrogels

had different capacities for selective transport (Labokha et al.,

2012). To effectively mimic the heterogeneous and dynamic

NPC environment, these systems will require constructing

single nanopores and hydrogels with multiple different FG

domains included. Because of the now known complexity, one

FG domain type cannot be considered in isolation; nor are all

FG domains the same.

Several different models have been proposed for the mecha-

nism of NPC translocation. These differ in how the intermolecular

interactions between FG domains contribute to facilitated trans-

port and a selective barrier (Terry and Wente, 2009; Aitchison

and Rout, 2012; Hulsmann et al., 2012). For example, the

entropic barrier model suggests that unstructured FG domains

function to exclude noninteracting molecules. Alternatively, the

selective phase model proposes that interdomain hydrophobic

interactions form a gel-like meshwork locally ‘‘dissolved’’ by

transport receptor interactions. For bothmodels, work is needed

to account for the heterogeneity of FG domains in vivo and

in vitro. A hybrid model is also quite appealing, wherein functions

for cohesive (for permeability barrier) and noncohesive (for

entropic bristles) interactions are considered (Yamada et al.,

2010). These complexities provide an exciting challenge for

further investigations.

PerspectiveCurrently, a single mechanism of nuclear transport across the

NPC likely does not exist; rather, layers of complexity lead to

multiple specialized pathways in a given NPC. Whether different

transport pathways allow multiple transport events to take

place within a single NPC is still unresolved. Classic EM experi-

ments demonstrated that an individual NPC is capable of

carrying out both import and export (Feldherr et al., 1984);

however, whether import and export can be simultaneous has

not been tested. Tracking single mRNA transcripts reveals tran-

sient association with multiple NPCs before exit (Grunwald and

Singer, 2010) possibly due to the inherent properties of

stochastic cargo movement with the NPC. Alternately, this might

reflect a full cargo load for a given NPC, inhibiting entry and new

translocation events. This may also involve the absence of

specific factors/Nups at a given NPC or quality control mecha-

nisms detecting incomplete processing of the transcript. To

directly address simultaneous transport, a future challenge will

be to monitor single-molecule facilitated transport of different

cargos at the same time within one cell/NPC.

Though specialized transport pathways exist within the

heterogeneous environment of the NPC, it is unclear whether

different NPCs in a cell are specialized for distinct types of

transport. Distinctions might exist in each NPC as a result of

dynamic Nup associations, posttranslational or conformational

changes, or temporal changes in expression. There is evidence

for differential NPC function in specific animal tissues at specific

times in cellular differentiation. A recent study found that a

transmembrane Nup (gp210) was absent in proliferating

myoblasts but was required for differentiation into neuroprogeni-

tors (D’Angelo et al., 2012). Using genome-wide RNA

sequencing, gp210 expression caused differential regulation of

a subset of transcripts without globally affecting NPC transport.

How a transmembrane Nup has these effects is unclear;

however, NPC function is evidently altered by differential Nup

association. Advances in imaging and NPC-wide, or genome-

wide, approaches will be needed to further analyze NPC mech-

anisms of specialization on cellular and organism levels.

Finally, the complexity of Nups extends beyond the NPC, as

independent functions have been uncovered for some Nups

(Raices and D’Angelo, 2012). Thus, a full understanding of

nuclear pore complexity is needed to position the field in evalu-

ating the molecular mechanisms underlying nup mutants linked

to human developmental diseases (Raices and D’Angelo,

2012). The wealth of innovations has unveiled NPC structure

and function as much more complex than anticipated at first

glance.

ACKNOWLEDGMENTS

We thank Joe Gall (Carnegie Institution for Science) and Markus Sauer (Julius-

Maximilians-University Wurzburg) for permission to reprint the images in

Figures 1A and 1B, and we thank Wente laboratory members and Elizabeth

Bowman for discussion. Due to space constraints, we regret not being able

to cite all primary references. The authors were supported by grants from

the National Institutes of Health (R37GM051219 [S.R.W.] and T32HD007502

[R.L.A.]).

REFERENCES

Aitchison, J.D., and Rout, M.P. (2012). Genetics 190, 855–883.

Alber, F., Dokudovskaya, S., Veenhoff, L.M., Zhang, W., Kipper, J., Devos, D.,

Suprapto, A., Karni-Schmidt, O., Williams, R., Chait, B.T., et al. (2007). Nature

450, 695–701.

Bilokapic, S., and Schwartz, T.U. (2012). Curr. Opin. Cell Biol. 24, 86–91.

D’Angelo, M.A., Gomez-Cavazos, J.S., Mei, A., Lackner, D.H., and Hetzer,

M.W. (2012). Dev. Cell 22, 446–458.

Fahrenkrog, B., Maco, B., Fager, A.M., Koser, J., Sauder, U., Ullman, K.S., and

Aebi, U. (2002). J. Struct. Biol. 140, 254–267.

Feldherr, C.M., Kallenbach, E., and Schultz, N. (1984). J. Cell Biol. 99, 2216–

2222.

Folkmann, A.W., Noble, K.N., Cole, C.N., and Wente, S.R. (2011). Nucleus 2,

540–548.

Gall, J.G. (1954). Exp. Cell Res. 7, 197–200.

Grunwald, D., and Singer, R.H. (2010). Nature 467, 604–607.

Hayakawa, A., Babour, A., Sengmanivong, L., and Dargemont, C. (2012).

J. Cell Biol. 196, 19–27.

Hulsmann, B.B., Labokha, A.A., and Gorlich, D. (2012). Cell 150, 738–751.

Jovanovic-Talisman, T., Tetenbaum-Novatt, J., McKenney, A.S., Zilman, A.,

Peters, R., Rout, M.P., and Chait, B.T. (2009). Nature 457, 1023–1027.

Labokha, A.A., Gradmann, S., Frey, S., Hulsmann, B.B., Urlaub, H., Baldus,

M., and Gorlich, D. (2012). EMBO J. 32, 204–218.

Laurell, E., Beck, K., Krupina, K., Theerthagiri, G., Bodenmiller, B., Horvath, P.,

Aebersold, R., Antonin, W., and Kutay, U. (2011). Cell 144, 539–550.

Loschberger, A., van de Linde, S., Dabauvalle, M.C., Rieger, B., Heilemann,

M., Krohne, G., and Sauer, M. (2012). J. Cell Sci. 125, 570–575.

Lowe, A.R., Siegel, J.J., Kalab, P., Siu, M., Weis, K., and Liphardt, J.T. (2010).

Nature 467, 600–603.

Ma, J., Goryaynov, A., Sarma, A., and Yang, W. (2012). Proc. Natl. Acad. Sci.

USA 109, 7326–7331.

Maimon, T., Elad, N., Dahan, I., and Medalia, O. (2012). Structure 20, 998–

1006.

Mor, A., Suliman, S., Ben-Yishay, R., Yunger, S., Brody, Y., and Shav-Tal, Y.

(2010). Nat. Cell Biol. 12, 543–552.

Rabut, G., Doye, V., and Ellenberg, J. (2004). Nat. Cell Biol. 6, 1114–1121.

Raices, M., and D’Angelo, M.A. (2012). Nat. Rev. Mol. Cell Biol. 13, 687–699.

Savas, J.N., Toyama, B.H., Xu, T., Yates, J.R., 3rd, and Hetzer, M.W. (2012).

Science 335, 942.

Terry, L.J., and Wente, S.R. (2009). Eukaryot. Cell 8, 1814–1827.

Yamada, J., Phillips, J.L., Patel, S., Goldfien, G., Calestagne-Morelli, A.,

Huang, H., Reza, R., Acheson, J., Krishnan, V.V., Newsam, S., et al. (2010).

Mol. Cell. Proteomics 9, 2205–2224.

Yang, W., and Musser, S.M. (2006). J. Cell Biol. 174, 951–961.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1221

Leading Edge

Minireview

Enclosing Chromatin: Reassemblyof the Nucleus after Open Mitosis

Cornelia Wandke1 and Ulrike Kutay1,*1Institute of Biochemistry, Department of Biology, ETH Zurich, Schafmattstrasse 18, 8093 Zurich, Switzerland

*Correspondence: ulrike.kutay@bc.biol.ethz.chhttp://dx.doi.org/10.1016/j.cell.2013.02.046

During mitosis in vertebrate cells, the nuclear envelope undergoes extensive structural reorganiza-tion, starting with the retraction of nuclear membranes into the ER at mitotic onset and ending withthe re-enclosure of chromatin by ER-derived membranes during mitotic exit. Here, we review ourcurrent understanding of postmitotic nuclear assembly.

Nuclear and cytoplasmic processes are separated from each

other by the nuclear envelope (NE), a double membrane perfo-

rated by nuclear pore complexes (NPCs). Extensive reorganiza-

tion of the NE accompanies many forms of cell division and is

required for spindle assembly. To establish the mitotic spindle,

microtubules (MTs) need to gain access to chromatin. In

‘‘open’’ mitosis, as employed by most metazoan cells, the

spindle is built in the cytoplasm, and chromatin is exposed to

cytoplasmic MTs by NE breakdown (NEBD). In extreme forms

of open mitosis, e.g., in vertebrate cells, the nuclear compart-

ment is completely taken apart, including the disintegration of

NPCs and the dispersal of NE membranes into the ER.

Consequently, to re-establish nucleocytoplasmic compart-

mentalization, the NE needs to be reassembled around the

segregated mass of chromatin in the future daughter cells. This

relies on the general spatiotemporal orchestration of mitotic

exit and requires coordination of chromosome decondensation,

membrane recruitment to chromatin, and NPC assembly. Here,

we discuss models of mitotic NE/ER remodeling and nuclear

assembly focusing on vertebrate systems and highlight the

crucial role for phosphatases as spatial and temporal regulators

of nuclear reformation.

NE ReformationTemporal control of nuclear reassembly is exerted by the

machinery governing mitotic exit and relies on the inactivation

of mitotic kinases like CDK1, as well as on the action of

protein phosphatases. These phosphatases revert phosphory-

lation events that have driven NEBD during mitotic entry. The

catalytic phosphatase subunits are constitutively active but

restricted in their intracellular localization, substrate speci-

ficity, and overall activity through association with a large

range of regulatory subunits. These regulatory subunits thus

contribute both to temporal and spatial control. Another

aspect of spatial control is conferred by RanGTP generation

around chromatin through the action of the chromatin-bound

RanGEF. Certain nucleoporins and membrane proteins are

kept in a reassembly-incompetent state in the mitotic cytosol

by association with RanGTP-binding import receptors. In the

vicinity of chromatin, RanGTP triggers their release from inhib-

itory importins and helps to spatially confine formation of

1222 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

NPCs and the NE to the surface of chromatin (reviewed in

Guttinger et al., 2009).

Nuclear Membrane Formation—from ER Sheets or

Tubules?

Early experiments using Xenopus egg extracts had suggested

that membranes utilized for postmitotic NE reformation originate

from vesicles. In vitro, these vesicles bind to chromatin,

then flatten and fuse, thereby forming the NE. However, it

became evident later that these vesicles represent a peculiarity

of the in vitro system as they arise by fragmentation of the fragile

ER network during fractionation of Xenopus eggs. It is now

commonly accepted that NE reformation in vivo involves

membranes derived from the mitotic ER, which start to be re-

cruited back to chromatin in late anaphase.

The rapid reattachment of ER membranes to chromatin is

protein mediated and redundantly facilitated by several inner

nuclear membrane (INM) proteins (Anderson et al., 2009). The

best-studied examples are the lamin B receptor (LBR), which

is recruited to chromatin by interaction with core histones

H3/H4 and heterochromatin-binding protein 1 (HP1), as well as

the LEM domain proteins Lap2b, Emerin, and MAN1, which

bind the chromatin-associated barrier-to-autointegration factor

(BAF) in a cell-cycle-regulated manner. A number of membrane

proteins, including LBR and the nucleoporins POM121 and

NDC1, may also directly bind to DNA as it is becoming

more exposed during chromatin decondensation (reviewed in

Guttinger et al., 2009). The redundant involvement of many

membrane proteins that reassociate with chromatin/DNA during

NE formation ensures fast and robust nuclear reassembly.

Although the role of INM proteins in NE reformation is un-

disputed, controversy exists on whether ER membranes

approach chromatin as tubules or sheet-like structures

(Figure 1A). This dispute is more than a scholarly quarrel on

different experimental observations because the mode of NE

reformation may impact on the mechanism of postmitotic NPC

assembly (see below).

Confocal microscopy and electron tomographic analysis of

chemically fixed samples had first indicated that the mitotic ER

is entirely tubular (Puhka et al., 2007). In agreement with this

description, Hetzer and coworkers observed tips of ER tubules

in the vicinity of chromatin during late anaphase, suggesting

Figure 1. Nuclear Envelope Reformation(A) The NE is reformed from ERmembranes, whichcontact chromatin either as tubules or sheets.(B) The ‘‘enclosure’’ and ‘‘insertion’’ models ofpostmitotic NPC assembly. For enclosure, ‘‘pre-NPCs’’ containing Nup107–160 complexesassemble on the chromatin surface, are engulfedby membranes, and mature into NPCs. Insertionrelies on INM-ONM fusion in NE sheets. Theseholes are then occupied by Nups, which assemblestepwise into mature NPCs.(C) Membranes start binding back to chromatin viafuture INM proteins during late anaphase. RanGTPreleases soluble Nups and membrane proteinsfrom inhibitory importins in the vicinity of chro-matin, thereby contributing to spatial control ofNE assembly. Phosphatases confer temporaland spatial control by reverting inhibitory phos-phorylations on chromatin and NE proteins. Theregulatory subunit Repo-Man targets PP1gto chromatin; H3 becomes dephosphorylated,allowing for chromatin restructuring and bindingof HP1. Both HP1 and H3 interact with the INMprotein LBR, connecting chromatin to the NE.PP2A is recruited to membranes by LEM4, whichinhibits the BAF kinase VRK-1 (not shown) andpromotes dephosphorylation of BAF by PP2A.This drives the interaction of BAF with chromatinand other LEM domain proteins.

that ER tubulesmight serve as source of NEmembranes. Enforc-

ing this idea, a preformed, largely tubular ER network could

efficiently support NE assembly in vitro (Anderson and Hetzer,

2007). To form the sheet-like structure of the NE, chromatin-

bound tubules must at some point flatten, expand, and seal

on the chromatin surface, involving chromatin/DNA interactions

of membrane proteins. Accordingly, overexpression of ER

tubule-forming proteins such as reticulons and DP1 delayed

NE formation and nuclear expansion in mammalian cells,

whereas their depletion accelerated the formation of a closed

NE (Anderson and Hetzer, 2008). These experiments were taken

as indication that remodeling of the ER from tubules to sheets

could present a rate-limiting step in nuclear assembly.

Yet, the same results would be expected if a sheet-like

morphology were required for NE reformation in the first place.

Recent studies have indeed questioned the existence of

a primarily tubular mitotic ER network. Using spinning-disk

Cell 152

confocal microscopy and EM tomog-

raphy after high-pressure freezing, Kirch-

hausen and coworkers demonstrated in

various cell types that the mitotic ER is

almost entirely composed of extended

sheets, or ‘‘cisternae,’’ which are contin-

uous with the nascent NE in the chro-

matin periphery (Lu et al., 2009, 2011).

Subsequent re-evaluation of EM fixation

methods and analyses of additional cell

lines by Puhka et al. (2012) revealed

cell-type-specific variations of mitotic

ER structure, softening some of the con-

troversy. Taking these results together,

cisternal organization of the mitotic ER

predominates in the majority of studied cells, with some cell

types displaying a transition to fenestrated ER sheets and

more tubular networks in mitosis.

But how does the ER approach anaphase chromatin—in the

form of sheets or tubules? Lu et al. (2011) indeed observed

sheet-like structures on the surface of chromatin during NE refor-

mation and suggest that ER cisternae attach to chromatin for re-

assembly of the NE. In contrast, chromatin-proximal ER tubules

seemed incompetent in generating NE membranes. Still, the

binding of an ER sheet to anaphase chromatin has not yet

been visualized at sufficient temporal and spatial resolution to

exclude a tubule-to-sheet transition, and additional evidence

will be required to strengthen the ‘‘sheet’’ hypothesis.

NPC Assembly—Insertion or Enclosure?

Concomitantly with the attachment of membranes to chro-

matin, NPCs assemble into the growing NE. The recruitment

of nucleoporins to chromatin is spatially controlled by

, March 14, 2013 ª2013 Elsevier Inc. 1223

RanGTP-dependent release of inhibitory importins in the vicinity

of chromatin and timely dephosphorylation of nucleoporins by

uncharacterized protein phosphatases.

Postmitotic NPC formation is a stepwise process that has

been visualized by live-cell imaging and recapitulated in vitro

using Xenopus egg extracts and sperm chromatin. An initial

event is the binding of the large Nup107–160 NPC scaffolding

complex to chromatin, mediated by the nucleoporin ELYS,

which directly interacts with AT-rich DNA sequences (reviewed

in Guttinger et al., 2009). Immunodepletion of Nup107–160

complexes from in vitro nuclear assembly reactions resulted in

nuclei with pore-free, closed NEs (Harel et al., 2003; Walther

et al., 2003), highlighting the central role of this subcomplex in

coordinating NPC assembly with NE reformation.

Originally, it had been suggested that the Nup107–160

complex is seeded onto chromatin in the form of ‘‘pre-NPCs’’

(Walther et al., 2003). However, careful EM analysis of early

NPC assembly intermediates in the absence of membranes

failed to visualize ring-like ‘‘pre-pores’’ but detected smaller

Nup107-containing seeds (Rotem et al., 2009), likely made up

from single subcomplexes. Ring-shaped NPC structures

were only formed when membranes were present, indicating

a requirement for membrane components to induce subsequent

steps in NPC assembly. These include recruitment of the

membrane nucleoporins POM121 and NDC1, which facilitate

integration of the central Nup53–93 scaffolding complex. At

the same time, Nup98, essential for the transport and barrier

properties of NPCs, associates, followed by other FG domain

Nups rendering the NPC competent for various transport path-

ways (reviewed in Guttinger et al., 2009).

Although an approximate order of nucleoporin recruitment has

been established, the mechanism of postmitotic NPC assembly

is a matter of debate. Two opposing models have been

proposed—‘‘enclosure’’ and ‘‘insertion’’ (Figure 1B). The inser-

tion model suggests that NPCs are introduced into chromatin-

associated NE sheets, necessitating a membrane fusion event

between INM and ONM for pore formation (Macaulay and

Forbes, 1996). Such NPC insertion mechanism is employed

during interphase, when a continuous NE surrounds chromatin;

yet the molecular mechanism of INM-ONM fusion is enigmatic.

For NPC assembly by insertion, it should not matter whether

membranes initially contact chromatin as sheets or tubules, as

long as a flattened double membrane is finally available. The

enclosure model, in contrast, proposes that chromatin-associ-

ated, preassembled NPCs are engulfed by membranes,

rendering membrane fusion between INM and ONM unneces-

sary for postmitotic NPC assembly (Anderson and Hetzer,

2007). Recruitment of large membrane sheets to chromatin

could hinder NPC formation by enclosure on the surface of

affected chromatin areas.

Forbes and coworkers have recently followed postmitotic

nuclear assembly in vitro at low temperatures, which decelerates

the reaction and allows for dissecting assembly steps (Fichtman

et al., 2010). Under these conditions, chromatin-associated

NPC assembly intermediates containing the Nup107–160

complex and POM121 were detected in a fully closed NE.

Completion of NPC assembly indeed occurred upon longer incu-

bation, indicating the requirement of INM-ONM fusion. Thus, in

1224 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

principle, postmitotic assembly can occur by insertion, but it

remains to be shown that this is relevant at normal assembly

kinetics.

Postmitotic and interphase NPC biogenesis differ in several

respects. After mitosis, there is a burst in NPC assembly

(�2,000 NPCs assemble from pre-existing building blocks within

10 min in cultured somatic cells). Compared to this fast, parallel

formation of NPCs, interphase assembly is the culmination of

many occasional events that double NPC number until the next

mitosis. In interphase, newly synthesized nucleoporin (sub-

complexes) must be generated and integrated into a continuous

NE. Notably, kinetic measurements have revealed that single

NPC assembly events are considerably slower in interphase

than after mitosis (Dultz and Ellenberg, 2010). Whether this

reflects mechanistic differences between the pathways, i.e.,

insertion versus enclosure, or simply the fact that de novo

synthesis of nucleoporins is rate limiting for interphase NPC

assembly is unclear.

Additional dissimilarities have been observed between post-

mitotic and interphase NPC biogenesis and have been inter-

preted as evidence for distinct mechanisms underlying both

modes of NPC assembly, although none of these studies has

directly addressed insertion or enclosure. First, ELYS is neces-

sary for NPC assembly at the end of mitosis but appears

dispensable during interphase assembly (Doucet et al., 2010).

Second, there are differences in the order of nucleoporin recruit-

ment, i.e., POM121 precedes the integration of Nup107–160

complexes into the NE during interphase, but not after mitosis

(Doucet et al., 2010; Dultz and Ellenberg, 2010). Yet, both differ-

ences could simply reflect the need for seeding Nup107–160-

containing NPC assembly sites on chromatin after open mitosis.

Third, proteins or domains involved in sensing or generating

membrane curvature are specifically required for interphase

NPC formation, potentially indicating a different pore formation

mechanism. These include the reticulons (Anderson and Hetzer,

2008), the ALPS motif of Nup133 (Doucet et al., 2010), and

a C-terminal membrane-bending domain in Nup53 (Vollmer

et al., 2012). It is, however, conceivable that these membrane

curvature modules are critical for interphase assembly only

because of its slower kinetics and the requirement to stabilize

pore assembly intermediates over longer times. Fourth, RNA

interference (RNAi) experiments indicated that POM121 and

the LINC complex component SUN1 might only be important

for interphase and not for postmitotic NPC assembly, perhaps

via a direct role in NPC insertion (Doucet et al., 2010; Talamas

and Hetzer, 2011). It should be noted, however, that the interpre-

tation of the POM121 RNAi data is controversial, and a postmi-

totic role for POM121 is supported by depletion (Antonin et al.,

2005) and dominant-negative experiments in vitro (Shaulov

et al., 2011).

Taken together, progress in the last years has revealed differ-

ences between postmitotic and interphase NPC assembly. Still,

the call is out whether these indeed reflect distinct mechanisms

of pore generation—i.e., enclosure versus insertion. All so-far-

described dissimilarities could also arise from differences in

chromatin accessibility, NPC assembly kinetics, and availability

of components (reservoir versus de novo synthesis). It will be

possible to directly distinguish between the enclosure and

insertion models for postmitotic assembly once the mechanism

of ONM-INM fusion has been delineated.

Phosphatases in Charge of Spatial and TemporalControlProtein phosphatases directly contribute to NE reformation by

reverting phosphorylation of chromatin and NE components,

thereby making both sides competent for reassociation

(Figure 1C). An important player is protein phosphatase 1g

(PP1g). During early anaphase, PP1g is targeted to chromatin

by its regulatory subunit Repo-Man, causing dephosphorylation

of histone H3 at several sites (Vagnarelli et al., 2011). Likely

by dephosphorylating H3 at Ser10, Repo-Man/PP1g controls

the association of HP1 with chromatin, contributing to hetero-

chromatin formation during mitotic exit (Vagnarelli et al., 2011).

Interestingly, HP1 has been suggested to promote NE reforma-

tion by assisting the recruitment of membranes to chromatin

through interaction with the INM protein LBR (Ye et al., 1997).

Repo-Man also harbors another function in nuclear assembly

that is independent of PP1 activity, which is to support the

recruitment of importin b and Nup153 to the periphery of

anaphase chromosomes. Based on these findings, it has been

speculated that importin b marks sites on chromatin for NPC

assembly (Vagnarelli et al., 2011).

Like PP1, the phosphatase PP2A is required for timely

dephosphorylation of CDK1 substrates and also functions as

a key factor in postmitotic nuclear reassembly. One critical tar-

get of PP2A is BAF. During interphase, BAF provides the chro-

matin-binding site for LEM domain proteins of the INM. At the

onset of mitosis, BAF is phosphorylated by VRK-1, which is

proposed to contribute to NEBD both by releasing BAF from

chromatin and weakening the interaction with LEM domain

proteins (Gorjanacz et al., 2007). Conversely, recruitment of

LEM domain proteins from the ER into the reforming NE requires

dephosphorylation of BAF and its reassociation with chromatin.

Strikingly, one LEM family member, Lem4/ANKLE2 in human

cells and LEM-4L in C. elegans, serves as a membrane-bound

platform that coordinates the dephosphorylation of BAF by

simultaneously inhibiting the BAF kinase VRK-1 and recruiting

PP2A. In this scenario, Lem4 potentially functions as a regulatory

PP2A subunit (Asencio et al., 2012). The interactions between

Lem4, VRK-1, and PP2A must differ between mitotic entry and

exit, suggesting that additional control mechanisms exist

upstream of the VRK1-BAF-PP2A-Lem4 axis. It will be inter-

esting to see whether the integrative role of the NE protein

Lem4 in regulating kinase and phosphatase activity on

a common substrate will emerge as paradigm for spatial coordi-

nation during mitotic exit.

OutlookKey mechanisms underlying the dynamic changes of the cell

nucleus during mitosis have been revealed, yet many exciting

questions remain. Nuclear membranes are retracted into the

mitotic ER, from where they re-emerge in anaphase, but is the

mitotic ER more than an inert spindle shell? Further, it is still

controversial whether NPCs are inserted into or rather enclosed

by the reforming NE. Similarly, it is debated whether the ER

approaches chromatin as tubules or sheets. As the development

of membrane probes for time-resolved superresolution micros-

copy progresses rapidly, these open questions should soon be

resolvable. Finally, protein phosphatases coordinate nuclear

reassembly in a spatiotemporal manner, but only few phospha-

tase-substrate relationships causal for specific steps of nuclear

reassembly have been established. Clearly, much remains to

be learned in this interesting area, especially with respect to

how changes in chromatin translate into competence for nuclear

reformation.

ACKNOWLEDGMENTS

We apologize for the few citations owing to space limitations, and we thank

Drs. M. Mayr and A. Rothballer for critical reading and the SNSF and ERC

for funding.

REFERENCES

Anderson, D.J., and Hetzer, M.W. (2007). Nat. Cell Biol. 9, 1160–1166.

Anderson, D.J., and Hetzer, M.W. (2008). J. Cell Biol. 182, 911–924.

Anderson, D.J., Vargas, J.D., Hsiao, J.P., and Hetzer, M.W. (2009). J. Cell Biol.

186, 183–191.

Antonin,W., Franz, C., Haselmann, U., Antony, C., andMattaj, I.W. (2005). Mol.

Cell 17, 83–92.

Asencio, C., Davidson, I.F., Santarella-Mellwig, R., Ly-Hartig, T.B., Mall, M.,

Wallenfang, M.R., Mattaj, I.W., and Gorjanacz, M. (2012). Cell 150, 122–135.

Doucet, C.M., Talamas, J.A., and Hetzer, M.W. (2010). Cell 141, 1030–1041.

Dultz, E., and Ellenberg, J. (2010). J. Cell Biol. 191, 15–22.

Fichtman, B., Ramos, C., Rasala, B., Harel, A., and Forbes, D.J. (2010). Mol.

Biol. Cell 21, 4197–4211.

Gorjanacz, M., Klerkx, E.P., Galy, V., Santarella, R., Lopez-Iglesias, C.,

Askjaer, P., and Mattaj, I.W. (2007). EMBO J. 26, 132–143.

Guttinger, S., Laurell, E., and Kutay, U. (2009). Nat. Rev. Mol. Cell Biol. 10,

178–191.

Harel, A., Orjalo, A.V., Vincent, T., Lachish-Zalait, A., Vasu, S., Shah, S.,

Zimmerman, E., Elbaum, M., and Forbes, D.J. (2003). Mol. Cell 11, 853–864.

Lu, L., Ladinsky, M.S., and Kirchhausen, T. (2009). Mol. Biol. Cell 20, 3471–

3480.

Lu, L., Ladinsky, M.S., and Kirchhausen, T. (2011). J. Cell Biol. 194, 425–440.

Macaulay, C., and Forbes, D.J. (1996). J. Cell Biol. 132, 5–20.

Puhka, M., Vihinen, H., Joensuu, M., and Jokitalo, E. (2007). J. Cell Biol. 179,

895–909.

Puhka, M., Joensuu, M., Vihinen, H., Belevich, I., and Jokitalo, E. (2012). Mol.

Biol. Cell 23, 2424–2432.

Rotem, A., Gruber, R., Shorer, H., Shaulov, L., Klein, E., and Harel, A. (2009).

Mol. Biol. Cell 20, 4031–4042.

Shaulov, L., Gruber, R., Cohen, I., and Harel, A. (2011). J. Cell Sci. 124, 3822–

3834.

Talamas, J.A., and Hetzer, M.W. (2011). J. Cell Biol. 194, 27–37.

Vagnarelli, P., Ribeiro, S., Sennels, L., Sanchez-Pulido, L., de Lima Alves, F.,

Verheyen, T., Kelly, D.A., Ponting, C.P., Rappsilber, J., and Earnshaw, W.C.

(2011). Dev. Cell 21, 328–342.

Vollmer, B., Schooley, A., Sachdev, R., Eisenhardt, N., Schneider, A.M.,

Sieverding, C., Madlung, J., Gerken, U., Macek, B., and Antonin, W. (2012).

EMBO J. 31, 4072–4084.

Walther, T.C., Alves, A., Pickersgill, H., Loıodice, I., Hetzer, M., Galy, V.,

Hulsmann, B.B., Kocher, T., Wilm,M., Allen, T., et al. (2003). Cell 113, 195–206.

Ye, Q., Callebaut, I., Pezhman, A., Courvalin, J.C., and Worman, H.J. (1997).

J. Biol. Chem. 272, 14983–14989.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1225

Leading Edge

Primer

Criteria for Inference of Chromothripsisin Cancer Genomes

Jan O. Korbel1,* and Peter J. Campbell2,3,4,*1Genome Biology Unit, European Molecular Biology Laboratory, 69117 Heidelberg, Germany2Cancer Genome Project, Wellcome Trust Sanger Institute, Hinxton CB10 1SA, UK3Department of Haematology, Addenbrooke’s Hospital, Cambridge CB2 0QQ, UK4Department of Haematology, University of Cambridge, Cambridge CB22XY, UK

*Correspondence: jan.korbel@embl.de (J.O.K.), pc8@sanger.ac.uk (P.J.C.)

http://dx.doi.org/10.1016/j.cell.2013.02.023

Chromothripsis scars the genome when localized chromosome shattering and repair occurs ina one-off catastrophe. Outcomes of this process are detectable as massive DNA rearrangementsaffecting one or a few chromosomes. Although recent findings suggest a crucial role of chromo-thripsis in cancer development, the reproducible inference of this process remains challenging,requiring that cataclysmic one-off rearrangements be distinguished from localized lesions thatoccur progressively. We describe conceptual criteria for the inference of chromothripsis, basedon ruling out the alternative hypothesis that stepwise rearrangements occurred. Robust meansof inferencemay facilitate in-depth studies on the impact of, and themechanisms underlying, chro-mothripsis.

IntroductionOften described as a disease of the genome, cancer typically

results from the acquisition of DNA alterations in somatic cells

leading to activation of oncogenes and inactivation of tumor

suppressor genes. As a result, cellular processes including

cell-cycle control, apoptosis, and DNA repair are impaired,

conferring a growth advantage to cells and fomenting tumori-

genesis (Stratton et al., 2009). According to a long-standing

presumption, a single genetic hit is typically insufficient for

a cell to develop into cancer. Instead, several progressive (i.e.,

gradually acquired or stepwise) DNA alteration events are

required, resulting in incremental development and progression

of cancer (Knudson, 1971; Stratton et al., 2009).

Recent cancer genome analyses, however, have revisited

this presumption by suggesting an alternative process that

involves massive de novo structural rearrangement formation

in a one-step catastrophic genomic event coined chromothrip-

sis (Stephens et al., 2011) (‘‘chromo’’ from chromosome;

‘‘thripsis’’ for shattering into pieces; illustrated in Figures 1A

and 1B). A key feature of chromothripsis is the formation of

tens to hundreds of locally clustered DNA rearrangements

through a singular, cataclysmic (one-off) event, resulting in

a large number of rearranged fragments (often tens to hundreds)

interspersed with widespread losses of sequence fragments

(Figure 1B). Occasionally, rearrangements resulting from chro-

mothripsis can lead to the formation of small circular DNA

molecules (double-minute chromosomes), which may subse-

quently become amplified if they harbor oncogenes (Rausch

et al., 2012a; Stephens et al., 2011) (Figure 1B). As a result

of the massive DNA alterations occurring, chromosomes

affected by chromothripsis show a characteristic pattern of

copy-number ‘‘oscillations,’’ whereby typically only two (or

occasionally three) copy-number states are detectable along

1226 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

the chromosome in the context of a large number of rearrange-

ments (Stephens et al., 2011).

This pattern distinguishes chromothripsis from other ‘‘punctu-

ated equilibrium’’-like mechanisms in which one-off events

precipitate multiple successive DNA rearrangements. An

example of the latter is the breakage-fusion-bridge cycle

(Figure 1C), in which one DNA double-strand break can result

in further DNA alterations acquired with each subsequent cell

cycle (Bignell et al., 2007; Rudolph et al., 2001). Such processes,

although occurring in a short period of time, are conceptually

different to chromothripsis because they are associated with

DNA replication interspersed with progressive rearrangements,

and thus copy-number states can vary extensively across

the derivative chromosome.

Impact of Chromothripsis on Cancer Developmentand ProgressionThe DNA breakpoints resulting from chromothripsis fre-

quently affect only one or a few chromosomes (Figure 2A). Spec-

tral karyotyping and fluorescent in situ hybridization (FISH)

experiments have further shown that only one of the two parental

chromosomes (or haplotypes) is typically affected by chromo-

thripsis (Stephens et al., 2011). DNA rearrangements arising

through chromothripsis can lead to several simultaneous tumor-

igenic DNA alterations (Rausch et al., 2012a; Stephens et al.,

2011) (illustrated in Figure 1A and Figure 2B). FISH experiments

further showed rearrangement outcomes of chromothripsis to

be detectable throughout practically all cells in a tumor and

not solely in tumor subclones (e.g., Figure 2B), suggesting

that chromothripsis occurs as a relatively early tumorigenic

event (Rausch et al., 2012a; Stephens et al., 2011). Hence,

chromothripsis is thought to contribute to, or even represent

a driving force of, cancer development and progression.

Figure 1. Cataclysmic DNA Rearrangement Processes(A) Tumorigenesis is classically thought to involve the stepwise acquisition of somatic DNA driver alterations (dashed blue arrows). Cellular ‘‘crises,’’ such aschromothripsis, may accelerate this process by resulting in several DNA alterations at once (solid black arrows). The red color symbolizes the acquisition ofmalignant phenotypes in the cell (white = nonmalignant cell; red = aggressive/highly malignant cell).(B)Chromothripsis,acellularcrisisalteringchromosomes inaone-offburst thought to involveasinglecell cycle (adaptedfromStephensetal.,2011,Rauschetal., 2012a).(C) The breakage-fusion-bridge cycle, a prototypic process (McClintock, 1941) involving chromosome end-to-end fusions that lead to clustered breakpoints butnot to extensive copy-number state oscillations. This form of ‘‘crisis’’ typically involves several subsequent cell cycles. Though in the classical breakage-fusionbridge cycle only a single DNA break is thought to occur in each cell-division cycle, it is hypothesized that chromosome end-to-end fusions may also lead tochromothripsis events (Stephens et al., 2011).

The characteristic signature of massive DNA rearrangements

resulting from chromothripsis has been observed in 2%–3% of

cancer samples (Stephens et al., 2011). Distinct malignancies

display different rates of chromothripsis (reviewed in Jones

and Jallepalli, 2012), and the outcomes of such one-off chromo-

somal crises have been reported in diverse cancer entities,

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1227

Figure 2. Appearances of Chromothripsis and Progressive DNA Rearrangements(A) DNA rearrangement pattern of SNU-C1, a tetraploid colorectal cancer cell line, with >200 rearrangements on chromosome 15 associated with widespreadDNA fragment loss (reproduced from Stephens et al., 2011). Oscillating copy-number profiles derived from SNP6microarray data are depicted in the upper panelof points. Allelic ratios for each SNP, depicting segments with retained heterozygosity interspersed with LOH, are shown in the lower panel of dots. HomozygousSNPs cluster at allelic ratios near 0 or 1. Heterozygous SNPs cluster around 0.5. The structural rearrangement graph with intrachromosomal rearrangements of allfour possible orientations is depicted as colored lines that connect DNA segments. The box to the right shows a zoomed-in version of the 15q region. Abundantregions with LOH indicate that chromothripsis preceded genome duplication in this cancer cell line.(B) Chromothripsis in a primary Shh-driven pediatric medulloblastoma sample LFS-MB4 associated with the formation of a circular double-minute chromosomederived from chromosome 2 fragments (reproduced from Rausch et al., 2012a). The outermost rings in the illustrated circular plot depict chromosome coor-dinates and annotated genes with known oncogenes shown in red. FISH analysis verified the colocalization of the synchronously amplifiedMYCN (red) andGLI2(green) oncogenes in the chromothripsis-associated, amplified double-minute chromosomes, and demonstrated their presence throughout virtually all tumorcells (reproduced from Rausch et al., 2012a).

including bone cancer, pediatric medulloblastoma, neuroblas-

toma, colorectal cancer, melanoma, and hematological malig-

nancies (Hirsch et al., 2012; Kloosterman et al., 2011b; Magran-

geas et al., 2011; Molenaar et al., 2012; Northcott et al., 2012;

1228 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Rausch et al., 2012a; Stephens et al., 2011). Furthermore, chro-

mothripsis has been associated with poor patient survival in

several cancers (Hirsch et al., 2012; Magrangeas et al., 2011;

Molenaar et al., 2012; Rausch et al., 2012a), indicating its

potential relevance as a prognostic marker, and suggesting

chromothripsis as a feature of some particularly aggressive

forms of cancer. In sonic hedgehog (Shh)-driven medulloblas-

toma, chromothripsis has been linked with predisposing

(germline) mutations in the gene encoding the p53 tumor sup-

pressor (TP53) (Rausch et al., 2012a), and in group-3-subtype

medulloblastoma and acute myeloid leukemia with somatic

DNA alterations of TP53 (Northcott et al., 2012; Rausch et al.,

2012a). Hence, chromothripsis appears to be prone to occur in

specific contexts—i.e., in conjunction with, or even instigated

by, progressively acquired DNA alterations.

Mechanisms Hypothesized to be Involved inChromothripsisAlthough we can find evidence of these cataclysmic events in

genomes, the mechanisms that give rise to them are still being

worked out. Computational analyses of breakpoint junction

sequences performed at nucleotide resolution have provided

initial clues on the mechanism for rejoining the shattered DNA

fragments. Abundant 2–4 nt long repeating sequences (i.e.,

observed ‘‘microhomology’’) at the respective rearrangement

breakpoints (Stephens et al., 2011) are consistent with the repair

of shattered DNA fragments by nonhomologous end-joining

(NHEJ). Simulation-based computational analyses, described

in more detail below, have further provided compelling evidence

that the complex chromosome aberrations resulting from

chromothripsis result from singular, catastrophic DNA rear-

rangement event (Rausch et al., 2012a; Stephens et al., 2011).

Several hypothetical mechanisms have been proposed to

lead to the massive DNA rearrangements observed in conjunc-

tion with chromothripsis (recently reviewed in Forment et al.,

2012; Jones and Jallepalli, 2012; Maher and Wilson, 2012).

Most proposed mechanisms assume that chromothripsis acts

on condensed chromosomes in association with mitosis, which

may explain the highly localized nature of DNA breakpoints on

a single (or few) chromosomes (Stephens et al., 2011)—although

localized DNA shattering could also occur in the context of the

regular spatial organization of interphase chromosomes (Lichter

et al., 1988; Rausch et al., 2012a). In brief, the following mecha-

nistic hypotheses have been presented and discussed: ionizing

radiation acting upon condensed chromosomes (Stephens et al.,

2011); critical telomere shortening followed by chromosome

end-to-end fusions and subsequent massive DNA breakage

(Stephens et al., 2011); abortive apoptosis events (Tubio and

Estivill, 2011); ‘‘premature chromosome compaction,’’ in which

chromosomes condense before completing DNA replication

and may consequently shatter (Johnson and Rao, 1970; Meyer-

son and Pellman, 2011); and DNA damage associated with the

packaging of mitotically ‘‘delayed’’ chromosomes into separate

cellular compartments known as micronuclei (Crasta et al.,

2012). In this regard, a particularly relevant observation made

by Crasta and coworkers is that of DNA fragmentation affecting

isolated chromosomes packaged into micronuclei, which

addresses the conceptual problem of how highly localized

DNA shattering, in the context of chromothripsis, might be

achieved at the molecular level.

Beyond reports of chromothripsis in many cancers, there is

evidence that a similar (or perhaps identical) process may act

upon germline DNA, resulting in constitutional disorders (Chiang

et al., 2012; Kloosterman et al., 2011a; Liu et al., 2011). Nucleo-

tide resolution analyses of the DNA breakpoint junctions of

‘‘constitutional chromothripsis’’ events revealed the presence

of microhomology compatible with NHEJ in some patients

(Kloosterman et al., 2011a; Kloosterman et al., 2012). In others,

sequence-based evidence for replication-associated structural

rearrangements involving the proposed microhomology-

mediated break-induced replication (MMBIR) mechanism was

reported (Liu et al., 2011), with MMBIR thought to be frequently

associated with duplication events and with the insertion of

short DNA-template-derived sequences (i.e., templated inser-

tions) at the respective breakpoint junctions. The frequent

association of chromothripsis in cancer with sequence loss

(Stephens et al., 2011), rather than with duplication, and the

lack of template-derived insertions at the respective DNA

breakpoints in medulloblastoma (Rausch et al., 2012a) suggests

that they may differ mechanistically from ‘‘constitutional chro-

mothripsis’’ events. As is the case for chromothripsis in cancer,

the molecular mechanism driving ‘‘constitutional chromothrip-

sis’’ has not yet been experimentally elucidated.

Challenges in the Assessment of Chromothripsisin Cancer GenomesAccurate inference of chromothripsis is crucial for further

characterization of the underlying molecular process. However,

the genomic signature left by other processes can resemble

that of chromothripsis potentially resulting in misclassification

of chromothripsis events that may hamper research on the

mechanistic basis of chromothripsis and impede attempts to

exploit chromothripsis as a biomarker for disease prognosis.

To robustly and reproducibly identify DNA rearrangements

arising from chromothripsis, those alterations underlying a

one-off event must be distinguished fromDNA alterations occur-

ring in a stepwise manner.

Different operational definitions have been applied for inferring

chromothripsis in microarray based copy-number profiling

data. These operational definitions have been geared toward

recognizing oscillating copy-number profiles, by requiring the

detection of at least 10, 20, or 50 copy-number alterations (i.e.,

identifiable shifts in the copy-number profile) on a particular

chromosome, with these alterations oscillating between only

two or three copy-number states (Hirsch et al., 2012; Jones

et al., 2012; Magrangeas et al., 2011; Molenaar et al., 2012;

Northcott et al., 2012; Rausch et al., 2012a; Stephens et al.,

2011). In addition to requiring a fixed number of copy-number

alterations (such as 50) as a threshold, the number of DNA

breakpoints associated with oscillating copy-number alterations

has been put in relation to the total number of breakpoints on

a chromosome to define a threshold for inferring chromothripsis

in microarray data (Kim et al., 2013).

Marked differences in the spatial distribution, number, and

types of somatically acquired DNA rearrangements observed

between cancer entities (Yates and Campbell, 2012), however,

limit the utility of a defined threshold in terms of identified

copy-number alterations for ascertaining chromothripsis.

Specifically, cancers displaying pronounced genomic instability,

such as ovarian cancer (Cancer Genome Atlas Research

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1229

Figure 3. Amalgam of DNA Rearrangements in a Cancer Genome from an Ovarian Cancer PatientThe large number and diversity of DNA rearrangements detectable in this cancer genome highlight the necessity to use rigorous statistics for distinguishingchromothripsis events from progressive DNA alterations. Ovarian cancers show widespread DNA copy-number alterations throughout the genome, most ofwhich involve progressive rearrangements (depicted by light blue arrows). Although chromosome 7 may potentially have undergone chromothripsis (purplearrow), the large genome-wide number of alterations limits the utility of operational definitions for inference—hence calling for rigorous statistical testing. Thiscancer genome copy-number alteration profile was determined using microarrays (Cancer Genome Atlas Research Network, 2011). Array data were reanalyzedwith Nexus 6v10 (Biodiscovery) copy-number software, as described in Rausch et al., 2012a. Scales corresponding to array log2 ratios of 1 (gain) and�1 (loss) areindicated beneath the axis corresponding to the X chromosome.

Network, 2011), can harbor such a high number of progressively

acquired somatic DNA alterations per chromosome (Figure 3)

that based on operational definitions, those cancers may

mistakenly be suspected to have undergone chromothripsis.

Additionally, accumulations of DNA alterations on the same

chromosome can be achieved by multistep processes, rather

than one-off events, e.g., through successional breakage-

fusion-bridge cycles (Bignell et al., 2007; Rudolph et al., 2001)

or through consecutive deletions that originate from fragile

sites or are driven by positive selection (Bignell et al., 2010).

Thus, although operational definitions can facilitate the

screening for chromothripsis in microarray copy-number

profiling data, from which copy-number state information but

not the relative order or orientation of rearrangements can be re-

constructed, their utility is noticeably limited—and because

operational definitions are prone to subjectivity, they can inter-

fere with reproducibility.

Criteria for Statistical Assessment of ChromothripsisAmore robust and accurate distinction between DNA rearrange-

ments arising from chromothripsis and those occurring in a

stepwise fashion can be achieved by applying criteria that

enable rigorous statistical evaluation of cancer genome

1230 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

sequencing data (Rausch et al., 2012a; Stephens et al., 2011).

The aim of these criteria is to evaluate the model that a particular

set of DNA rearrangements resulted from stepwise somatic

DNA alterations as compared to the alternative model that the

rearrangements arose through a single catastrophic event (i.e.,

chromothripsis).

The following sections outline the rationale behind several

different criteria, each of which can facilitate the statistical infer-

ence of chromothripsis, allowing for more reproducible and

accurate ascertainment of chromothripsis than otherwise

possible using solely operational definitions. Most of these

criteria take into account the entire set of structural rearrange-

ments that have occurred on a chromosome in question,

including the relative order and orientation of rearranged

segments, which are typically detected using whole-genome

paired-end DNA sequencing data, and which can be repre-

sented in the form of a structural rearrangement graph (Figure 2A

and Box 1).

Clustering of BreakpointsDNA breakpoints occurring in conjunction with chromothripsis

typically show pronounced clustering (depicted in Figure 4A).

Often, 5–10 breaks can be observed within 50 kb, followed by

Box 1. Construction of DNA Structural Rearrangement Graphs

A crucial prerequisite for the inference of chromothripsis is the accu-

rate mapping of somatically acquired DNA structural rearrangements

in samples of interest to obtain a structural rearrangement graph,

which represents the set of somatic rearrangements that occurred

on a chromosome, comprising copy-number state information and

data on the relative order and orientation of segments subsequent to

rearrangement (see e.g., Figure 2A). Accurate structural rearrange-

ment graphs can be obtained using sequence variant discovering

approaches in massively parallel DNA sequencing data. These

approaches include paired-end mapping, which is based on

sequencing the ends of size-selected DNA fragments, and detecting

DNA rearrangements by identifying paired ends that map abnormally

onto the human reference assembly (Campbell et al., 2008; Korbel

et al., 2007; Mills et al., 2011). Deletion-type rearrangements (tail-to-

head) are inferred based on the abnormal distance of mapped ends,

tandem duplication-type (head-to-tail) alterations based on their

abnormal relative mapping order, and inversion-type alterations

(head-to-head or tail-to-tail) based on their abnormal relative mapping

orientation. The sensitivity of paired-end mapping for detecting DNA

alterations is improved when DNA sequencing libraries with different

library insert sizes are used (Mills et al., 2011; Rausch et al., 2012a).

Read-depth analysis (Campbell et al., 2008; Chiang et al., 2009), an

approach based on identifying copy-number alterations by analyzing

the DNA read depth of coverage, can also be used to discover

structural rearrangements and to infer the copy-number status of

segments. Split-read (or clipped-read) analysis, which is based on

evaluating gapped read alignments onto the human reference genome

assembly, enables the fine-mapping of DNA rearrangement break-

points (Rausch et al., 2012b; Wang et al., 2011; Ye et al., 2009). In

theory, DNA sequence assembly can further improve the detection

of structural rearrangement events, although recent analyses suggest

that assembly using short DNA read data displays low sensitivity

compared to the aforementioned sequence variant discovery

approaches (Mills et al., 2011).

Data from several of these rearrangement discovery approaches are

typically combined to describe the somatic DNA structural rearrange-

ment graph. This graph serves as the starting point for the described

criteria for inferring chromothripsis.

long tracts of intact chromosomal sequence. Breakpoints can be

confined to individual chromosome arms with the clustering

presumably resulting from whatever process drives the chromo-

some fragmentation (Stephens et al., 2011). Thus, an analysis of

breakpoint clustering can be used as means to obtain evidence

for chromothripsis (Rausch et al., 2012a; Stephens et al., 2011),

as outlined in Box 2.

Under a progressive rearrangements model, tendencies of

breakpoints to cluster substantially imply a ‘‘memory’’ of

previous rearrangements from one cell division to the next.

Although less pronounced than in chromothripsis, local accumu-

lation of breakpoints can be observed in progressive rearrange-

ment scenarios where it may be driven by either chromosomal

fragility or selection for particular genes within a chromosomal

region (Campbell et al., 2010). As a consequence, under

progressive rearrangement scenarios, breakpoint clustering

tends to be recurrent across patients because both the locations

of cancer genes and fragile sites represent intrinsic features of

the human genome (a priori information that can be taken into

consideration for rigorous statistical evaluation of breakpoint

clustering).

Regularity of Oscillating Copy-Number StatesThe aforementioned oscillating behavior of copy-number states

resulting from chromothripsis (e.g., as evident from the chromo-

thripsis example shown in Figure 2A) can be evaluated rigor-

ously, as illustrated in Figure 4B, by simulating a gradual process

in which each of the structural rearrangements detected on

a chromosome, according to the rearrangement graph, are

introduced onto an in silico (modeled) chromosome one-after-

another (Rausch et al., 2012a; Stephens et al., 2011). By intro-

ducing these rearrangements in a stepwise fashion using Monte

Carlo simulations, we can assess the ability of the progressive

rearrangement null model to reproduce the regular (oscillating)

nature of copy-number state switches characteristic for chromo-

thripsis. Support for chromothripsis is obtained in cases where

the null model is ruled out based on these simulations.

Prevalence of Regions with Interspersed Loss andRetention of HeterozygosityChromothripsis frequently leads to massive loss of segments on

the affected chromosome with segmental losses being inter-

spersed with regions displaying normal (disomic) copy-number

(e.g., copy-number states oscillating between copy-number =

1 and copy-number = 2). Although monosomic regions have

evidently lost heterozygosity, the key feature of chromothripsis

is that the segments in the higher (disomic) copy-number state

have retained heterozygosity (Stephens et al., 2011). The result

is a highly regular (oscillating) pattern of segments with retained

heterozygosity interspersed with loss-of-heterozygosity (LOH)

(see Figure 2A, and illustration in Figure 4C). Once lost to the

cell through deletion, heterozygosity cannot be regained. Hence,

in the presence of an abundance of copy-number states oscil-

lating between the states ‘‘1’’ and ‘‘2,’’ perfect concordance

between disomic regions and heterozygous regions will be

unlikely in the event of gradually acquired rearrangements

(Figure 4C). A simulation in which rearrangements are randomly

and sequentially drawn from the available structural rearrange-

ment graph (Box 1) can be employed to assess this concordance

and hence to evaluate the hypothesis that DNA rearrangements

were gradually acquired.

It is worth noting that if chromothripsis occurs in the context of

polyploidy, the lower copy-number state may not display LOH,

but instead may reflect the resulting allelic contribution in lost

genomic segments (e.g., alternating between allelic ratios of

1:1 and 2:1, if genome duplication precedes chromothripsis).

Nonetheless, in the case of chromothripsis, the resulting allelic

ratios will oscillate between segments that are lost and retained,

and evaluation of concordance of this oscillating behavior with

the segmental copy-number state changes can hence facilitate

the discrimination of one-off from progressive rearrangements

in the context of polyploidy.

Prevalence of Rearrangements Affecting a SingleHaplotypeWhen chromothripsis occurs, fragments resulting from chromo-

somal DNA shattering typically originate from a single parental

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1231

Figure 4. Criteria for the Inference of Chromothripsis(A) Breakpoint clustering can yield evidence for chromothripsis (left), as stepwise alterations (right) do not typically lead to a similar level of clustering of DNAbreaks. Curved colored lines depict individual rearrangements.(B) Oscillating copy-number profiles. The left panel depicts a particular set of rearrangements resulting in oscillating copy number, indicative of chromo-thripsis. The null hypothesis of stepwise alterations can be rejected if simulations making use of all rearrangements depicted in the rearrangement graph fail toresult in oscillations involving so few (in this case two) copy-number states. This is illustrated in the right panel, where the copy-number profile displays fourdifferent states.(C) Interspersed regions with loss and retention of heterozygosity often result from chromothripsis (left) and can be used for statistical testing as in the presence ofstepwise alterations (right) such regularity of patterns is unlikely to occur.(D) Chromothripsis-associated rearrangements are typically detectable on a single parental copy (haplotype) of affected chromosomes (referred to as H1 in theleft), whereas stepwise alterations do not typically show such preference.(E) Because fragments are randomly joined following DNA shattering (left), it follows that the relative order of rearranged fragments and the type of fragment joinsshould be uniformly distributed. By comparison, clustered stepwise alterations often show biases toward certain rearrangement forms (right), and are thus notexpected to result in such uniform joining and ordering of segments.(F) In a region of chromothripsis, each fragment is either retained in or lost from the derivative chromosome, enabling an unambiguous walk through the re-arrangements created. As a result, when viewed on the reference genome, adjacent reads demarcating breakpoints inferred by paired-end mapping showperfect alternations between head (‘‘h’’) and tail (‘‘t’’) paired-end reads (left). In contrast, most progressive DNA alteration scenarios that result in nested re-arrangements (right) do not have this property.

chromosome (or haplotype). Considering that DNA rearrange-

ments can be associatedwith a specific haplotype using phasing

(Box 3), the extent to which rearrangements are biased toward

a single haplotype, rather than occurring on both haplotypes

(assuming disomy), can be used to obtain further evidence for

chromothripsis (Figure 4D). Under the assumption that progres-

sive rearrangements affect eachhaplotype randomly, a statistical

test can provide evidence for chromothripsis by defining the

extent to which rearrangements are concentrated on a single

haplotype—for example, by using the Poisson assumption that

in the presence of progressive rearrangements structural rear-

rangements occur on both haplotypes (null hypothesis), rather

than only on a single one.

1232 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Selection for particular genes within a chromosomal region

can bias progressive rearrangements to occur preferentially on

only one rather than on both haplotypes. However, the rear-

ranged genomic regions would be recurrent across patients if

driven by selection, providing a possible rationale to account

for such potentially confounding factor.

Randomness of DNA Fragment JoinsThe assumption underlying the chromothripsis theory is that the

chromosome fragments are randomly stitched together

(‘‘joined’’), involving a DNA double-strand repair process. The

implication is that at each join, the orientation of the two DNA

fragment ends should be random (illustrated in Figure 4E), in

Box 2. Outline of Statistical Algorithms for Inferring Chromo-thripsis

Guidelines for evaluating the following four criteria are outlined:

1. Clustering of breakpoints: Let fxð1Þ; xð2Þ;.; xðnÞg be the set of break-

point locations on a given chromosome, ordered from the lowest to the

highest (as positioned on the reference genome). The null model of

random breakpoint locations implies that the distances between adja-

cent breakpoints, fxð2Þ � xð1Þ; xð3Þ � xð2Þ;.; xðnÞ � xðn�1Þg, should be

distributed according to an exponential distribution with meanPn�1

1 ðxði + 1Þ � xðiÞÞ=ðn� 1Þ which can be readily evaluated using

a goodness-of-fit test. In our experience, chromothripsis is typically

associated with a strong departure from this null distribution, although

some situations of progressive rearrangements (e.g., rearrangements

arising through successive breakage fusion bridge cycles; Figure 1C)

are too.

2. Randomness of DNA fragment joins: Let frDel; rTD; rH2H; rT2Tg be the

counts of observed rearrangements that have a deletion-type, tandem

duplication-type, head-to-head-inverted, and tail-to-tail-inverted

orientation respectively. If more than one chromosome is involved,

then interchromosomal rearrangements can be interpreted in the

same four categories using orientation of the strands at the breakpoint.

Then, in a region of chromothripsis, we would expect these counts to

be distributed as a multinomial distribution with parameters n=P

riand probability pi = 1=4. A departure from this distribution can be

employed as evidence against the rearrangements arising from a chro-

mothripsis process.

3. Randomness of DNA fragment order: In a chromothripsis event, the

presumption is that the original position of a fragment on the reference

genome carries no information about the origins of the fragments it is

joined to at either end. To test this, let fxð1Þ; xð2Þ;.; xðnÞg be the set of

breakpoint locations, ordered from the lowest to the highest (as posi-

tioned on the reference genome). Each observed rearrangement

consists of two DNA breaks joined together and can be denoted as

fðI1; I2Þg, where I refers to the index of the ordered breakpoints fxðIÞg.Under the chromothripsis model, the paired indices should be random

draws without replacement from f1; 2;.; ng. There are suites of tools

available for statistically assessing randomness that could be adapted

here. One possibility, for example, would be to calculate the mean of

fjI2 � I1jg and compare this to 1,000 Monte Carlo simulations. When

we have tested this in practice, the fragment order from a chromothrip-

sis process is not entirely random, implying some spatial structure to

the DNA repair process but considerably more random than most

progressive rearrangement scenarios.

4. Ability to walk the derivative chromosome: As can be seen in

Figure 4F, the ability to walk the derivative chromosome implies that

adjacent DNA reads demarcating breakpoints inferred by paired-end

mapping (Box 1) must alternate between the head of the paired-end

fragment and the tail of that fragment. Let fxð1Þ; xð2Þ;.; xðnÞg be the

set of breakpoint locations, ordered from the lowest to the highest

(as positioned on the reference genome), and let fsð1Þ; sð2Þ;.; sðnÞgbe the paired-end DNA read (head or tail) associated with each of these

breakpoints. In a region of chromothripsis, if all rearrangements were

observed, fsð1Þ; sð2Þ;.; sðnÞg would be a perfect alternating sequence

of heads and tails when ordered along the reference genome assembly

(Figure 4F). Because some rearrangements are likely to be missed in

the sequencing, the problem is one of whether there are longer runs

of alternating heads and tails than expected by chance. This circum-

stance could be assessed by adapting the Wald-Wolfowitz test for

runs. Note that some progressive rearrangement processes could

give similar runs, such as a series of deletions on a given chromosome,

but that many processes, especially those associated with amplifica-

tion, will not.

Box 3. Application of Haplotype Phasing to Improve StructuralRearrangement Analysis

A normal human genome is diploid (2n), and cancer genomes can

display different karyotype configurations (e.g., tetraploidy, 4n). Ac-

cording to the theory of chromothripsis, structural rearrangements

arising should normally display a bias toward occurring on a single

chromosome homolog (i.e., haplotype), rather than on both haplotypes

for disomic karyotypes (or all four in the case of 4n). Hence, the ability

to relate rearrangements to a specific haplotype would allow inferring

chromothripsis events with increased power. Using short read DNA

sequencing, haplotype phases of 300–400 kb could be used tomonitor

whether adjacent DNA breakpoints arose on a single DNA molecule,

using the 1000 Genomes Project integrated haplotype reference

panel (1000 Genomes Project Consortium et al., 2012) in conjunction

with computational approaches based on imputation (Browning and

Browning, 2011). Chromosome-wide phasing data can be obtained

when germline whole-genomic sequencing data from both parents

or somatic genome sequencing data from aneuploid secondary

tumors (which are common in the context of hereditary disorders

such as Li-Fraumeni syndrome; Li and Fraumeni, 1969) are available

for a patient sample in question.

analogy to a pearl necklace that after being disrupted is put

together, with the pearls added to the chain in random order

and orientation.

To evaluate rearrangement patterns for chromothripsis, the

uniformity of orientation of joined DNA fragments can be inferred

by interpreting the structural rearrangement graph (Box 1)—that

is, the number of tail-to-head (deletion-type), head-to-tail

(tandem-duplication-type), head-to-head and tail-to-tail (inver-

sion-type) rearrangements observed should be broadly equal

(Box 2). This criterion applies whether the rearrangements

are intrachromosomal or interchromosomal. In contrast, for

many other types of clustered rearrangements, this property

does not apply. For example, in regions affected by recurrent

breakage-fusion-bridge cycles, there will be predominance of

head-to-head and tail-to-tail inverted rearrangements, whereas

for chromosomal fragile sites, deletions tend to dominate

among the spectrum of rearrangements (Campbell et al., 2010).

Randomness of DNA Fragment OrderBecause chromosome fragments are randomly joined, their

relative order, namely their position on the derivative chromo-

some, also should be approximately random (Figure 4E) pro-

vided that there is no preference for joining particular ends

together, such as maintaining a centromere or telomere. Hence,

as an extension to the criterion to evaluate the randomness of

DNA fragment orientation, an assessment of the randomness

of DNA fragment order can be used to obtain further evidence

for chromothripsis (Box 2). This criterion applies to both intra-

chromosomal and interchromosomal rearrangements.

‘‘Walking’’ the Derivative ChromosomeIf all DNA rearrangements in a region with chromothripsis are

detectable, it should be possible to reconstruct the relative order

in which segments are joined based on the structural rearrange-

ment graph. Computational approaches for piecing together

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1233

such ‘‘digital karyotypes’’ are being developed (Greenman et al.,

2012). For our purposes here, the chromothripsis model means

that each DNA segment included in the derivative chromosome

resulting from chromothripsis has consistent orientation (namely

it has a head at one end and a tail at the other). The derivative

chromosome then forms a single, coherent chain of segments

with the constraint that either end of each segment must have

consistent configuration. Each DNA segment retained in the

derivative chromosome must be demarcated at either end by

genomic rearrangements—when viewed from the perspective

of the reference genome, each separate segment will start with

a rearrangement from the head of the segment and finish with

a rearrangement from the tail of the segment. This constraint

will lead to an alternating head/tail sequence of DNA rearrange-

ments, detected by paired-end mapping (Box 1), when paired-

ends demarcating breakpoints are represented along the refer-

ence genome (see illustration in Figure 4F).

Importantly, this organization of alternating heads and tails

need not be the case under the alternative model because

when the rearrangements occur sequentially, some segments

can be ‘‘reused’’ in the derivative chromosome. This would

generally break the perfectly alternating head/tail series.

Consider, for example, two tandem duplications, one nested

entirely within the other. There are four breakpoints from two

DNA rearrangements. The two breakpoints at the lowest

genomic reference coordinates are both demarcated by tails

(see Figure 4F, right). This would break the alternating head/tail

sequence and would be inconsistent with chromothripsis. As

described in Box 2, it is relatively straightforward to test for

consistency with this criterion.

Summary and OutlookIn this primer, we describe the characterization of chromothripsis

within a genome as a statistical question geared toward discrim-

inating rearrangements resulting from chromothripsis from

those that result from subsequent stepwise DNA alterations.

Approaches for inferring the presence of chromothripsis in

genomes harboring appreciable levels of gradually acquired

alterations can be viewed as conceptually similar to detecting

driver alterations among the tumult of passenger mutations

and structural abnormalities typically observed in a cancer

genome. In this analogy, the discrimination of driver from

passenger alterations in studies focusing on generating ‘‘cancer

gene catalogs’’ benefits from statistical approaches for rejecting

the hypothesis that an event corresponds to a stochastically

occurring, inconsequential, passenger alteration (Dees et al.,

2012).

Not all of the aforementioned criteria for inferring chromothrip-

sis can be applied to each cancer sample. In cancers harboring

extreme levels of genomic instability, the characteristic stamp of

chromothripsis may be hidden behind the mass of stepwise

alterations in such a way that it may not be confidently detect-

able with the approaches described here. Tumor heterogeneity

and ploidy may affect the inference of chromothripsis. Heteroge-

neity, as a confounding factor, can be partially dealt with by

focusing analyses on those DNA alterations that affected the

same subset of cells based on haplotype-specific analyses of

subclonal alterations (e.g., using approaches described in Nik-

1234 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Zainal et al., 2012). Exome sequencing or microarray based

copy-number profiles cannot be used to infer order and orienta-

tion of rearranged segments, limiting criteria that can be used for

inferring chromothripsis to the evaluation of breakpoint clus-

tering, or to operational definitions (such as the enumeration of

copy-number state changes). Even the most widely used

massively parallel DNA sequencing techniques have remaining

limitations, with short DNA reads (%150 nt) and the most

commonly used paired-end library (Box 1) insert sizes

(<400 bp) remaining ineffective for ascertaining sequence varia-

tion in highly repetitive DNA (Onishi-Seebacher and Korbel,

2011). This technological constraint inevitably limits analyses

to ‘‘mappable’’ genomic regions, which have been estimated

to comprise �90% of the human reference assembly (1000

Genomes Project Consortium et al., 2012). Hence, the available

data may in some cases not be sufficient to infer chromothripsis

reliably, in which case the criteria we describe may be biased

toward presuming that progressive DNA rearrangements

occurred.

Despite these challenges, the criteria described here will

enable researchers to ascertain chromothripsis in cancer

genomes in a rigorous, and more reliable, fashion than feasible

on the basis of operational definitions. We recommend assess-

ment of each of the criteria we described on cancer samples

harboring rearrangements that can be clearly attributed to

chromothripsis as well as on such harboring DNA alterations

that undoubtedly underlie a stepwise process, because this

will facilitate identifying optimal parameters for discriminating

one-off from progressive alterations, which may depend on

sequencing depth and protocol used. With massively parallel

DNA sequencing technology increasingly prevailing over micro-

array-based approaches for cancer genome analysis, we

propose that future studies should verify the occurrence of

chromothripsis by using sequencing data, and by demon-

strating the applicability of different—e.g., at least two—criteria

as ‘‘minimal evidence’’ for discriminating stepwise from one-off

events. We foresee that using robust, reproducible criteria for

classification, future research will reveal electrifying insights

into the functional consequences and mechanistic basis of

chromothripsis.

ACKNOWLEDGMENTS

J.O.K. acknowledges funding from the European Commission (Health-

F2-2010-260791). P.J.C. is a Wellcome Trust Senior Clinical Fellow. We

thank Tobias Rausch, Stephanie Sungalee, Balca Mardin, JoachimWeischen-

feldt, Christopher Buccitelli, and Wolfgang Huber for their thoughtful

comments.

REFERENCES

1000 Genomes Project Consortium, Abecasis, G.R., Auton, A., Brooks, L.D.,

DePristo, M.A., Durbin, R.M., Handsaker, R.E., Kang, H.M., Marth, G.T., and

McVean, G.A. (2012). An integratedmap of genetic variation from 1,092 human

genomes. Nature 491, 56–65.

Bignell, G.R., Santarius, T., Pole, J.C., Butler, A.P., Perry, J., Pleasance, E.,

Greenman, C., Menzies, A., Taylor, S., Edkins, S., et al. (2007). Architectures

of somatic genomic rearrangement in human cancer amplicons at

sequence-level resolution. Genome Res. 17, 1296–1303.

Bignell, G.R., Greenman, C.D., Davies, H., Butler, A.P., Edkins, S., Andrews,

J.M., Buck, G., Chen, L., Beare, D., Latimer, C., et al. (2010). Signatures of

mutation and selection in the cancer genome. Nature 463, 893–898.

Browning, S.R., and Browning, B.L. (2011). Haplotype phasing: existing

methods and new developments. Nat. Rev. Genet. 12, 703–714.

Campbell, P.J., Stephens, P.J., Pleasance, E.D., O’Meara, S., Li, H., Santarius,

T., Stebbings, L.A., Leroy, C., Edkins, S., Hardy, C., et al. (2008). Identification

of somatically acquired rearrangements in cancer using genome-wide

massively parallel paired-end sequencing. Nat. Genet. 40, 722–729.

Campbell, P.J., Yachida, S., Mudie, L.J., Stephens, P.J., Pleasance, E.D.,

Stebbings, L.A., Morsberger, L.A., Latimer, C., McLaren, S., Lin, M.L., et al.

(2010). The patterns and dynamics of genomic instability in metastatic pancre-

atic cancer. Nature 467, 1109–1113.

Cancer Genome Atlas Research Network. (2011). Integrated genomic anal-

yses of ovarian carcinoma. Nature 474, 609–615.

Chiang, D.Y., Getz, G., Jaffe, D.B., O’Kelly, M.J., Zhao, X., Carter, S.L., Russ,

C., Nusbaum, C., Meyerson, M., and Lander, E.S. (2009). High-resolution

mapping of copy-number alterations with massively parallel sequencing.

Nat. Methods 6, 99–103.

Chiang, C., Jacobsen, J.C., Ernst, C., Hanscom, C., Heilbut, A., Blumenthal, I.,

Mills, R.E., Kirby, A., Lindgren, A.M., Rudiger, S.R., et al. (2012). Complex reor-

ganization and predominant non-homologous repair following chromosomal

breakage in karyotypically balanced germline rearrangements and transgenic

integration. Nat. Genet. 44, 390–397, S1.

Crasta, K., Ganem, N.J., Dagher, R., Lantermann, A.B., Ivanova, E.V., Pan, Y.,

Nezi, L., Protopopov, A., Chowdhury, D., and Pellman, D. (2012). DNA

breaks and chromosome pulverization from errors in mitosis. Nature 482,

53–58.

Dees, N.D., Zhang, Q., Kandoth, C., Wendl, M.C., Schierding, W., Koboldt,

D.C., Mooney, T.B., Callaway, M.B., Dooling, D., Mardis, E.R., et al. (2012).

MuSiC: identifying mutational significance in cancer genomes. Genome Res.

22, 1589–1598.

Forment, J.V., Kaidi, A., and Jackson, S.P. (2012). Chromothripsis and cancer:

causes and consequences of chromosome shattering. Nat. Rev. Cancer 12,

663–670.

Greenman, C.D., Pleasance, E.D., Newman, S., Yang, F., Fu, B., Nik-Zainal,

S., Jones, D., Lau, K.W., Carter, N., Edwards, P.A., et al. (2012). Estimation

of rearrangement phylogeny for cancer genomes. Genome Res. 22,

346–361.

Hirsch, D., Kemmerling, R., Davis, S., Camps, J., Meltzer, P.S., Ried, T., and

Gaiser, T. (2012). Chromothripsis and focal copy number alterations determine

poor outcome inmalignant melanoma. Cancer Res., in press. Published online

December 27, 2012. http://dx.doi.org/10.1158/0008-5472.

Johnson, R.T., and Rao, P.N. (1970). Mammalian cell fusion: induction of

premature chromosome condensation in interphase nuclei. Nature 226,

717–722.

Jones, M.J., and Jallepalli, P.V. (2012). Chromothripsis: chromosomes in

crisis. Dev. Cell 23, 908–917.

Jones, D.T., Jager, N., Kool, M., Zichner, T., Hutter, B., Sultan, M., Cho, Y.J.,

Pugh, T.J., Hovestadt, V., Stutz, A.M., et al. (2012). Dissecting the genomic

complexity underlying medulloblastoma. Nature 488, 100–105.

Kim, T.M., Xi, R., Luquette, L.J., Park, R.W., Johnson, M.D., and Park, P.J.

(2013). Functional genomic analysis of chromosomal aberrations in a compen-

dium of 8000 cancer genomes. Genome Res. 23, 217–227.

Kloosterman, W.P., Guryev, V., van Roosmalen, M., Duran, K.J., de Bruijn, E.,

Bakker, S.C., Letteboer, T., van Nesselrooij, B., Hochstenbach, R., Poot, M.,

and Cuppen, E. (2011a). Chromothripsis as a mechanism driving complex

de novo structural rearrangements in the germline. Hum. Mol. Genet. 20,

1916–1924.

Kloosterman, W.P., Hoogstraat, M., Paling, O., Tavakoli-Yaraki, M., Renkens,

I., Vermaat, J.S., van Roosmalen, M.J., van Lieshout, S., Nijman, I.J., Roes-

singh, W., et al. (2011b). Chromothripsis is a common mechanism driving

genomic rearrangements in primary and metastatic colorectal cancer.

Genome Biol. 12, R103.

Kloosterman, W.P., Tavakoli-Yaraki, M., van Roosmalen, M.J., van Binsber-

gen, E., Renkens, I., Duran, K., Ballarati, L., Vergult, S., Giardino, D., Hansson,

K., et al. (2012). Constitutional chromothripsis rearrangements involve clus-

tered double-stranded DNA breaks and nonhomologous repair mechanisms.

Cell Rep. 1, 648–655.

Knudson, A.G., Jr. (1971). Mutation and cancer: statistical study of retinoblas-

toma. Proc. Natl. Acad. Sci. USA 68, 820–823.

Korbel, J.O., Urban, A.E., Affourtit, J.P., Godwin, B., Grubert, F., Simons, J.F.,

Kim, P.M., Palejev, D., Carriero, N.J., Du, L., et al. (2007). Paired-end mapping

reveals extensive structural variation in the human genome. Science 318,

420–426.

Li, F.P., and Fraumeni, J.F., Jr. (1969). Soft-tissue sarcomas, breast cancer,

and other neoplasms. A familial syndrome? Ann. Intern. Med. 71, 747–752.

Lichter, P., Cremer, T., Borden, J., Manuelidis, L., and Ward, D.C. (1988).

Delineation of individual human chromosomes in metaphase and interphase

cells by in situ suppression hybridization using recombinant DNA libraries.

Hum. Genet. 80, 224–234.

Liu, P., Erez, A., Nagamani, S.C., Dhar, S.U., Ko1odziejska, K.E., Dharmadhi-

kari, A.V., Cooper, M.L., Wiszniewska, J., Zhang, F., Withers, M.A., et al.

(2011). Chromosome catastrophes involve replication mechanisms generating

complex genomic rearrangements. Cell 146, 889–903.

Magrangeas, F., Avet-Loiseau, H., Munshi, N.C., and Minvielle, S. (2011).

Chromothripsis identifies a rare and aggressive entity among newly diagnosed

multiple myeloma patients. Blood 118, 675–678.

Maher, C.A., and Wilson, R.K. (2012). Chromothripsis and human disease:

piecing together the shattering process. Cell 148, 29–32.

McClintock, B. (1941). The Stability of Broken Ends of Chromosomes in Zea

Mays. Genetics 26, 234–282.

Meyerson, M., and Pellman, D. (2011). Cancer genomes evolve by pulverizing

single chromosomes. Cell 144, 9–10.

Mills, R.E., Walter, K., Stewart, C., Handsaker, R.E., Chen, K., Alkan, C.,

Abyzov, A., Yoon, S.C., Ye, K., Cheetham, R.K., et al.; 1000 Genomes Project.

(2011). Mapping copy number variation by population-scale genome

sequencing. Nature 470, 59–65.

Molenaar, J.J., Koster, J., Zwijnenburg, D.A., van Sluis, P., Valentijn, L.J., van

der Ploeg, I., Hamdi, M., van Nes, J., Westerman, B.A., van Arkel, J., et al.

(2012). Sequencing of neuroblastoma identifies chromothripsis and defects

in neuritogenesis genes. Nature 483, 589–593.

Nik-Zainal, S., Van Loo, P., Wedge, D.C., Alexandrov, L.B., Greenman, C.D.,

Lau, K.W., Raine, K., Jones, D., Marshall, J., Ramakrishna, M., et al.; Breast

Cancer Working Group of the International Cancer Genome Consortium.

(2012). The life history of 21 breast cancers. Cell 149, 994–1007.

Northcott, P.A., Shih, D.J., Peacock, J., Garzia, L., Morrissy, A.S., Zichner, T.,

Stutz, A.M., Korshunov, A., Reimand, J., Schumacher, S.E., et al. (2012).

Subgroup-specific structural variation across 1,000 medulloblastoma

genomes. Nature 488, 49–56.

Onishi-Seebacher, M., and Korbel, J.O. (2011). Challenges in studying

genomic structural variant formation mechanisms: the short-read dilemma

and beyond. Bioessays 33, 840–850.

Rausch, T., Jones, D.T., Zapatka, M., Stutz, A.M., Zichner, T., Weischenfeldt,

J., Jager, N., Remke, M., Shih, D., Northcott, P.A., et al. (2012a). Genome

sequencing of pediatric medulloblastoma links catastrophic DNA rearrange-

ments with TP53 mutations. Cell 148, 59–71.

Rausch, T., Zichner, T., Schlattl, A., Stutz, A.M., Benes, V., and Korbel, J.O.

(2012b). DELLY: structural variant discovery by integrated paired-end and

split-read analysis. Bioinformatics 28, i333–i339.

Rudolph, K.L., Millard, M., Bosenberg, M.W., and DePinho, R.A. (2001).

Telomere dysfunction and evolution of intestinal carcinoma in mice and hu-

mans. Nat. Genet. 28, 155–159.

Stephens, P.J., Greenman, C.D., Fu, B., Yang, F., Bignell, G.R., Mudie, L.J.,

Pleasance, E.D., Lau, K.W., Beare, D., Stebbings, L.A., et al. (2011). Massive

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1235

genomic rearrangement acquired in a single catastrophic event during cancer

development. Cell 144, 27–40.

Stratton, M.R., Campbell, P.J., and Futreal, P.A. (2009). The cancer genome.

Nature 458, 719–724.

Tubio, J.M., and Estivill, X. (2011). Cancer: When catastrophe strikes a cell.

Nature 470, 476–477.

Wang, J., Mullighan, C.G., Easton, J., Roberts, S., Heatley, S.L., Ma, J., Rusch,

M.C., Chen, K., Harris, C.C., Ding, L., et al. (2011). CREST maps somatic

1236 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

structural variation in cancer genomes with base-pair resolution. Nat. Methods

8, 652–654.

Yates, L.R., and Campbell, P.J. (2012). Evolution of the cancer genome. Nat.

Rev. Genet. 13, 795–806.

Ye, K., Schulz, M.H., Long, Q., Apweiler, R., and Ning, Z. (2009). Pindel:

a pattern growth approach to detect break points of large deletions and

medium sized insertions from paired-end short reads. Bioinformatics 25,

2865–2871.

Leading Edge

Review

Transcriptional Regulation andIts Misregulation in Disease

Tong Ihn Lee1 and Richard A. Young1,2,*1Whitehead Institute for Biomedical Research, Cambridge, MA 02142, USA2Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA*Correspondence: young@wi.mit.edu

http://dx.doi.org/10.1016/j.cell.2013.02.014

The gene expression programs that establish and maintain specific cell states in humans arecontrolled by thousands of transcription factors, cofactors, and chromatin regulators. Misregula-tion of these gene expression programs can cause a broad range of diseases. Here, we reviewrecent advances in our understanding of transcriptional regulation and discuss how these haveprovided new insights into transcriptional misregulation in disease.

IntroductionThe key concepts of transcriptional control were established

half a century ago in bacterial systems (Jacob and Monod,

1961). That pioneering work and many subsequent studies

established that DNA binding transcription factors (also known

as trans-factors) occupy specific DNA sequences at control

elements (cis-elements) and recruit and regulate the transcrip-

tion apparatus. In eukaryotic systems, there has been extensive

study of specific transcription factors and their cofactors, the

general transcription apparatus, and various chromatin regula-

tors, leading to a present-day consensus model for selective

gene control (Adelman and Lis, 2012; Bannister and Kouzar-

ides, 2011; Bonasio et al., 2010; Conaway and Conaway,

2011; Fuda et al., 2009; Ho and Crabtree, 2010; Roeder,

2005; Spitz and Furlong, 2012; Taatjes, 2010; Zhou et al.,

2012b).

Our knowledge of mammalian regulatory elements and the

transcriptional and chromatin regulators that operate at these

sites has increased considerably in the last decade. There

have also been substantial advances in our understanding of

the control of large portions of the gene expression program in

embryonic stem cells (ESCs) and in a number of more differenti-

ated cell types. In these relatively well-studied cells, for example,

it is now understood that a small fraction of the hundreds of tran-

scription factors that are present dominate the control of much of

the active gene expression program (Graf, 2011; Ng and Surani,

2011; Orkin and Hochedlinger, 2011; Young, 2011).

The recent insights into control of cellular gene expression

programs have had an important impact on our understanding

of misregulation of gene expression in disease. Many different

diseases and syndromes, including cancer, autoimmunity,

neurological disorders, diabetes, cardiovascular disease, and

obesity, can be caused by mutations in regulatory sequences

and in the transcription factors, cofactors, chromatin regulators,

and noncoding RNAs that interact with these regions. New

insights into the global effects of some of these mutations have

recently emerged. These insights alter our view of the underlying

cause of some diseases and are the primary focus of this

Review.

We begin with a brief review of the basic features of human

genes and the fundamentals of gene regulation. This leads to

a discussion of cellular gene expression programs and the

mechanisms involved in global regulation of transcription. We

then describe how recent advances in our understanding of

the control of gene expression have led to new insights into

the mechanisms involved in misregulation of gene expression

in various human diseases and disorders.

Genes and Enhancer ElementsThere are a remarkable variety and number of genes that are

transcribed into protein-coding and noncoding RNA (ncRNA)

species in mammalian cells (Table 1). The human genome is

thought to contain �20,000 protein-coding genes and at least

as many ncRNA genes (Djebali et al., 2012). Functions have

been determined or inferred for many of the protein-coding

genes, but less is understood about the functions of the ncRNA

genes.Many of the ncRNAs contribute to control of gene expres-

sion through modulation of transcriptional or posttranscriptional

processes (Bartel, 2009; Ebert and Sharp, 2012; Lee, 2012;

Orom and Shiekhattar, 2011; Rinn and Chang, 2012; Wright

and Ciosk, 2013). For example, the microRNAs (miRNAs), which

are the best studied of the various classes of ncRNAs, fine tune

the levels of target messenger RNAs (mRNAs). Some of the long

ncRNAs (lncRNAs) recruit chromatin regulators to specific

regions of the genome and thereby modify gene expression,

and some apparently do not have a function but are simply

a product of a transcriptional event that is itself regulatory (Latos

et al., 2012).

Transcription factors typically regulate gene expression by

binding enhancer elements and recruiting cofactors and

RNA polymerase II to target genes (Lelli et al., 2012; Ong and

Corces, 2011; Spitz and Furlong, 2012). Multiple transcription

factors typically bind in a cooperative fashion to individual

enhancers (Panne, 2008) and regulate transcription from the

core promoters of nearby or distant genes through physical

contacts that involve looping of the DNA between enhancers

and the core promoters (Krivega and Dean, 2012). The core

promoter elements, which include sites where transcription

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1237

Table 1. Human Genes

Class Abbreviation Function Estimated Number

Messenger RNA mRNA protein coding 20,078a

Ribosomal RNA rRNA structural and functional component of ribosome 531a

Transfer RNA tRNA translational adaptor molecule 512b

Small nuclear RNA snRNA processing of pre-mRNA 1,923a

Small nucleolar RNA snoRNA processing of rRNA, tRNA, and snRNA 1,529a

Antisense RNA aRNA gene regulation 4,424a

Long noncoding RNA lncRNA gene regulation 12,933a

MicroRNA miRNA translational inhibition and mRNA degradation >500c

Small interfering RNA siRNA posttranscriptional gene silencing n/ad

Piwi-interacting RNA piRNA protect genome integrity n/ae

Enhancer RNA eRNA unknown variablef

aGENCODE (Harrow et al., 2012).bHGNC database (Seal et al., 2011; http://www.genenames.org).c2,000–3,000 putative miRNA have been annotated (Harrow et al., 2012), but the majority are not validated.dHuman endogenous siRNA are rare and have not been systematically identified.ePiwi-interacting RNA (piRNA) have not been systematically identified, although estimates indicate that hundreds of piRNAs are derived from each of

more than 100 loci (Aravin et al., 2006; Brennecke et al., 2007; Girard et al., 2006).fEnhancer RNAs (eRNAs) are generated from active enhancers, thus the number of eRNAs depends on the set of active enhancers in a cell (Kim et al.,

2010; Wang et al., 2011). Current estimates indicate that 25%–80% of active enhancers generate eRNAs.

initiation occurs, can also be bound by certain transcription

factors (Dikstein, 2011; Goodrich and Tjian, 2010).

Enhancers can be identified by profiling the locations of key

transcriptional regulators genome wide and by testing whether

these DNA elements are active in enhancer-reporter vectors,

and a large population of ESC enhancers has been identified in

this manner (Chen et al., 2008). Enhancers are occupied by

nucleosomes with specific modifications and are sensitive to

DNase treatment, and these features can be used to identify

putative enhancers when the key transcriptional regulators are

not known (Buecker and Wysocka, 2012; Thurman et al.,

2012). Approximately one million putative enhancers have

recently been identified in the human genome by using, in

multiple cell types, a variety of high-throughput techniques that

detect these features of enhancers (Dunham et al., 2012;

Thurman et al., 2012). These putative enhancers provide a

resource for identifying regions of the genome where sequence

variation may impact factor binding and gene regulation and

thus contribute to disease. Recent studies suggest that a con-

siderable portion of the genetic variation that is associated with

disease occurs in these regulatory regions (Maurano et al., 2012).

Transcriptional Control of GenesTranscriptional regulation occurs at two interconnected levels:

the first involves transcription factors and the transcription appa-

ratus, and the second involves chromatin and its regulators

(Figure 1). We briefly discuss the fundamentals of transcriptional

control in this order, noting recent advances and reviews where

the reader can obtain more detailed information.

Transcription factors can be separated into two classes based

on their regulatory responsibilities: control of initiation versus

control of elongation (Adelman and Lis, 2012; Fuda et al.,

2009; Rahl et al., 2010; Yankulov et al., 1994; Zhou et al.,

2012b). This distinction is not absolute, as some transcription

1238 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

factors may contribute to control of both initiation and elonga-

tion. Transcription factors typically bind cofactors, which are

protein complexes that contribute to activation (coactivators)

and repression (corepressors) but do not have DNA-binding

properties of their own. Most transcription factors are thought

to contribute to transcription initiation and do so by recruiting

coactivators. These coactivators include the Mediator com-

plex, P300, and general transcription factors, among others

(Juven-Gershon and Kadonaga, 2010; Malik and Roeder, 2010;

Sikorski and Buratowski, 2009; Taatjes, 2010). Recent studies

have highlighted the importance of Mediator in integrating infor-

mation from transcriptional activators, repressors, signaling

pathways, and other regulators during transcription initiation

and during the switch to elongation (Berk, 2012; Borggrefe and

Yue, 2011; Conaway and Conaway, 2011; Kagey et al., 2010;

Kornberg, 2005; Lariviere et al., 2012; Malik and Roeder, 2010;

Spaeth et al., 2011; Taatjes, 2010).

Once the recruited RNA polymerase II molecules initiate tran-

scription, they generally transcribe a short distance, typically

20–50 bp, and then pause (Figure 1) (Adelman and Lis, 2012).

This process is controlled by the pause control factors DSIF

and NELF, which are physically associated with the paused

RNA polymerase II molecules. The paused polymerases may

transition to active elongation through pause release, or they

may ultimately terminate transcription with release of the small

RNA species. Pause release and subsequent elongation occur

through recruitment and activation of positive transcription

elongation factor b (P-TEFb), which phosphorylates the paused

polymerase and its associated pause control factors. P-TEFb

can be brought to these sites in the form of a large complex

called the super elongation complex (SEC) (Luo et al., 2012a;

Smith et al., 2011b). Additional complexes, such as PAFc, also

contribute to the regulation of elongation (Jaehning, 2010). Tran-

scription factors such as c-Myc stimulate P-TEFb-mediated

Figure 1. Transcriptional Regulation(A) Formation of a preinitiation complex. Transcription factors bind to specific DNA elements (enhancers) and to coactivators, which bind to RNA polymerase II,which in turn binds to general transcription factors at the transcription start site (arrow). The DNA loop formed between the enhancer and the start site is stabilizedby cofactors such as the Mediator complex and cohesin.(B) Initiation and pausing by RNA polymerase II. RNA polymerase II begins transcription from the initiation site, but pause control factors cause it to stall some tensof base pairs downstream.(C) Pause release and elongation. Various transcription factors and cofactors recruit elongation factors such as P-TEFb, which phosphorylates the pause releasefactors and polymerase, allowing elongation to proceed.(D) Chromatin structure is regulated by ATP-dependent remodeling complexes that can mobilize the nucleosome, allowing regulators and the transcriptionapparatus increased access to DNA sequences.(E) Transcriptional activity is influenced by proteins that modify and bind the histone components of nucleosomes. Some proteins add modifications (writers),some remove modifications (erasers), and others bind via these modifications (readers). The modifications include acetylation (Ac), methylation (Me), phos-phorylation (P), sumoylation (Su), and ubiquitination (Ub).(F) Histone modifications occur in characteristic patterns associated with different transcriptional activities. As an example, the characteristic patterns observedat actively transcribed genes are shown for histone H3 lysine 27 acetylation (H3K27Ac), histone H3 lysine 4 trimethylation (H3K4me3), histone H3 lysine 79dimethylation (H3K79me2), and histone H3 lysine 36 trimethylation (H3K36me3).

release of RNA polymerase II from these pause sites and thus

contribute to the control of transcription elongation (Rahl et al.,

2010).

Recent studies have provided new insights into cofactors that

play important roles in DNA loop formation and maintenance,

which are key to proper gene control. During transcription initia-

tion, the DNA loop formed between enhancers and core

promoter elements is stabilized by cohesin, which is recruited

by the NIPBL cohesin-loading protein that is associated with

Mediator (Kagey et al., 2010). The cohesin complex has circular

dimensions capable of encircling two nucleosome-bound mole-

cules of DNA. Reducing the levels of cohesin or NIPBL has the

same adverse effect on transcription as reducing the levels of

Mediator, so these cofactors apparently play a similarly impor-

tant role in gene activity (Kagey et al., 2010). Although cohesin

is recruited to active promoters, it also becomes associated

with the DNA-binding factor CTCF, which has been implicated

in formation of insulator elements. Thus, cohesin is thought to

have roles in transcription activation at some genes and in

silencing at others (Dorsett, 2011; Hadjur et al., 2009; Parelho

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1239

Figure 2. Master Transcriptional Regulators and Reprogramming FactorsTranscription factors that have dominant roles in the control of specific cell states and that are capable of reprogramming cell states when ectopically expressedin various cell types (Buganim et al., 2012; Davis et al., 1987; Huang et al., 2011; Ieda et al., 2010; Kajimura et al., 2009; Marro et al., 2011; Pang et al., 2011; Sekiyaand Suzuki, 2011; Takahashi and Yamanaka, 2006; Vierbuchen et al., 2010; Xie et al., 2004; Zhou et al., 2008).

et al., 2008; Phillips and Corces, 2009; Schmidt et al., 2010;

Seitan and Merkenschlager, 2012; Wendt et al., 2008).

The fundamental unit of chromatin, the nucleosome, is regu-

lated by protein complexes that can mobilize the nucleosome or

modify its histone components (Figure 1). Gene activation is

accompanied by recruitment of ATP-dependent chromatin re-

modeling complexes of the SWI/SNF family, which mobilize

nucleosomes to facilitate access of the transcription apparatus

and its regulators to DNA (Clapier and Cairns, 2009; Hargreaves

and Crabtree, 2011). In addition, there is recruitment, by tran-

scription factors and the transcription apparatus, of an array

of histone-modifying enzymes that acetylate, methylate, ubiqu-

tinylate, and otherwise chemically modify nucleosomes in

a stereotypical fashion across the span of each active gene

(Bannister and Kouzarides, 2011; Campos and Reinberg,

2009; Gardner et al., 2011; Rando, 2012; Zhu et al., 2013).

These modifications provide interaction surfaces for protein

complexes that contribute to transcriptional control. Enzymes

that remove these modifications are also typically present at

the active genes, producing a highly dynamic process of chro-

matin modification as RNA polymerase is recruited and goes

through the various steps of initiation and elongation of the

RNA species.

Repressed genes are embedded in chromatin with modifica-

tions that are characteristic of specific repression mechanisms

(Beisel and Paro, 2011; Cedar and Bergman, 2012; Jones,

2012; Moazed, 2009; Reyes-Turcu and Grewal, 2012). One

type of repressed chromatin, which contains nucleosome modi-

fications generated by the Polycomb complex (e.g., histone

H3K27me3), is found at genes that are silent but poised for acti-

vation at some later stage of development and differentiation

(Orkin and Hochedlinger, 2011; Young, 2011). Another type of

repressed chromatin is found in regions of the genome that are

fully silenced, such as that containing retrotransposons and

1240 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

other repetitive elements (Feng et al., 2010; Lejeune and Allshire,

2011). The mechanisms that silence this latter set of genes can

involve both nucleosome modification (e.g., histone H3K9me3)

and DNA methylation.

Control of Gene Expression ProgramsThe set of genes that are transcribed largely defines the cell. The

gene expression program of a specific cell type includes RNA

species from genes that are active in most cells (housekeeping

genes) and genes that are active predominantly in one or a limited

number of cell types (cell-type-specific genes). In ESCs, for

example, at least 60% of the protein-coding genes are tran-

scribed into full-length mRNA species, but only a minority are

cell-type specific and thus defining for ESCs (Assou et al.,

2007). Mammals contain hundreds and possibly thousands of

cell types, and most of these have yet to be studied with respect

to the set of transcripts they contain. Thus, the terms ‘‘house-

keeping’’ and ‘‘cell-type specific’’ are relative rather than

absolute and have yet to be precisely defined. Furthermore,

the ‘‘transcriptome’’ of specific cells, derived from high-

throughput sequencing, does not show a distinct boundary

between ‘‘active’’ and ‘‘silent’’ genes, but rather shows a broad

distribution of RNA levels that ranges from less than one RNA

molecule/gene/cell to millions of RNA molecules/gene/cell, and

it is not clear what level is functionally sufficient for each RNA

species.

The particular set of transcription factors that are expressed in

any one cell type controls the selective transcription of a subset

of genes by RNA polymerase II, thereby producing the gene

expression program of the cell. Studies of the transcription

factors that are key to establishing and maintaining specific

cell states suggest that only a small number of the transcription

factors that are expressed in cells are necessary to establish cell-

type-specific gene expression programs (Figure 2). For example,

Figure 3. Features of Master Transcription Factors of ES Cells that Likely Extend to Other Cell Types(A) Master transcription factors are expressed at high levels (30,000–300,000 molecules/cell) relative to other transcription factors.(B) Master transcription factors dominate control of the gene expression program by forming enhancers that are associated with most active ESC genes.(C) Master transcription factors positively regulate transcription of cell-type-specifying genes and, together with Polycomb group proteins, negatively regulatethe expression of genes that specify other cell types.(D) Master transcription factors (circles) positively regulate their own genes (boxes), forming interconnected autoregulatory loops.

althoughmore than half of the�1,200 genes encoding transcrip-

tion factors show some evidence of transcription in ESCs, only

a few of these transcription factors are needed to reprogram

a broad range of cell types into induced pluripotent stem cells

(iPSCs) with features essentially indistinguishable from ESCs

(Graf, 2011; Ng and Surani, 2011; Orkin and Hochedlinger,

2011; Yamanaka, 2012; Yeo and Ng, 2013; Young, 2011). These

ESC transcription factors, which include OCT4, SOX2, and

NANOG, are expressed at high levels, bind regulatory elements

associated with most active ESC genes, are involved in

Polycomb-mediated repression of genes that specify other cell

types, and positively regulate their own gene expression through

interconnected autoregulatory loops (Figure 3) (Young, 2011).

Activation of these endogenous interconnected autoregulatory

loops may be key to cellular reprogramming by introduction of

exogenous transcription factors. Other cell types express cell-

type-specific, or lineage-specific, master transcription factors

that are likely to share these key properties of the ESC master

transcription factors.

Most of the transcription factors that are key to control of cell

state and that can act as reprogramming factors are thought to

control transcription initiation at the genes they regulate. For

example, the ESC transcription factors OCT4 and NANOG

bind to the P300 and Mediator coactivators (Chen et al., 2008;

Kagey et al., 2010), which can then drive the formation of

open chromatin and recruitment of the transcription apparatus.

Similarly, many of the transcription factors that can reprogram

or transdifferentiate cells, including MYOD, C/EBPb, HNF1a,

HNF4a, BRN2, and GATA4, bind to at least one of these

coactivators (Borggrefe and Yue, 2011).

Recent studies have revealed that certain transcription factors

can exert a broad effect on the gene expression programs of

cells through elongation control (Figure 4). The c-Myc transcrip-

tion factor can stimulate increased elongation from essentially

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1241

Figure 4. Global Alterations in Gene Expression Programs through

Transcription Elongation(A) Transcriptional amplification of the gene expression program. Thetranscription factor c-Myc stimulates increased elongation of most activelytranscribed genes, producing increased levels of transcripts for most genes inthe gene expression program of the cell. (B) Expanded pause release extendsthe gene expression program. In some cells, RNA polymerase will initiatetranscription at some genes but fails to transition to elongation. AIREstimulates pause release for many of these initiated genes, thus producingtranscripts for many genes that are normally expressed only in peripheraltissues. (C) Specific pause release. Some elongation factors stimulate pauserelease at specific sets of genes that are important for a particular cell’sfunction.

the entire active gene expression program in diverse cell types

(Lin et al., 2012; Nie et al., 2012; Rahl et al., 2010). The transcrip-

tion factor AIRE functions to expand the set of genes that

undergo RNA polymerase II pause release in specialized thymic

1242 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

stromal cells, allowing expression of the broad spectrum of self-

antigens necessary to induce immune tolerance (Abramson

et al., 2010; Giraud et al., 2012; Oven et al., 2007; �Zumer et al.,

2011). In hematopoiesis, the TIF1g transcription factor controls

erythroid cell fate by interacting with P-TEFb and regulating tran-

scription elongation at a specific set of target genes (Bai et al.,

2010). Development generally appears to be dependent on

proper elongation control; the transcription elongation factor

Tcea3 (TFIIS) contributes to the ability of ESCs to respond

appropriately to differentiation cues (Park et al., 2013), andmuta-

tions in the P-TEFb repressor HEXIM cause gross develop-

mental defects (Nguyen et al., 2012).

Summary of Recent Advances in Gene RegulationThe key themes that have emerged from recent studies in

transcriptional control and that are highlighted here are the

following. Sequence variation in enhancers plays an important

role in misregulation of gene expression and disease. A small

number of key transcription factors dominate control of gene

expression programs. Some transcription factors regulate tran-

scription initiation, whereas other factors control elongation,

and factors that control this latter step can have profound effects

on cell state. The Mediator coactivator complex integrates

signals from diverse regulators and recruits cohesin complexes

to active genes, which, in turn, contributes to both chromatin

looping and gene activity. Diverse chromatin regulators mobilize

nucleosomes and dynamically modify nucleosomes during

active gene transcription and in gene silencing, and some chro-

matin regulators are regulated by lncRNAs. These advances in

our understanding of sequences involved in gene control, tran-

scriptional circuitry, the transcription apparatus, and chromatin

regulation have led to new insights into themechanisms involved

in misregulation of gene expression in various human diseases

and disorders. We discuss some of these below.

Misregulated Gene Expression in DiseaseMany diseases and syndromes are associated with mutations in

regulatory regions and in transcription factors, cofactors, chro-

matin regulators and noncoding RNAs (Table S1 available on-

line). These mutations can contribute to cancer, autoimmunity,

neurological disorders, developmental syndromes, diabetes,

cardiovascular disease, and obesity, among others. We highlight

here several insights into disease mechanisms that have

emerged from advances in our understanding of gene regulation.

Cancer

Recent studies have highlighted the link between disease-

associated variants in regulatory DNA and breast cancer (Jiang

et al., 2011), prostate cancer (Demichelis et al., 2012), colorectal

cancer (Lubbe et al., 2012), renal cancer (Schodel et al., 2012),

lung cancer (Liu et al., 2011), nasopharyngeal cancer (Yew

et al., 2012), and melanoma (Huang et al., 2013; Horn et al.,

2013). The genome instability that is a hallmark of cancer almost

certainly contributes to further alter sequences in regulatory

regions that can promote tumor progression.

Mutations in transcription factors have long been known

to contribute to tumorigenesis, and recent studies indicate

that overexpressed oncogenic transcription factors can alter

the core autoregulatory circuitry of the cell. The oncogenic

transcription factor TAL1, which is overexpressed in approxi-

mately half of the cases of T cell acute lymphoblastic leukemia

(T-ALL), forms an interconnected autoregulatory loop with

several key transcription factor partners, and this circuitry

contributes to the sustained activation of TAL1-regulated onco-

genic program (Sanda et al., 2012). Thus, high levels of TAL1

produce amodified autoregulatory circuitry that drives the onco-

genic program in T-ALL.

Most tumor cells depend on the transcription factor c-Myc for

their growth and proliferation (Littlewood et al., 2012).MYC is the

most frequently amplified oncogene, and the elevated expres-

sion of its gene product is associated with tumor aggression

and poor clinical outcome. Elevated levels of c-Myc can promote

tumorigenesis in a wide range of tissues. In tumor cells express-

ing high levels of c-Myc, the transcription factor accumulates in

the promoter regions of most active genes, recruits the tran-

scription elongation factor P-TEFb, and causes transcriptional

amplification, producing increased levels of transcripts within

the cell’s gene expression program (Lin et al., 2012; Nie et al.,

2012). Thus, rather than binding and regulating a new set of

genes when overexpressed, c-Myc amplifies the output of the

existing gene expression program (Figure 4). These results

suggest that transcriptional amplification reduces rate-limiting

constraints for tumor cell growth and proliferation.

Mutations in the Mediator coactivator complex have recently

been implicated in the development of various tumors. Uterine

leiomyomas, or fibroids, are benign tumors that affect millions

of women. The MED12 gene is altered in the majority of uterine

leiomyomas, and its expression is absent in many uterine

leiomyosarcomas, the malignant counterparts of leiomyomas

(Makinen et al., 2011a, 2011b). MED12 mutations also occur

frequently in prostate cancer (Barbieri et al., 2012). MED12 is

part of the CDK module of the Mediator complex, and the

CDK8 subunit of this module has been reported to act as an

oncogene in both colon cancer and melanoma (Firestein et al.,

2008; Kapoor et al., 2010; Morris et al., 2008). Mediator has roles

in gene activation and repression and can function both in

transcription initiation and elongation, so further study is needed

to establish howMediator mutations contribute to these tumors.

Mediator-associated NIPBL recruits cohesin. Alterations in co-

hesin expression and function have been noted in some cancer

cells, and there is speculation that cohesin misregulation may

also contribute to development of various cancers, but direct

evidence for a role of cohesin in cancer remains to be estab-

lished (Mannini and Musio, 2011; Xu et al., 2011).

Mutations in a variety of chromatin regulators have been

implicated in development of cancer cells, and the normal func-

tions of these regulators provide some clues to the mechanisms

involved in altered gene expression. Loss-of-function mutations

in several nucleosome remodeling proteins, including ARID1A,

SMARCA4 (BRG1), and SMARCB1 (INI1), are associated with

multiple types of cancer (Dawson and Kouzarides, 2012;

Hargreaves and Crabtree, 2011; Tsai and Baylin, 2011; Wilson

and Roberts, 2011), suggesting that defects in mobilizing nucle-

osomes near the promoters of active genes are involved. Simi-

larly, various mutations in the Polycomb components EZH2

and SUZ12 and in the DNA methylation apparatus occur in

multiple cancers, suggesting that, in these instances, it is the

loss of proper gene silencing that contributes to tumorigenesis

(Cedar and Bergman, 2012; Jones, 2012; Margueron and

Reinberg, 2011; Mills, 2010). The majority of malignant mela-

nomas overexpress SETDB1, a histone H3K9 methyltransferase

that can contribute to gene activation or silencing, and this

causes deregulation of HOX genes and accelerates melanoma

(Ceol et al., 2011).

Gene fusions with the chromatin regulator MLL in leukemias

are now known to alter transcription elongation (Luo et al.,

2012b; Marschalek, 2010; Slany, 2009; Smith et al., 2011a).

Several translocation partners of MLL are components of

a SEC that includes P-TEFb and ELL proteins, which have also

been shown to control transcription elongation (Lin et al., 2010,

2011; Luo et al., 2012a; Smith et al., 2011b). It is thought that

translocation of any of the SEC subunits to the amino-terminal

domain of MLL abnormally stabilizes the localization of the

SEC at MLL target genes, including HOXA9 and HOXA10, which

leads to excessive stimulation of RNA polymerase II into produc-

tive elongation at these genomic loci, which in turn contributes to

aggressive acute leukemia.

Specific lncRNAs have recently been implicated in cancer

progression. The ANRIL lncRNAmediates transcriptional repres-

sion of members of the INK4A/ARF/INK4B locus, which encode

tumor suppressors whose repression is associated with various

cancers (Aguilo et al., 2011; Popov and Gil, 2010). ANRIL func-

tions by recruiting polycomb repressive complexes 1 and 2

(PRC1 and PRC2), and misregulation of ANRIL may lead to

abnormal silencing of tumor suppressors and thus contribute

to cancer progression (Kotake et al., 2011; Yap et al., 2010).

Interestingly, genome-wide association studies have identified

numerous polymorphisms that affect the expression and pro-

cessing of ANRIL and are associated with increased suscepti-

bility to an increasing variety of disease states, including multiple

types of cancer, coronary artery disease, and type 2 diabetes

(Pasmant et al., 2011; Harismendy et al., 2011).

Autoimmunity and Inflammation

Mutations in the autoimmune regulator (AIRE) protein cause type

I autoimmune polyendocrinopathy syndrome. AIRE is a tran-

scription factor whose role in promoting transcriptional elonga-

tion at genes with paused RNA polymerase II in the thymus

explains why loss of AIRE function leads to autoimmune disease.

Self-reactive T cells are normally eliminated during maturation in

the thymus due to the specialized ability of thymic stromal cells

and, in particular, medullary epithelial cells (MECs) to transcribe

a large repertoire of genes encoding peripheral tissue antigens

(Kyewski and Klein, 2006). This ectopic gene expression is

controlled in a large part by AIRE, which is expressed almost

exclusively in MECs. Mice and humans with an AIRE gene defect

express only a fraction of the peripheral tissue antigens and

develop immune infiltrates and autoantibodies directed at

multiple peripheral tissues (Akirav et al., 2011; Gardner et al.,

2009; Mathis and Benoist, 2009; Metzger and Anderson,

2011). Recent studies have shown that AIRE interacts with

P-TEFb and influences transcription elongation in primary

MECs (Abramson et al., 2010; Giraud et al., 2012; Oven et al.,

2007; �Zumer et al., 2011). AIRE is physically associated with all

the active genes in MECs but has its greatest effect on genes

that do not experience pause release in its absence (Figure 4).

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1243

These results are consistent with the idea that AIRE causes the

release of RNA polymerase II molecules that are nonproductively

paused at the promoters of a broad spectrum of genes that are

otherwise expressed only in peripheral tissues. Thus, in MECs,

AIRE’s function is to expand the set of genes that undergoes

RNA polymerase II pause release.

Misregulation of the immune response transcriptional regu-

lator NF-kB has been linked to inflammatory and autoimmune

diseases, improper immune development, and cancer. NF-kB

is found in most cell types and is involved in cellular responses

to stimuli such as infection and stress (Hayden and Ghosh,

2012). The transcription factor controls genes involved in inflam-

mation and is chronically active in inflammatory diseases such as

inflammatory bowel disease, arthritis, sepsis, gastritis, asthma,

and atherosclerosis. Although most research into the mecha-

nism of transcriptional activation by this and other regulators

has focused on coactivator recruitment, evidence that NF-kB

interacts with BRD4 and P-TEFb suggests that this ubiquitous

regulator plays a role in elongation control at inflammatory genes

during immune and stress responses (Barboric et al., 2001;

Huang et al., 2009; Nowak et al., 2008). This view is supported

by evidence that inhibitors of BRD4, which contributes to recruit-

ing active P-TEFb, suppress expression of key inflammatory

genes in activated macrophages and confer protection against

lipopolysaccharide-induced endotoxic shock and bacteria-

induced sepsis (Nicodeme et al., 2010).

Developmental Disorders: Neurological

Mutations in various components of the Mediator coactivator

have been linked to a variety of neurological disorders and other

developmental deficiencies (Ding et al., 2008; Goh and Grants,

2012; Hashimoto et al., 2011; Kaufmann et al., 2010; Leal

et al., 2009; Philibert et al., 2007; Risheg et al., 2007; Rump

et al., 2011; Schwartz et al., 2007; Zhou et al., 2012a). Mutations

in MED23 alter the interaction between enhancer-bound tran-

scription factors and Mediator, leading to transcriptional dysre-

gulation of mitogen-responsive immediate-early genes that

affect brain development and plasticity. A similar defect in imme-

diate-early gene expression is observed in cells from patients

with another intellectual disability, Opitz-Kaveggia syndrome,

which is caused byMED12mutations. It would not be surprising

to find that additional Mediator mutations contribute to neurolog-

ical disorders, given the role of this coactivator in integrating

information from transcriptional activators, repressors, and

signaling pathways.

Heterozygous germline mutations in components of the SWI/

SNF chromatin remodeling complex were recently identified in

patients with various neurological syndromes whose common

features are severe intellectual disability and speech delay

(Hoyer et al., 2012; Santen et al., 2012a, 2012b; Tsurusaki

et al., 2012; Van Houdt et al., 2012). These mutations were found

in SMARCB1, SMARCA4, SMARCA2, SMARCE1, ARID1A, and

ARID1B. It has been suggested that up to 3% of unexplained

intellectual disability may be caused by mutations in genes

encoding SWI/SNF components (Santen et al., 2012b). It is inter-

esting to note that ARID1B component of human SWI/SNF

interacts with elongin C (Li et al., 2010), a component of the

SIII transcription elongation factor, which enhances transcrip-

tion elongation by suppressing transient pausing of RNA

1244 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

polymerase II (Aso et al., 1995). Thus, alterations in SWI/SNF

complexes have the potential to affect both chromatin remodel-

ing and transcription elongation.

Developmental Disorders: Cohesinopathies

Cohesinopathies are characterized by a wide variety of develop-

mental defects, including growth and mental retardation, limb

deformities, and craniofacial anomalies (Bose and Gerton,

2010; Liu and Krantz, 2008). This broad spectrum of phenotypes

is now thought to be due to reduced cohesin loading and cohesin

function in gene expression during development. A variety of

cohesinopathies have been described, including Cornelia de

Lange syndrome and Roberts syndrome, in which patients

have mutations in the cohesin loading protein NIPBL or the

proteins that constitute the cohesin complex. With recent

evidence for roles of cohesin complexes in regulation of gene

expression and DNA looping (Kagey et al., 2010; Kawauchi

et al., 2009; Liu et al., 2009; Schaaf et al., 2009; Seitan et al.,

2011), it has become apparent that these deficiencies lead to

defects in transcriptional regulation and probably to the overall

structure of chromatin in the nucleus of disease cells.

Diabetes

Diabetes mellitus is a group of metabolic diseases in which

a person has elevated blood sugar either because the pancreas

fails to produce adequate amounts of insulin or because cells do

not respond properly to the insulin that is produced. Mutations in

pancreatic master transcription factors and the sequences they

bind have been implicated in diabetes. The gene expression

programs of pancreatic cells appear to be controlled by a small

set of key transcription factors, including HNF1a, HNF1b,

HNF4a, PDX1, and NEUROD1, some of which contribute to

the interconnected autoregulatory circuitry of these cells

(Odom et al., 2004). Mutations in any of these factors can result

in various forms of maturity-onset diabetes of the young (MODY)

(Maestro et al., 2007; Malecki, 2005). These mutations almost

certainly have a deleterious effect on the interconnected autore-

gulatory circuitry formed by these factors and their target genes.

The frequency of SNPs that are linked to defects in glucose

homeostasis and diabetes is greatly enriched in the binding sites

for these transcription factors (Maurano et al., 2012). This obser-

vation indicates that perturbations that affect the regulatory

circuitry of pancreatic cells may contribute to diabetes. It also

suggests that previously undiscovered regulatory networks

and network architectures may be uncovered by incorporating

information about disease-associated genetic variants and

knowledge of the binding sites of diverse transcription factors.

Cardiovascular Disease

Misregulated development of the cardiovascular system is

among the most common class of congenital birth defects,

and diseases of the cardiovascular system are among the

most prevalent clinical issues for adult populations (Bruneau,

2008; Kathiresan and Srivastava, 2012; Roger et al., 2012). It is

well established that loss-of-function mutations in certain tran-

scription factors cause various cardiovascular deficiencies

(Table S1), but new studies have highlighted the role that muta-

tions in ncRNA species can play in cardiovascular diseases.

Specific miRNAs have been implicated in both the promo-

tion and inhibition of differentiation into cardiac lineages,

cardiac hypertrophy, vascular differentiation, and erythropoiesis

(Han et al., 2011; Papageorgiou et al., 2012; Small and Olson,

2011). MicroRNAs have also been linked to causative and

protective roles for multiple types of cardiovascular disease,

including arrhythmia, fibrosis, hypertrophy due to high pressure,

and misregulation of cardiac energy metabolism (Callis et al.,

2009; Care et al., 2007; Grueter et al., 2012; Thum et al., 2008;

van Rooij et al., 2007, 2008; Yang et al., 2007). MicroRNAs are

thought to fine-tune gene expression, and thus, the alterations

in these cases are thought to lead to deficiencies in fine-tuning

the cardiovascular gene expression program.

Summary and OutlookSeveral concepts have emerged from recent studies of gene

expression programs in healthy and in diseased cells. Genetic

variationmay contribute to disease largely throughmisregulation

of gene expression. Mutations in the transcription factors that

control cell state may impact the autoregulatory loops that are

at the core of cellular regulatory circuitry, leading to the loss of

a normal healthy cell state. Some transcription factors control

RNA polymerase II pause release and elongation and, when their

expression or function is altered, can produce aggressive tumor

cells (c-Myc) or some forms of autoimmunity (AIRE). Mutations in

the coactivator complexes that integrate information from many

transcription factors and contribute to DNA looping can cause

a broad spectrum of developmental diseases. Alterations in

specific chromatin regulators can contribute to development of

cancer and many other diseases. Misregulation of noncoding

RNAs can also contribute to disease. Additional insights into

the role of transcriptional misregulation in human disease will

require improved genome annotation, knowledge of the DNA

sequences whose alterations contribute to disease, identifica-

tion of the key transcriptional regulators of all cells of medical

relevance, and further understanding of the roles of cofactors,

chromatin regulators, and ncRNAs.

It is essential to improve human genome annotation in order to

more fully understand gene expression programs and their regu-

lation and, thus, gene misregulation in disease. As a first step, it

would be ideal to identify the complete set of protein-coding and

noncoding genes and to ascertain which of these are actively

transcribed in specific cell types. There are considerable chal-

lenges associated with defining the complete set of expressed

genes in any one type of mammalian cell. Such characterization

has traditionally required large numbers of cells, and for most

primary cell types, it is challenging to obtain a homogeneous

population of cells. Whereas protein-coding genes can be

recognized, at least in part, by the presence of a coding

sequence, it is challenging to produce a complete and accurate

annotation of ncRNA genes due to limitations in the read length

of widely used sequence technologies and the short lifetime of

many ncRNAs. Nonetheless, recent studies have identified

a vast number and variety of ncRNAs in human cells, so there

is promise that improved human genome annotation is at hand

(Djebali et al., 2012). Furthermore, new technologies allow inves-

tigators tomonitor RNA polymerase II molecules that are actively

engaged in transcription (Core et al., 2008). Such approaches

have recently provided evidence that most lncRNA species are

the product of divergent transcription from the promoters of

active protein coding genes (Sigova et al., 2013). This suggests

that most active protein-coding genes in humans are actually

divergently transcribed mRNA/lncRNA gene pairs.

Knowledge of the sequence variation that contributes to

disease is being gained at a rapid pace, and this will improve

our understanding of disease mechanisms and lead to new

approaches to disease diagnosis and therapy. Several lines of

evidence suggest that much of genetic variation contributes to

disease through misregulation of gene expression. A substantial

portion of the genomic sequences that are under positive selec-

tion are thought to be regulatory (Grossman et al., 2010).

Disease-associated SNPs are enriched in regulatory regions

(Ernst et al., 2011; Hindorff et al., 2009; Maurano et al., 2012).

Many recent studies have identified links between disease-asso-

ciated variants in regulatory DNA and a broad spectrum of

human diseases, including cancer (Demichelis et al., 2012;

Huang et al., 2013; Horn et al., 2013; Jiang et al., 2011; Liu

et al., 2011; Lubbe et al., 2012; Schodel et al., 2012; Yew

et al., 2012), congenital heart disease (Zhao et al., 2012), inflam-

matory lung disease (Han et al., 2012), multiple sclerosis (Alcina

et al., 2013), Alzheimer’s disease (Gaj et al., 2012), abdominal

aortic aneurysm (Bown et al., 2011), amyotrophic lateral scle-

rosis (Iida et al., 2011), and coronary artery disease (Harismendy

et al., 2011). Disease-associated genetic variants in regulatory

regions are most often found in regions that are utilized in

a cell-type-specific manner and are associated with diseases

of the corresponding cell type (Ernst et al., 2011; Maurano

et al., 2012). Thus, cell-type-specific enhancer use can explain

how genetic variants produce tissue-specific diseases.

Disease-associated genetic variants can exist in regulatory

regions that are very distant from the genes they control, but

knowledge of the nature of loops between such distal enhancers

and their target genes can explain how these distant variants

affect a specific gene and its biological functions (Li et al.,

2012; Maurano et al., 2012).

Genetic investigations and reprogramming studies suggest

that only a small number of the hundreds of transcription factors

that are expressed in cells are essential for establishing and

maintaining the regulatory networks that produce specific cell

states. If this holds true for most cell types, then it would be ideal

to identify the key transcription factors for all cell types of

medical relevance. It should be possible to identify these tran-

scription factors if they have features identified for their counter-

parts in well-studied cells: relative high expression, occupancy

of enhancers associated with a large fraction of active genes,

and formation of interconnected autoregulatory loops. Discov-

ering how gene expression programs are controlled in many

different cell types should lead to further understanding of regu-

latory circuitry, facilitate cellular reprogramming, and accelerate

the new field of regenerative medicine.

Cofactors and chromatin regulators are generally expressed in

most cell types, but mutations in these genes often produce

diseases or syndromes that exhibit tissue-specific disease

phenotypes (Table S1). For example, defects in Mediator

subunits contribute to nonsyndromic intellectual disability,

Charcot-Marie-Tooth disease, Opitz-Kaveggia and Lujan

syndromes, and infantile cerebral and cerebellar atrophy and

have been implicated in prostate cancer, deficiencies in

systemic energy homeostasis (Grueter et al., 2012), and altered

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1245

hair-cycling (Nakajima et al., 2013; Oda et al., 2012). Improved

understanding of the interactions between cofactors and tran-

scription factors and the mechanisms involved in information

integration by these complex apparatuses will be valuable for

understanding the mechanisms that produce tissue-specific

phenotypes. Similarly, it will be important to further understand

the collaboration between the transcription apparatus and chro-

matin regulators in global control of gene expression programs.

Recent efforts to target chromatin regulators for cancer therapy

(Dawson et al., 2012) would benefit from a fuller understanding of

the regulatory mechanisms and pathways that are impacted by

these potential therapeutics.

Our future understanding of disease and the advance of

personalized medicine will benefit from models of human tran-

scriptional regulatory circuitry that integrate information about

regulatory sequences and the key transcription factors, cofac-

tors, chromatin regulators, and ncRNAs that operate at regula-

tory sites. The development of these models should thus be

among the priorities of biomedical research.

SUPPLEMENTAL INFORMATION

Supplemental Information includes one table and can be found with this article

online at http://dx.doi.org/10.1016/j.cell.2013.02.014.

ACKNOWLEDGMENTS

Our description of the themes highlighted in this review benefited from

discussions with Karen Adelman, Jay Bradner, Gerald Crabtree, Rudolf

Jaenisch, Ian Krantz, Lee Lawton, David Levens, John Lis, Alex Marson,

Matthias Merkenschlager, Alan Mullen, Duncan Odom, David Price, Peter

Rahl, Robert Roeder, Ali Shilatifard, Phil Sharp, Alla Sigova, Alexander Stark,

Dylan Taatjes, Robert Weinberg, and Leonard Zon. We thank David Orlando

for help with data collation and analysis. This work was supported by National

Institutes of Health grants HG002668 (R.A.Y.) and CA146445 (R.A.Y.

and T.I.L.).

REFERENCES

Abramson, J., Giraud, M., Benoist, C., and Mathis, D. (2010). Aire’s partners in

the molecular control of immunological tolerance. Cell 140, 123–135.

Adelman, K., and Lis, J.T. (2012). Promoter-proximal pausing of RNA

polymerase II: emerging roles in metazoans. Nat. Rev. Genet. 13, 720–731.

Aguilo, F., Zhou, M.M., and Walsh, M.J. (2011). Long noncoding RNA,

polycomb, and the ghosts haunting INK4b-ARF-INK4a expression. Cancer

Res. 71, 5365–5369.

Akirav, E.M., Ruddle, N.H., and Herold, K.C. (2011). The role of AIRE in human

autoimmune disease. Nat. Rev. Endocrinol. 7, 25–33.

Alcina, A., Fedetz, M., Fernandez, O., Saiz, A., Izquierdo, G., Lucas, M., Leyva,

L., Garcıa-Leon, J.A., Abad-Grau, Mdel.M., Alloza, I., et al. (2013). Identifica-

tion of a functional variant in the KIF5A-CYP27B1-METTL1-FAM119B locus

associated with multiple sclerosis. J. Med. Genet. 50, 25–33.

Aravin, A., Gaidatzis, D., Pfeffer, S., Lagos-Quintana, M., Landgraf, P., Iovino,

N., Morris, P., Brownstein, M.J., Kuramochi-Miyagawa, S., Nakano, T., et al.

(2006). A novel class of small RNAs bind to MILI protein in mouse testes.

Nature 442, 203–207.

Aso, T., Lane, W.S., Conaway, J.W., and Conaway, R.C. (1995). Elongin (SIII):

a multisubunit regulator of elongation by RNA polymerase II. Science 269,

1439–1443.

Assou, S., Le Carrour, T., Tondeur, S., Strom, S., Gabelle, A., Marty, S., Nadal,

L., Pantesco, V., Reme, T., Hugnot, J.P., et al. (2007). A meta-analysis of

1246 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

human embryonic stem cells transcriptome integrated into a web-based

expression atlas. Stem Cells 25, 961–973.

Bai, X., Kim, J., Yang, Z., Jurynec, M.J., Akie, T.E., Lee, J., LeBlanc, J., Sessa,

A., Jiang, H., DiBiase, A., et al. (2010). TIF1gamma controls erythroid cell fate

by regulating transcription elongation. Cell 142, 133–143.

Bannister, A.J., and Kouzarides, T. (2011). Regulation of chromatin by histone

modifications. Cell Res. 21, 381–395.

Barbieri, C.E., Baca, S.C., Lawrence, M.S., Demichelis, F., Blattner, M.,

Theurillat, J.P., White, T.A., Stojanov, P., Van Allen, E., Stransky, N., et al.

(2012). Exome sequencing identifies recurrent SPOP, FOXA1 and MED12

mutations in prostate cancer. Nat. Genet. 44, 685–689.

Barboric, M., Nissen, R.M., Kanazawa, S., Jabrane-Ferrat, N., and Peterlin,

B.M. (2001). NF-kappaB binds P-TEFb to stimulate transcriptional elongation

by RNA polymerase II. Mol. Cell 8, 327–337.

Bartel, D.P. (2009). MicroRNAs: target recognition and regulatory functions.

Cell 136, 215–233.

Beisel, C., and Paro, R. (2011). Silencing chromatin: comparing modes and

mechanisms. Nat. Rev. Genet. 12, 123–135.

Berk, A.J. (2012). Yin and yang of mediator function revealed by human

mutants. Proc. Natl. Acad. Sci. USA 109, 19519–19520.

Bonasio, R., Tu, S., and Reinberg, D. (2010). Molecular signals of epigenetic

states. Science 330, 612–616.

Borggrefe, T., and Yue, X. (2011). Interactions between subunits of the

Mediator complex with gene-specific transcription factors. Semin. Cell Dev.

Biol. 22, 759–768.

Bose, T., and Gerton, J.L. (2010). Cohesinopathies, gene expression, and

chromatin organization. J. Cell Biol. 189, 201–210.

Bown, M.J., Jones, G.T., Harrison, S.C., Wright, B.J., Bumpstead, S., Baas,

A.F., Gretarsdottir, S., Badger, S.A., Bradley, D.T., Burnand, K., et al.;

CARDIoGRAM Consortium; Global BPgen Consortium; DIAGRAM Consor-

tium; VRCNZ Consortium. (2011). Abdominal aortic aneurysm is associated

with a variant in low-density lipoprotein receptor-related protein 1. Am. J.

Hum. Genet. 89, 619–627.

Brennecke, J., Aravin, A.A., Stark, A., Dus, M., Kellis, M., Sachidanandam, R.,

and Hannon, G.J. (2007). Discrete small RNA-generating loci as master regu-

lators of transposon activity in Drosophila. Cell 128, 1089–1103.

Bruneau, B.G. (2008). The developmental genetics of congenital heart disease.

Nature 451, 943–948.

Buecker, C., and Wysocka, J. (2012). Enhancers as information integration

hubs in development: lessons from genomics. Trends Genet. 28, 276–284.

Buganim, Y., Itskovich, E., Hu, Y.C., Cheng, A.W., Ganz, K., Sarkar, S., Fu, D.,

Welstead, G.G., Page, D.C., and Jaenisch, R. (2012). Direct reprogramming of

fibroblasts into embryonic Sertoli-like cells by defined factors. Cell Stem Cell

11, 373–386.

Callis, T.E., Pandya, K., Seok, H.Y., Tang, R.H., Tatsuguchi, M., Huang, Z.P.,

Chen, J.F., Deng, Z., Gunn, B., Shumate, J., et al. (2009). MicroRNA-208a is

a regulator of cardiac hypertrophy and conduction in mice. J. Clin. Invest.

119, 2772–2786.

Campos, E.I., and Reinberg, D. (2009). Histones: annotating chromatin. Annu.

Rev. Genet. 43, 559–599.

Care, A., Catalucci, D., Felicetti, F., Bonci, D., Addario, A., Gallo, P., Bang,

M.L., Segnalini, P., Gu, Y., Dalton, N.D., et al. (2007). MicroRNA-133 controls

cardiac hypertrophy. Nat. Med. 13, 613–618.

Cedar, H., and Bergman, Y. (2012). Programming of DNAmethylation patterns.

Annu. Rev. Biochem. 81, 97–117.

Ceol, C.J., Houvras, Y., Jane-Valbuena, J., Bilodeau, S., Orlando, D.A.,

Battisti, V., Fritsch, L., Lin, W.M., Hollmann, T.J., Ferre, F., et al. (2011). The

histone methyltransferase SETDB1 is recurrently amplified in melanoma and

accelerates its onset. Nature 471, 513–517.

Chen, X., Xu, H., Yuan, P., Fang, F., Huss,M., Vega, V.B.,Wong, E., Orlov, Y.L.,

Zhang, W., Jiang, J., et al. (2008). Integration of external signaling pathways

with the core transcriptional network in embryonic stem cells. Cell 133, 1106–

1117.

Clapier, C.R., and Cairns, B.R. (2009). The biology of chromatin remodeling

complexes. Annu. Rev. Biochem. 78, 273–304.

Conaway, R.C., and Conaway, J.W. (2011). Function and regulation of the

Mediator complex. Curr. Opin. Genet. Dev. 21, 225–230.

Core, L.J., Waterfall, J.J., and Lis, J.T. (2008). Nascent RNA sequencing

reveals widespread pausing and divergent initiation at human promoters.

Science 322, 1845–1848.

Davis, R.L., Weintraub, H., and Lassar, A.B. (1987). Expression of a single

transfected cDNA converts fibroblasts to myoblasts. Cell 51, 987–1000.

Dawson, M.A., and Kouzarides, T. (2012). Cancer epigenetics: from mecha-

nism to therapy. Cell 150, 12–27.

Dawson, M.A., Kouzarides, T., and Huntly, B.J. (2012). Targeting epigenetic

readers in cancer. N. Engl. J. Med. 367, 647–657.

Demichelis, F., Setlur, S.R., Banerjee, S., Chakravarty, D., Chen, J.Y., Chen,

C.X., Huang, J., Beltran, H., Oldridge, D.A., Kitabayashi, N., et al. (2012).

Identification of functionally active, low frequency copy number variants at

15q21.3 and 12q21.31 associated with prostate cancer risk. Proc. Natl.

Acad. Sci. USA 109, 6686–6691.

Dikstein, R. (2011). The unexpected traits associated with core promoter

elements. Transcription 2, 201–206.

Ding, N., Zhou, H., Esteve, P.O., Chin, H.G., Kim, S., Xu, X., Joseph, S.M.,

Friez, M.J., Schwartz, C.E., Pradhan, S., and Boyer, T.G. (2008). Mediator links

epigenetic silencing of neuronal gene expression with x-linked mental retarda-

tion. Mol. Cell 31, 347–359.

Djebali, S., Davis, C.A., Merkel, A., Dobin, A., Lassmann, T., Mortazavi, A.,

Tanzer, A., Lagarde, J., Lin, W., Schlesinger, F., et al. (2012). Landscape of

transcription in human cells. Nature 489, 101–108.

Dorsett, D. (2011). Cohesin: genomic insights into controlling gene transcrip-

tion and development. Curr. Opin. Genet. Dev. 21, 199–206.

Dunham, I., Kundaje, A., Aldred, S.F., Collins, P.J., Davis, C.A., Doyle, F.,

Epstein, C.B., Frietze, S., Harrow, J., Kaul, R., et al.; ENCODE Project Consor-

tium. (2012). An integrated encyclopedia of DNA elements in the human

genome. Nature 489, 57–74.

Ebert, M.S., and Sharp, P.A. (2012). Roles for microRNAs in conferring robust-

ness to biological processes. Cell 149, 515–524.

Ernst, J., Kheradpour, P., Mikkelsen, T.S., Shoresh, N., Ward, L.D., Epstein,

C.B., Zhang, X., Wang, L., Issner, R., Coyne, M., et al. (2011). Mapping and

analysis of chromatin state dynamics in nine human cell types. Nature 473,

43–49.

Feng, S., Jacobsen, S.E., and Reik, W. (2010). Epigenetic reprogramming in

plant and animal development. Science 330, 622–627.

Firestein, R., Bass, A.J., Kim, S.Y., Dunn, I.F., Silver, S.J., Guney, I., Freed, E.,

Ligon, A.H., Vena, N., Ogino, S., et al. (2008). CDK8 is a colorectal cancer

oncogene that regulates beta-catenin activity. Nature 455, 547–551.

Fuda, N.J., Ardehali, M.B., and Lis, J.T. (2009). Defining mechanisms that

regulate RNA polymerase II transcription in vivo. Nature 461, 186–192.

Gaj, P., Paziewska, A., Bik, W., Dabrowska, M., Baranowska-Bik, A.,

Styczynska, M., Chodakowska- _Zebrowska, M., Pfeffer-Baczuk, A.,

Barcikowska, M., Baranowska, B., and Ostrowski, J. (2012). Identification of

a late onset Alzheimer’s disease candidate risk variant at 9q21.33 in Polish

patients. J. Alzheimers Dis. 32, 157–168.

Gardner, J.M., Fletcher, A.L., Anderson, M.S., and Turley, S.J. (2009). AIRE in

the thymus and beyond. Curr. Opin. Immunol. 21, 582–589.

Gardner, K.E., Allis, C.D., and Strahl, B.D. (2011). Operating on chromatin,

a colorful language where context matters. J. Mol. Biol. 409, 36–46.

Girard, A., Sachidanandam, R., Hannon, G.J., and Carmell, M.A. (2006). A

germline-specific class of small RNAs binds mammalian Piwi proteins. Nature

442, 199–202.

Giraud, M., Yoshida, H., Abramson, J., Rahl, P.B., Young, R.A., Mathis, D., and

Benoist, C. (2012). Aire unleashes stalled RNA polymerase to induce ectopic

gene expression in thymic epithelial cells. Proc. Natl. Acad. Sci. USA 109,

535–540.

Goh, Y.S., and Grants, J.M. (2012). Mutations in the Mediator subunit

MED23 link intellectual disability to immediate early gene regulation. Clin.

Genet. 81, 430.

Goodrich, J.A., and Tjian, R. (2010). Unexpected roles for core promoter

recognition factors in cell-type-specific transcription and gene regulation.

Nat. Rev. Genet. 11, 549–558.

Graf, T. (2011). Historical origins of transdifferentiation and reprogramming.

Cell Stem Cell 9, 504–516.

Grossman, S.R., Shlyakhter, I., Karlsson, E.K., Byrne, E.H., Morales, S.,

Frieden, G., Hostetter, E., Angelino, E., Garber, M., Zuk, O., et al. (2010). A

composite of multiple signals distinguishes causal variants in regions of posi-

tive selection. Science 327, 883–886.

Grueter, C.E., van Rooij, E., Johnson, B.A., DeLeon, S.M., Sutherland, L.B., Qi,

X., Gautron, L., Elmquist, J.K., Bassel-Duby, R., and Olson, E.N. (2012). A

cardiac microRNA governs systemic energy homeostasis by regulation of

MED13. Cell 149, 671–683.

Hadjur, S., Williams, L.M., Ryan, N.K., Cobb, B.S., Sexton, T., Fraser, P.,

Fisher, A.G., and Merkenschlager, M. (2009). Cohesins form chromosomal

cis-interactions at the developmentally regulated IFNG locus. Nature 460,

410–413.

Han, M., Toli, J., and Abdellatif, M. (2011). MicroRNAs in the cardiovascular

system. Curr. Opin. Cardiol. 26, 181–189.

Han, Y.J., Ma, S.F., Wade, M.S., Flores, C., and Garcia, J.G. (2012). An intronic

MYLK variant associated with inflammatory lung disease regulates promoter

activity of the smooth muscle myosin light chain kinase isoform. J. Mol.

Med. 90, 299–308.

Hargreaves, D.C., and Crabtree, G.R. (2011). ATP-dependent chromatin

remodeling: genetics, genomics and mechanisms. Cell Res. 21, 396–420.

Harismendy, O., Notani, D., Song, X., Rahim, N.G., Tanasa, B., Heintzman, N.,

Ren, B., Fu, X.D., Topol, E.J., Rosenfeld, M.G., and Frazer, K.A. (2011). 9p21

DNA variants associated with coronary artery disease impair interferon-g sig-

nalling response. Nature 470, 264–268.

Harrow, J., Frankish, A., Gonzalez, J.M., Tapanari, E., Diekhans, M.,

Kokocinski, F., Aken, B.L., Barrell, D., Zadissa, A., Searle, S., et al. (2012).

GENCODE: the reference human genome annotation for The ENCODE

Project. Genome Res. 22, 1760–1774.

Hashimoto, S., Boissel, S., Zarhrate, M., Rio, M., Munnich, A., Egly, J.M., and

Colleaux, L. (2011). MED23 mutation links intellectual disability to dysregula-

tion of immediate early gene expression. Science 333, 1161–1163.

Hayden, M.S., and Ghosh, S. (2012). NF-kB, the first quarter-century: remark-

able progress and outstanding questions. Genes Dev. 26, 203–234.

Hindorff, L.A., Sethupathy, P., Junkins, H.A., Ramos, E.M., Mehta, J.P.,

Collins, F.S., and Manolio, T.A. (2009). Potential etiologic and functional impli-

cations of genome-wide association loci for human diseases and traits. Proc.

Natl. Acad. Sci. USA 106, 9362–9367.

Ho, L., and Crabtree, G.R. (2010). Chromatin remodelling during development.

Nature 463, 474–484.

Horn, S., Figl, A., Rachakonda, P.S., Fischer, C., Sucker, A., Gast, A., Kadel,

S., Moll, I., Nagore, E., Hemminki, K., et al. (2013). TERT promoter mutations

in familial and sporadic melanoma. Science 339, 959–961.

Hoyer, J., Ekici, A.B., Endele, S., Popp, B., Zweier, C., Wiesener, A.,

Wohlleber, E., Dufke, A., Rossier, E., Petsch, C., et al. (2012). Haploinsuffi-

ciency of ARID1B, a member of the SWI/SNF-a chromatin-remodeling

complex, is a frequent cause of intellectual disability. Am. J. Hum. Genet.

90, 565–572.

Huang, B., Yang, X.D., Zhou, M.M., Ozato, K., and Chen, L.F. (2009). Brd4

coactivates transcriptional activation of NF-kappaB via specific binding to

acetylated RelA. Mol. Cell. Biol. 29, 1375–1387.

Huang, P., He, Z., Ji, S., Sun, H., Xiang, D., Liu, C., Hu, Y., Wang, X., and Hui, L.

(2011). Induction of functional hepatocyte-like cells from mouse fibroblasts by

defined factors. Nature 475, 386–389.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1247

Huang, F.W., Hodis, E., Xu, M.J., Kryukov, G.V., Chin, L., and Garraway, L.A.

(2013). Highly recurrent TERT promoter mutations in human melanoma.

Science 339, 957–959.

Ieda, M., Fu, J.D., Delgado-Olguin, P., Vedantham, V., Hayashi, Y., Bruneau,

B.G., and Srivastava, D. (2010). Direct reprogramming of fibroblasts into func-

tional cardiomyocytes by defined factors. Cell 142, 375–386.

Iida, A., Takahashi, A., Kubo, M., Saito, S., Hosono, N., Ohnishi, Y., Kiyotani,

K., Mushiroda, T., Nakajima, M., Ozaki, K., et al. (2011). A functional variant

in ZNF512B is associated with susceptibility to amyotrophic lateral sclerosis

in Japanese. Hum. Mol. Genet. 20, 3684–3692.

Jacob, F., and Monod, J. (1961). Genetic regulatory mechanisms in the

synthesis of proteins. J. Mol. Biol. 3, 318–356.

Jaehning, J.A. (2010). The Paf1 complex: platform or player in RNA poly-

merase II transcription? Biochim. Biophys. Acta 1799, 379–388.

Jiang, Y., Shen, H., Liu, X., Dai, J., Jin, G., Qin, Z., Chen, J.,Wang, S.,Wang, X.,

Hu, Z., and Shen, H. (2011). Genetic variants at 1p11.2 and breast cancer risk:

a two-stage study in Chinese women. PLoS ONE 6, e21563.

Jones, P.A. (2012). Functions of DNA methylation: islands, start sites, gene

bodies and beyond. Nat. Rev. Genet. 13, 484–492.

Juven-Gershon, T., and Kadonaga, J.T. (2010). Regulation of gene expression

via the core promoter and the basal transcriptional machinery. Dev. Biol. 339,

225–229.

Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van

Berkum, N.L., Ebmeier, C.C., Goossens, J., Rahl, P.B., Levine, S.S., et al.

(2010). Mediator and cohesin connect gene expression and chromatin archi-

tecture. Nature 467, 430–435.

Kajimura, S., Seale, P., Kubota, K., Lunsford, E., Frangioni, J.V., Gygi, S.P.,

and Spiegelman, B.M. (2009). Initiation of myoblast to brown fat switch by

a PRDM16-C/EBP-beta transcriptional complex. Nature 460, 1154–1158.

Kapoor, A., Goldberg, M.S., Cumberland, L.K., Ratnakumar, K., Segura, M.F.,

Emanuel, P.O., Menendez, S., Vardabasso, C., Leroy, G., Vidal, C.I., et al.

(2010). The histone variant macroH2A suppresses melanoma progression

through regulation of CDK8. Nature 468, 1105–1109.

Kathiresan, S., and Srivastava, D. (2012). Genetics of human cardiovascular

disease. Cell 148, 1242–1257.

Kaufmann, R., Straussberg, R., Mandel, H., Fattal-Valevski, A., Ben-Zeev, B.,

Naamati, A., Shaag, A., Zenvirt, S., Konen, O., Mimouni-Bloch, A., et al. (2010).

Infantile cerebral and cerebellar atrophy is associated with a mutation in the

MED17 subunit of the transcription preinitiation mediator complex. Am. J.

Hum. Genet. 87, 667–670.

Kawauchi, S., Calof, A.L., Santos, R., Lopez-Burks,M.E., Young, C.M., Hoang,

M.P., Chua, A., Lao, T., Lechner, M.S., Daniel, J.A., et al. (2009). Multiple organ

system defects and transcriptional dysregulation in the Nipbl(+/-) mouse,

a model of Cornelia de Lange Syndrome. PLoS Genet. 5, e1000650.

Kim, T.K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin,

D.A., Laptewicz, M., Barbara-Haley, K., Kuersten, S., et al. (2010). Widespread

transcription at neuronal activity-regulated enhancers. Nature 465, 182–187.

Kornberg, R.D. (2005). Mediator and the mechanism of transcriptional activa-

tion. Trends Biochem. Sci. 30, 235–239.

Kotake, Y., Nakagawa, T., Kitagawa, K., Suzuki, S., Liu, N., Kitagawa, M., and

Xiong, Y. (2011). Long non-coding RNA ANRIL is required for the PRC2 recruit-

ment to and silencing of p15(INK4B) tumor suppressor gene. Oncogene 30,

1956–1962.

Krivega, I., and Dean, A. (2012). Enhancer and promoter interactions-long

distance calls. Curr. Opin. Genet. Dev. 22, 79–85.

Kyewski, B., and Klein, L. (2006). A central role for central tolerance. Annu. Rev.

Immunol. 24, 571–606.

Lariviere, L., Seizl, M., and Cramer, P. (2012). A structural perspective on

Mediator function. Curr. Opin. Cell Biol. 24, 305–313.

Latos, P.A., Pauler, F.M., Koerner, M.V., Sxenergin, H.B., Hudson, Q.J.,

Stocsits, R.R., Allhoff, W., Stricker, S.H., Klement, R.M., Warczok, K.E.,

1248 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

et al. (2012). Airn transcriptional overlap, but not its lncRNA products, induces

imprinted Igf2r silencing. Science 338, 1469–1472.

Leal, A., Huehne, K., Bauer, F., Sticht, H., Berger, P., Suter, U., Morera, B., Del

Valle, G., Lupski, J.R., Ekici, A., et al. (2009). Identification of the variant

Ala335Val of MED25 as responsible for CMT2B2: molecular data, functional

studies of the SH3 recognition motif and correlation between wild-type

MED25 and PMP22 RNA levels in CMT1A animal models. Neurogenetics 10,

275–287.

Lee, J.T. (2012). Epigenetic regulation by long noncoding RNAs. Science 338,

1435–1439.

Lejeune, E., and Allshire, R.C. (2011). Common ground: small RNA program-

ming and chromatin modifications. Curr. Opin. Cell Biol. 23, 258–265.

Lelli, K.M., Slattery, M., and Mann, R.S. (2012). Disentangling the many layers

of eukaryotic transcriptional regulation. Annu. Rev. Genet. 46, 43–68.

Li, X.S., Trojer, P., Matsumura, T., Treisman, J.E., and Tanese, N. (2010).

Mammalian SWI/SNF—a subunit BAF250/ARID1 is an E3 ubiquitin ligase

that targets histone H2B. Mol. Cell. Biol. 30, 1673–1688.

Li, G., Ruan, X., Auerbach, R.K., Sandhu, K.S., Zheng, M., Wang, P., Poh,

H.M., Goh, Y., Lim, J., Zhang, J., et al. (2012). Extensive promoter-centered

chromatin interactions provide a topological basis for transcription regulation.

Cell 148, 84–98.

Lin, C., Smith, E.R., Takahashi, H., Lai, K.C., Martin-Brown, S., Florens, L.,

Washburn, M.P., Conaway, J.W., Conaway, R.C., and Shilatifard, A. (2010).

AFF4, a component of the ELL/P-TEFb elongation complex and a shared

subunit of MLL chimeras, can link transcription elongation to leukemia. Mol.

Cell 37, 429–437.

Lin, C., Garrett, A.S., De Kumar, B., Smith, E.R., Gogol, M., Seidel, C.,

Krumlauf, R., and Shilatifard, A. (2011). Dynamic transcriptional events in

embryonic stem cells mediated by the super elongation complex (SEC). Genes

Dev. 25, 1486–1498.

Lin, C.Y., Loven, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Bradner, J.E., Lee,

T.I., and Young, R.A. (2012). Transcriptional amplification in tumor cells with

elevated c-Myc. Cell 151, 56–67.

Littlewood, T.D., Kreuzaler, P., and Evan, G.I. (2012). All things to all people.

Cell 151, 11–13.

Liu, J., and Krantz, I.D. (2008). Cohesin and human disease. Annu. Rev.

Genomics Hum. Genet. 9, 303–320.

Liu, J., Zhang, Z., Bando, M., Itoh, T., Deardorff, M.A., Clark, D., Kaur, M.,

Tandy, S., Kondoh, T., Rappaport, E., et al. (2009). Transcriptional dysregula-

tion in NIPBL and cohesin mutant human cells. PLoS Biol. 7, e1000119.

Liu, G., Gramling, S., Munoz, D., Cheng, D., Azad, A.K., Mirshams, M., Chen,

Z., Xu, W., Roberts, H., Shepherd, F.A., et al. (2011). Two novel BRM insertion

promoter sequence variants are associated with loss of BRM expression and

lung cancer risk. Oncogene 30, 3295–3304.

Lubbe, S.J., Pittman, A.M., Olver, B., Lloyd, A., Vijayakrishnan, J., Naranjo, S.,

Dobbins, S., Broderick, P., Gomez-Skarmeta, J.L., and Houlston, R.S. (2012).

The 14q22.2 colorectal cancer variant rs4444235 shows cis-acting regulation

of BMP4. Oncogene 31, 3777–3784.

Luo, Z., Lin, C., Guest, E., Garrett, A.S., Mohaghegh, N., Swanson, S.,

Marshall, S., Florens, L., Washburn, M.P., and Shilatifard, A. (2012a). The

super elongation complex family of RNA polymerase II elongation factors:

gene target specificity and transcriptional output. Mol. Cell. Biol. 32, 2608–

2617.

Luo, Z., Lin, C., and Shilatifard, A. (2012b). The super elongation complex

(SEC) family in transcriptional control. Nat. Rev. Mol. Cell Biol. 13, 543–547.

Maestro, M.A., Cardalda, C., Boj, S.F., Luco, R.F., Servitja, J.M., and Ferrer, J.

(2007). Distinct roles of HNF1beta, HNF1alpha, and HNF4alpha in regulating

pancreas development, beta-cell function and growth. Endocr. Dev. 12, 33–45.

Makinen, N., Heinonen, H.R., Moore, S., Tomlinson, I.P., van der Spuy, Z.M.,

and Aaltonen, L.A. (2011a). MED12 exon 2 mutations are common in uterine

leiomyomas from South African patients. Oncotarget 2, 966–969.

Makinen, N., Mehine, M., Tolvanen, J., Kaasinen, E., Li, Y., Lehtonen, H.J.,

Gentile, M., Yan, J., Enge, M., Taipale, M., et al. (2011b). MED12, the mediator

complex subunit 12 gene, is mutated at high frequency in uterine leiomyomas.

Science 334, 252–255.

Malecki, M.T. (2005). Genetics of type 2 diabetes mellitus. Diabetes Res. Clin.

Pract. 68(Suppl 1), S10–S21.

Malik, S., and Roeder, R.G. (2010). The metazoan Mediator co-activator

complex as an integrative hub for transcriptional regulation. Nat. Rev. Genet.

11, 761–772.

Mannini, L., and Musio, A. (2011). The dark side of cohesin: the carcinogenic

point of view. Mutat. Res. 728, 81–87.

Margueron, R., and Reinberg, D. (2011). The Polycomb complex PRC2 and its

mark in life. Nature 469, 343–349.

Marro, S., Pang, Z.P., Yang, N., Tsai, M.C., Qu, K., Chang, H.Y., Sudhof, T.C.,

and Wernig, M. (2011). Direct lineage conversion of terminally differentiated

hepatocytes to functional neurons. Cell Stem Cell 9, 374–382.

Marschalek, R. (2010). Mixed lineage leukemia: roles in human malignancies

and potential therapy. FEBS J. 277, 1822–1831.

Mathis, D., and Benoist, C. (2009). Aire. Annu. Rev. Immunol. 27, 287–312.

Maurano, M.T., Humbert, R., Rynes, E., Thurman, R.E., Haugen, E., Wang, H.,

Reynolds, A.P., Sandstrom, R., Qu, H., Brody, J., et al. (2012). Systematic

localization of common disease-associated variation in regulatory DNA.

Science 337, 1190–1195.

Metzger, T.C., and Anderson, M.S. (2011). Control of central and peripheral

tolerance by Aire. Immunol. Rev. 241, 89–103.

Mills, A.A. (2010). Throwing the cancer switch: reciprocal roles of polycomb

and trithorax proteins. Nat. Rev. Cancer 10, 669–682.

Moazed, D. (2009). Small RNAs in transcriptional gene silencing and genome

defence. Nature 457, 413–420.

Morris, E.J., Ji, J.Y., Yang, F., Di Stefano, L., Herr, A., Moon, N.S., Kwon, E.J.,

Haigis, K.M., Naar, A.M., and Dyson, N.J. (2008). E2F1 represses beta-catenin

transcription and is antagonized by both pRB andCDK8. Nature 455, 552–556.

Nakajima, T., Inui, S., Fushimi, T., Noguchi, F., Kitagawa, Y., Reddy, J.K., and

Itami, S. (2013). Roles of MED1 in quiescence of hair follicle stem cells and

maintenance of normal hair cycling. J. Invest. Dermatol. 133, 354–360.

Ng, H.H., and Surani, M.A. (2011). The transcriptional and signalling networks

of pluripotency. Nat. Cell Biol. 13, 490–496.

Nguyen, D., Krueger, B.J., Sedore, S.C., Brogie, J.E., Rogers, J.T., Rajendra,

T.K., Saunders, A., Matera, A.G., Lis, J.T., Uguen, P., and Price, D.H. (2012).

The Drosophila 7SK snRNP and the essential role of dHEXIM in development.

Nucleic Acids Res. 40, 5283–5297.

Nicodeme, E., Jeffrey, K.L., Schaefer, U., Beinke, S., Dewell, S., Chung, C.W.,

Chandwani, R., Marazzi, I., Wilson, P., Coste, H., et al. (2010). Suppression of

inflammation by a synthetic histone mimic. Nature 468, 1119–1123.

Nie, Z., Hu, G., Wei, G., Cui, K., Yamane, A., Resch, W., Wang, R., Green, D.R.,

Tessarollo, L., Casellas, R., et al. (2012). c-Myc is a universal amplifier of

expressed genes in lymphocytes and embryonic stem cells. Cell 151, 68–79.

Nowak, D.E., Tian, B., Jamaluddin, M., Boldogh, I., Vergara, L.A., Choudhary,

S., and Brasier, A.R. (2008). RelA Ser276 phosphorylation is required for

activation of a subset of NF-kappaB-dependent genes by recruiting cyclin-

dependent kinase 9/cyclin T1 complexes. Mol. Cell. Biol. 28, 3623–3638.

Oda, Y., Hu, L., Bul, V., Elalieh, H., Reddy, J.K., and Bikle, D.D. (2012). Coac-

tivator MED1 ablation in keratinocytes results in hair-cycling defects and

epidermal alterations. J. Invest. Dermatol. 132, 1075–1083.

Odom, D.T., Zizlsperger, N., Gordon, D.B., Bell, G.W., Rinaldi, N.J., Murray,

H.L., Volkert, T.L., Schreiber, J., Rolfe, P.A., Gifford, D.K., et al. (2004). Control

of pancreas and liver gene expression by HNF transcription factors. Science

303, 1378–1381.

Ong, C.T., and Corces, V.G. (2011). Enhancer function: new insights into the

regulation of tissue-specific gene expression. Nat. Rev. Genet. 12, 283–293.

Orkin, S.H., and Hochedlinger, K. (2011). Chromatin connections to pluripo-

tency and cellular reprogramming. Cell 145, 835–850.

Orom, U.A., and Shiekhattar, R. (2011). Noncoding RNAs and enhancers:

complications of a long-distance relationship. Trends Genet. 27, 433–439.

Oven, I., Brdickova, N., Kohoutek, J., Vaupotic, T., Narat, M., and Peterlin,

B.M. (2007). AIRE recruits P-TEFb for transcriptional elongation of target

genes in medullary thymic epithelial cells. Mol. Cell. Biol. 27, 8815–8823.

Pang, Z.P., Yang, N., Vierbuchen, T., Ostermeier, A., Fuentes, D.R., Yang,

T.Q., Citri, A., Sebastiano, V., Marro, S., Sudhof, T.C., and Wernig, M.

(2011). Induction of human neuronal cells by defined transcription factors.

Nature 476, 220–223.

Panne, D. (2008). The enhanceosome. Curr. Opin. Struct. Biol. 18, 236–242.

Papageorgiou, N., Tousoulis, D., Androulakis, E., Siasos, G., Briasoulis, A.,

Vogiatzi, G., Kampoli, A.M., Tsiamis, E., Tentolouris, C., and Stefanadis, C.

(2012). The role of microRNAs in cardiovascular disease. Curr. Med. Chem.

19, 2605–2610.

Parelho, V., Hadjur, S., Spivakov, M., Leleu, M., Sauer, S., Gregson, H.C.,

Jarmuz, A., Canzonetta, C., Webster, Z., Nesterova, T., et al. (2008). Cohesins

functionally associate with CTCF on mammalian chromosome arms. Cell 132,

422–433.

Park, K.S., Cha, Y., Kim, C.H., Ahn, H.J., Kim, D., Ko, S., Kim, K.H., Chang,

M.Y., Ko, J.H., Noh, Y.S., et al. (2013). Transcription elongation factor tcea3

regulates the pluripotent differentiation potential of mouse embryonic stem

cells via the lefty1-nodal-smad2 pathway. Stem Cells 31, 282–292.

Pasmant, E., Sabbagh, A., Vidaud, M., and Bieche, I. (2011). ANRIL, a long,

noncoding RNA, is an unexpected major hotspot in GWAS. FASEB J. 25,

444–448.

Philibert, R.A., Bohle, P., Secrest, D., Deaderick, J., Sandhu, H., Crowe, R.,

and Black, D.W. (2007). The association of the HOPA(12bp) polymorphism

with schizophrenia in the NIMH Genetics Initiative for Schizophrenia sample.

Am. J. Med. Genet. B. Neuropsychiatr. Genet. 144B, 743–747.

Phillips, J.E., and Corces, V.G. (2009). CTCF: master weaver of the genome.

Cell 137, 1194–1211.

Popov, N., and Gil, J. (2010). Epigenetic regulation of the INK4b-ARF-INK4a

locus: in sickness and in health. Epigenetics 5, 685–690.

Rahl, P.B., Lin, C.Y., Seila, A.C., Flynn, R.A., McCuine, S., Burge, C.B., Sharp,

P.A., and Young, R.A. (2010). c-Myc regulates transcriptional pause release.

Cell 141, 432–445.

Rando, O.J. (2012). Combinatorial complexity in chromatin structure and func-

tion: revisiting the histone code. Curr. Opin. Genet. Dev. 22, 148–155.

Reyes-Turcu, F.E., and Grewal, S.I. (2012). Different means, same end-hetero-

chromatin formation by RNAi and RNAi-independent RNA processing factors

in fission yeast. Curr. Opin. Genet. Dev. 22, 156–163.

Rinn, J.L., and Chang, H.Y. (2012). Genome regulation by long noncoding

RNAs. Annu. Rev. Biochem. 81, 145–166.

Risheg, H., Graham, J.M., Jr., Clark, R.D., Rogers, R.C., Opitz, J.M.,

Moeschler, J.B., Peiffer, A.P., May, M., Joseph, S.M., Jones, J.R., et al.

(2007). A recurrent mutation in MED12 leading to R961W causes Opitz-

Kaveggia syndrome. Nat. Genet. 39, 451–453.

Roeder, R.G. (2005). Transcriptional regulation and the role of diverse coacti-

vators in animal cells. FEBS Lett. 579, 909–915.

Roger, V.L., Go, A.S., Lloyd-Jones, D.M., Benjamin, E.J., Berry, J.D., Borden,

W.B., Bravata, D.M., Dai, S., Ford, E.S., Fox, C.S., et al.; American Heart Asso-

ciation Statistics Committee and Stroke Statistics Subcommittee. (2012).

Executive summary: heart disease and stroke statistics—2012 update: a report

from the American Heart Association. Circulation 125, 188–197.

Rump, P., Niessen, R.C., Verbruggen, K.T., Brouwer, O.F., de Raad, M., and

Hordijk, R. (2011). A novel mutation in MED12 causes FG syndrome (Opitz-

Kaveggia syndrome). Clin. Genet. 79, 183–188.

Sanda, T., Lawton, L.N., Barrasa, M.I., Fan, Z.P., Kohlhammer, H., Gutierrez,

A., Ma,W., Tatarek, J., Ahn, Y., Kelliher, M.A., et al. (2012). Core transcriptional

regulatory circuit controlled by the TAL1 complex in human T cell acute

lymphoblastic leukemia. Cancer Cell 22, 209–221.

Santen, G.W., Aten, E., Sun, Y., Almomani, R., Gilissen, C., Nielsen, M., Kant,

S.G., Snoeck, I.N., Peeters, E.A., Hilhorst-Hofstee, Y., et al. (2012a). Mutations

in SWI/SNF chromatin remodeling complex gene ARID1B cause Coffin-Siris

syndrome. Nat. Genet. 44, 379–380.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1249

Santen, G.W., Kriek, M., and van Attikum, H. (2012b). SWI/SNF complex in

disorder: SWItching from malignancies to intellectual disability. Epigenetics

7, 1219–1224.

Schaaf, C.A., Misulovin, Z., Sahota, G., Siddiqui, A.M., Schwartz, Y.B., Kahn,

T.G., Pirrotta, V., Gause, M., and Dorsett, D. (2009). Regulation of the

Drosophila Enhancer of split and invected-engrailed gene complexes by sister

chromatid cohesion proteins. PLoS ONE 4, e6202.

Schmidt, D., Schwalie, P.C., Ross-Innes, C.S., Hurtado, A., Brown, G.D.,

Carroll, J.S., Flicek, P., and Odom, D.T. (2010). A CTCF-independent role for

cohesin in tissue-specific transcription. Genome Res. 20, 578–588.

Schodel, J., Bardella, C., Sciesielski, L.K., Brown, J.M., Pugh, C.W., Buckle,

V., Tomlinson, I.P., Ratcliffe, P.J., and Mole, D.R. (2012). Common genetic

variants at the 11q13.3 renal cancer susceptibility locus influence binding

of HIF to an enhancer of cyclin D1 expression. Nat. Genet. 44, 420–425,

S421–422.

Schwartz, C.E., Tarpey, P.S., Lubs, H.A., Verloes, A., May, M.M., Risheg, H.,

Friez, M.J., Futreal, P.A., Edkins, S., Teague, J., et al. (2007). The original Lujan

syndrome family has a novel missense mutation (p.N1007S) in the MED12

gene. J. Med. Genet. 44, 472–477.

Seal, R.L., Gordon, S.M., Lush, M.J., Wright, M.W., and Bruford, E.A.

(2011). genenames.org: the HGNC resources in 2011. Nucleic Acids Res.

39(Database issue), D514–D519.

Seitan, V.C., andMerkenschlager, M. (2012). Cohesin and chromatin organisa-

tion. Curr. Opin. Genet. Dev. 22, 93–100.

Seitan, V.C., Hao, B., Tachibana-Konwalski, K., Lavagnolli, T., Mira-Bonten-

bal, H., Brown, K.E., Teng, G., Carroll, T., Terry, A., Horan, K., et al. (2011). A

role for cohesin in T-cell-receptor rearrangement and thymocyte differentia-

tion. Nature 476, 467–471.

Sekiya, S., and Suzuki, A. (2011). Direct conversion of mouse fibroblasts to

hepatocyte-like cells by defined factors. Nature 475, 390–393.

Sigova, A.A., Mullen, A.C., Moline, B., Gupta, S., Orlando, D.A., Guenther,

M.G., Almada, A.A., Lin, C., Burge, C.B., Sharp, P.A., et al. (2013). Divergent

transcription of long noncoding RNA/mRNA gene pairs in embryonic stem

cells. Proc. Natl. Acad. Sci. USA. 110, 2876–2881.

Sikorski, T.W., and Buratowski, S. (2009). The basal initiation machinery:

beyond the general transcription factors. Curr. Opin. Cell Biol. 21, 344–351.

Slany, R.K. (2009). The molecular biology of mixed lineage leukemia.

Haematologica 94, 984–993.

Small, E.M., and Olson, E.N. (2011). Pervasive roles of microRNAs in cardio-

vascular biology. Nature 469, 336–342.

Smith, E., Lin, C., and Shilatifard, A. (2011a). The super elongation complex

(SEC) and MLL in development and disease. Genes Dev. 25, 661–672.

Smith, E.R., Lin, C., Garrett, A.S., Thornton, J., Mohaghegh, N., Hu, D.,

Jackson, J., Saraf, A., Swanson, S.K., Seidel, C., et al. (2011b). The little

elongation complex regulates small nuclear RNA transcription. Mol. Cell 44,

954–965.

Spaeth, J.M., Kim, N.H., and Boyer, T.G. (2011). Mediator and human disease.

Semin. Cell Dev. Biol. 22, 776–787.

Spitz, F., and Furlong, E.E. (2012). Transcription factors: from enhancer

binding to developmental control. Nat. Rev. Genet. 13, 613–626.

Taatjes, D.J. (2010). The human Mediator complex: a versatile, genome-wide

regulator of transcription. Trends Biochem. Sci. 35, 315–322.

Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells

from mouse embryonic and adult fibroblast cultures by defined factors. Cell

126, 663–676.

Thum, T., Gross, C., Fiedler, J., Fischer, T., Kissler, S., Bussen, M., Galuppo,

P., Just, S., Rottbauer, W., Frantz, S., et al. (2008). MicroRNA-21 contributes

to myocardial disease by stimulating MAP kinase signalling in fibroblasts.

Nature 456, 980–984.

Thurman, R.E., Rynes, E., Humbert, R., Vierstra, J., Maurano, M.T.,

Haugen, E., Sheffield, N.C., Stergachis, A.B., Wang, H., Vernot, B., et al.

1250 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

(2012). The accessible chromatin landscape of the human genome. Nature

489, 75–82.

Tsai, H.C., and Baylin, S.B. (2011). Cancer epigenetics: linking basic biology to

clinical medicine. Cell Res. 21, 502–517.

Tsurusaki, Y., Okamoto, N., Ohashi, H., Kosho, T., Imai, Y., Hibi-Ko, Y.,

Kaname, T., Naritomi, K., Kawame, H., Wakui, K., et al. (2012). Mutations

affecting components of the SWI/SNF complex cause Coffin-Siris syndrome.

Nat. Genet. 44, 376–378.

Van Houdt, J.K., Nowakowska, B.A., Sousa, S.B., van Schaik, B.D., Seuntjens,

E., Avonce, N., Sifrim, A., Abdul-Rahman, O.A., van den Boogaard, M.J.,

Bottani, A., et al. (2012). Heterozygous missense mutations in SMARCA2

cause Nicolaides-Baraitser syndrome. Nat. Genet. 44, 445–449, S441.

van Rooij, E., Sutherland, L.B., Qi, X., Richardson, J.A., Hill, J., and Olson, E.N.

(2007). Control of stress-dependent cardiac growth and gene expression by

a microRNA. Science 316, 575–579.

van Rooij, E., Sutherland, L.B., Thatcher, J.E., DiMaio, J.M., Naseem, R.H.,

Marshall, W.S., Hill, J.A., and Olson, E.N. (2008). Dysregulation of microRNAs

after myocardial infarction reveals a role of miR-29 in cardiac fibrosis. Proc.

Natl. Acad. Sci. USA 105, 13027–13032.

Vierbuchen, T., Ostermeier, A., Pang, Z.P., Kokubu, Y., Sudhof, T.C., and

Wernig, M. (2010). Direct conversion of fibroblasts to functional neurons by

defined factors. Nature 463, 1035–1041.

Wang, D., Garcia-Bassets, I., Benner, C., Li, W., Su, X., Zhou, Y., Qiu, J., Liu,

W., Kaikkonen, M.U., Ohgi, K.A., et al. (2011). Reprogramming transcription

by distinct classes of enhancers functionally defined by eRNA. Nature 474,

390–394.

Wendt, K.S., Yoshida, K., Itoh, T., Bando, M., Koch, B., Schirghuber, E.,

Tsutsumi, S., Nagae, G., Ishihara, K., Mishiro, T., et al. (2008). Cohesin

mediates transcriptional insulation by CCCTC-binding factor. Nature 451,

796–801.

Wilson, B.G., and Roberts, C.W. (2011). SWI/SNF nucleosome remodellers

and cancer. Nat. Rev. Cancer 11, 481–492.

Wright, J.E., and Ciosk, R. (2013). RNA-based regulation of pluripotency.

Trends Genet. 29, 99–107.

Xie, H., Ye, M., Feng, R., and Graf, T. (2004). Stepwise reprogramming of B

cells into macrophages. Cell 117, 663–676.

Xu, H., Tomaszewski, J.M., and McKay, M.J. (2011). Can corruption of chro-

mosome cohesion create a conduit to cancer? Nat. Rev. Cancer 11, 199–210.

Yamanaka, S. (2012). Induced pluripotent stem cells: past, present, and future.

Cell Stem Cell 10, 678–684.

Yang, B., Lin, H., Xiao, J., Lu, Y., Luo, X., Li, B., Zhang, Y., Xu, C., Bai, Y.,Wang,

H., et al. (2007). The muscle-specific microRNA miR-1 regulates cardiac

arrhythmogenic potential by targeting GJA1 and KCNJ2. Nat. Med. 13,

486–491.

Yankulov, K., Blau, J., Purton, T., Roberts, S., and Bentley, D.L. (1994). Tran-

scriptional elongation by RNA polymerase II is stimulated by transactivators.

Cell 77, 749–759.

Yap, K.L., Li, S., Munoz-Cabello, A.M., Raguz, S., Zeng, L., Mujtaba, S., Gil, J.,

Walsh, M.J., and Zhou,M.M. (2010). Molecular interplay of the noncoding RNA

ANRIL and methylated histone H3 lysine 27 by polycomb CBX7 in transcrip-

tional silencing of INK4a. Mol. Cell 38, 662–674.

Yeo, J.C., and Ng, H.H. (2013). The transcriptional regulation of pluripotency.

Cell Res. 23, 20–32.

Yew, P.Y., Mushiroda, T., Kiyotani, K., Govindasamy, G.K., Yap, L.F., Teo,

S.H., Lim, P.V., Govindaraju, S., Ratnavelu, K., Sam, C.K., et al.; Malaysian

NPC Study Group. (2012). Identification of a functional variant in SPLUNC1

associated with nasopharyngeal carcinoma susceptibility among Malaysian

Chinese. Mol. Carcinog. 51(Suppl 1), E74–E82.

Young, R.A. (2011). Control of the embryonic stemcell state.Cell 144, 940–954.

Zhao, J.Y., Yang, X.Y., Gong, X.H., Gu, Z.Y., Duan, W.Y., Wang, J., Ye, Z.Z.,

Shen, H.B., Shi, K.H., Hou, J., et al. (2012). Functional variant in methionine

synthase reductase intron-1 significantly increases the risk of congenital heart

disease in the Han Chinese population. Circulation 125, 482–490.

Zhou, Q., Brown, J., Kanarek, A., Rajagopal, J., andMelton, D.A. (2008). In vivo

reprogramming of adult pancreatic exocrine cells to beta-cells. Nature 455,

627–632.

Zhou, H., Spaeth, J.M., Kim, N.H., Xu, X., Friez, M.J., Schwartz, C.E., and

Boyer, T.G. (2012a). MED12 mutations link intellectual disability syndromes

with dysregulated GLI3-dependent Sonic Hedgehog signaling. Proc. Natl.

Acad. Sci. USA 109, 19763–19768.

Zhou, Q., Li, T., and Price, D.H. (2012b). RNA polymerase II elongation control.

Annu. Rev. Biochem. 81, 119–143.

Zhu, J., Adli, M., Zou, J.Y., Verstappen, G., Coyne, M., Zhang, X., Durham, T.,

Miri, M., Deshpande, V., De Jager, P.L., et al. (2013). Genome-wide chromatin

state transitions associated with developmental and environmental cues. Cell

152, 642–654.

�Zumer, K., Plemenita�s, A., Saksela, K., and Peterlin, B.M. (2011). Patient muta-

tion in AIRE disrupts P-TEFb binding and target gene transcription. Nucleic

Acids Res. 39, 7908–7919.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1251

Leading Edge

Review

Dynamic Integration of Splicingwithin Gene Regulatory Pathways

Ulrich Braunschweig,1,4 Serge Gueroussov,1,2,4 Alex M. Plocik,3,4 Brenton R. Graveley,3,* and Benjamin J. Blencowe1,2,*1Banting and Best Department of Medical Research, Donnelly Centre2Department of Molecular GeneticsUniversity of Toronto, Toronto, ON M5S 1A8, Canada3Department of Genetics and Developmental Biology, Institute for Systems Genomics, University of Connecticut Health Center,

400 Farmington Avenue, Farmington, CT 06030-6403, USA4These authors contributed equally to this work*Correspondence: graveley@neuron.uchc.edu (B.R.G.), b.blencowe@utoronto.ca (B.J.B.)

http://dx.doi.org/10.1016/j.cell.2013.02.034

Precursor mRNA splicing is one of the most highly regulated processes in metazoan species. Inaddition to generating vast repertoires of RNAs and proteins, splicing has a profound impact onother gene regulatory layers, including mRNA transcription, turnover, transport, and translation.Conversely, factors regulating chromatin and transcription complexes impact the splicing process.This extensive crosstalk between gene regulatory layers takes advantage of dynamic spatial,physical, and temporal organizational properties of the cell nucleus, and further emphasizes theimportance of developing a multidimensional understanding of splicing control.

IntroductionThe splicing of messenger RNA precursors (pre-mRNA) to

mature mRNAs is a highly dynamic and flexible process that

impacts almost every aspect of eukaryotic cell biology. The

formation of active splicing complexes—or ‘‘spliceosomes’’—

occurs via step-wise assembly pathways on pre-mRNAs. Small

nuclear ribonucleoprotein particles (snRNPs): U1, U2, U4/U6,

and U5, in the case of the major spliceosome, and U11, U12,

U4atac/U6atac, and U5, in the case of the minor spliceosome,

together with an additional �150 proteins, associate with pre-

mRNAs, initially through direct recognition of short sequences

at the exon/intron boundaries. Key features of spliceosome

formation are shown in Figure 1 and have been reviewed in detail

elsewhere (Hoskins and Moore, 2012; Wahl et al., 2009).

Spliceosome assembly can be regulated in extraordinarily

diverse ways, particularly in metazoans. The major steps involve

formation of the commitment complex followed by the pre-

splicing complex and culminating with assembly of the active

spliceosome. These steps appear to be reversible and potential

points of regulation (Hoskins et al., 2011), and accumulating

evidence indicates that formation of the commitment and pre-

splicing complexes may be the most often subject to control

(Chen and Manley, 2009).

Analysis of human genome architecture emphasizes a major

challenge for accurate recognition and regulation of splice sites

by the splicing machinery, namely that exons represent only 3%

of the human genome (ENCODE Project Consortium, 2012).

Accumulating evidence indicates that the high-fidelity process

of splice site selection is not simply governed by the interaction

of snRNPs and non-snRNP protein factors with pre-mRNA but

that factors associated with chromatin and the transcriptional

machinery are also important (Luco et al., 2011). Moreover,

splicing can ‘‘reach back’’ to impact chromatin composition

1252 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

and transcriptional activity, as well as influence parallel or

downstream steps in gene expression including 30-end pro-

cessing, mRNA turnover, and translation (de Almeida and

Carmo-Fonseca, 2012; Moore and Proudfoot, 2009). Therefore,

understanding fundamental biological processes such as cell

differentiation and development, as well as disease mecha-

nisms, will require knowledge of the crosstalk between splicing

and other regulatory layers in cells. A major facet of developing

such knowledge is to understand how splicing is physically,

spatially, and temporally integrated with other gene expression

processes in the cell nucleus. This review focuses on these

topics, with an emphasis on knowledge that has been gained

from the application of genome-wide strategies, together with

focusedmolecular, biochemical, and cell biological approaches.

Regulation of Splicing at the Level of RNARegulatory RNA Sequences

Alternative splicing (AS) is the process by which different pairs of

splice sites are selected in a pre-mRNA transcript to produce

distinct mRNA and protein isoforms. The importance of under-

standing AS regulation is underscored by its widespread nature

and its numerous defined roles in critical biological processes

including cell growth, cell death, pluripotency, cell differentia-

tion, development, circadian rhythms, responses to environ-

mental challenge, pathogen exposure, and disease (Irimia and

Blencowe, 2012; Kalsotra and Cooper, 2011). Analysis of data

from high-throughput RNA sequencing (RNA-Seq) of organ tran-

scriptomes has indicated that at least 95% of human multi-exon

genes produce alternatively spliced transcripts (Pan et al., 2008;

Wang et al., 2008) and that the frequency of AS scales with cell

type and species complexity (Barbosa-Morais et al., 2012;

Nilsen and Graveley, 2010). The main types of AS found in

eukaryotes are ‘‘cassette’’ exon skipping, alternative 50 and 30

Figure 1. Cotranscriptional and Posttranscriptional Aspects of Pre-mRNA SplicingCotranscriptional spliceosome assembly initiates with the binding of U1 snRNP to the 50 splice site, which is enhanced by exon-bound SR proteins and, for thefirst exon, the cap binding complex (CBC). A cross-intron commitment complex is formed upon association of U2 snRNP auxiliary factor (U2AF) with the 30 splicesite and adjacent intronic polypyrimidine tract, and branch point binding protein (BBP/SF1) with the branch site. Bridging interactions between these factorsacross internal exons, or ‘‘exon definition,’’ occurs within the commitment complex. Transition from a commitment complex to a presplicing complex entailscommunication between 50 and 30 splice sites, and the addition of U2 snRNP to the branch site along with numerous additional proteins (not shown). Subsequentassociation of U4/U6/U5 tri-snRNP, together with still more protein factors, and dynamic remodeling of RNA-protein, protein-protein, and RNA-RNA interactions,ultimately leads to formation of the catalytically active spliceosome. The two trans-esterification steps of splicing yield the excised intron in the form of thecharacteristic branched ‘‘lariat’’ structure and the ligated exons that formmaturemRNA. The assembly of most splicing factors and splicing of constitutive intronsis thought to occur cotranscriptionally, whereas splicing of regulated alternative introns often occurs posttranscriptionally. In the example shown, exon 4 isa regulated alternative exon controlled by an hnRNP protein, which prevents the splicing factors bound to flanking splice sites from engaging in productiveinteractions and therefore promotes exon skipping. At terminal exons (exon 5), interactions between the splicing factors bound to the upstream 30 splice site andthe exon interact with components of the cleavage and polyadenlyationmachinery (CPSF andCstF are shown; see also Figure 4A). The association of the splicingfactors with the pre-mRNA is enhanced throughout the transcription process by interactions with the C-terminal domain of RNA polymerase II. The EJC isrecruited upstream of splice junctions upon splicing. The EJC and SR proteins mutually stabilize one another to generate the mature mRNP, which is thenexported to the cytoplasm.

splice site selection, alternative retained introns, and mutually

exclusive exons. The vast majority of AS events have not been

functionally characterized on any level, and this represents

a major challenge for biological research. However, large-scale

studies of splice variants employing a mix of computational

and experimental approaches have provided evidence for

widespread roles of regulated alternative exons in the control

of protein interaction networks, and in cell signaling (Buljan

et al., 2012; Ellis et al., 2012; Weatheritt and Gibson, 2012).

The selection of correct pairs of 50 and 30 splice sites in

pre-mRNA is governed in part by cis-acting RNA sequences

that collectively comprise the ‘‘splicing code’’ (Wang and Burge,

2008). The code utilizes a surprisingly minimal set of highly

conserved features; these are the intronic dinucleotides GU

and AG (with variations used by the minor spliceosome) at the

50 and 30 splice sites, respectively, and the intronic adenosine

residue that forms the branched lariat structure. Additional

nucleotides surrounding these positions display sequence

preferences that reflect requirements for base-pairing interac-

tions with the snRNA components of snRNPs during spliceo-

some formation (Wahl et al., 2009). Although these minimal

core elements delineate sites of splicing, they lack sufficient

information to discriminate correct from incorrect splice sites

and to regulate AS.

Combinations of additional sequence elements referred to

as exonic/intronic splicing enhancers (E/ISEs) and silencers

(E/ISSs) serve to promote and repress splice site selection.

They operate in the context of achieving fidelity and in the regu-

lation of this process (Wang andBurge, 2008). Themajority of the

code elements comprise short and degenerate linear motifs,

although interesting examples of structured RNA elements

have been discovered that function in splice site selection

(Graveley, 2005; McManus and Graveley, 2011). The major

contribution of linear motifs to splicing regulation is reflected

by the ability of increasingly sophisticated computer algorithms

to predict splicing outcomes from genomic sequence alone

(Barash et al., 2010; Zhang et al., 2010). The emerging picture,

supported by site-directed mutagenesis of cis elements, is that

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1253

splice site selection involves the concerted action of multiple

enhancer and silencer elements that are concentrated in regions

proximal (typically within �300 nts) to splice sites (Barash et al.,

2010). In particular, enhancers that support constitutive exon

splicing are typically concentrated in exons, whereas enhancers

and silencers that function in the regulation of AS can be located

in alternative exons, although they are most often concentrated

in the immediate flanking intronic regions (Barash et al., 2010).

Additionally, silencer elements are enriched in sequences

surrounding cryptic splice sites—sequences that resemble

splice sites but are not functional splice sites (Wang and Burge,

2008).

Regulatory Proteins

Two major classes of widely expressed trans-acting factors that

control splice site recognition are the SR proteins and heteroge-

neous ribonucleoproteins (hnRNPs) (Long and Caceres, 2009;

Martinez-Contreras et al., 2007). Depending on their binding

location and the surrounding sequence context, members

of each class can promote or repress splice site selection

through associating with enhancers or silencers, respectively.

For example, members of the SR family of proteins contain

one or two RNA recognition motifs that bind ESEs and are

thought to promote splicing by facilitating exon-spanning inter-

actions that occur between splice sites (referred to as ‘‘exon

definition’’) and also by forging interactions with core spliceo-

somal proteins (Figure 1). In addition to widely expressed

trans-acting factors, several tissue-specific RNA-binding

splicing regulators have been characterized (Irimia and Blen-

cowe, 2012; Licatalosi and Darnell, 2010). These include

the neural-specific factors Nova, PTBP2/nPTB/brPTB, and

nSR100/SRRM4, and factors such as RBFOX, MBNL, CELF,

TIA, and STAR family proteins that are differentially expressed

between a variety of cell and tissue types. Through the use of

splicing-sensitive microarrays and RNA-Seq to detect exons

affected by the knockout or knockdown of these factors, in

combination with splicing code predictions and in vivo cross-

linking coupled to immunoprecipitation and sequencing (HITS-

CLIP or CLIP-Seq), ‘‘maps’’ of several of these proteins have

been generated that correlate their binding location (i.e., within

alternative exons and/or the flanking introns) with functions in

promoting exon inclusion or skipping (Licatalosi and Darnell,

2010; Witten and Ule, 2011). As mentioned earlier, where

studied, these proteins appear to act primarily at the earliest

stages of spliceosome formation to control splice site selection.

Integration of Splicingwith Chromatin and Transcription

Despite major progress in the characterization of factors that

control splicing at the level of RNA, the impact of linked steps in

gene regulation and of nuclear organization on the splicing

process is less well understood. The fact that synthetic

pre-mRNAs can be efficiently spliced in nuclear extracts

demonstrates that splicing can be uncoupled from other nuclear

processes in vitro. However, mounting evidence indicates that

splicing, transcription, and chromatin modification are highly

integrated in the cell. Thus, key to understanding the role of chro-

matin and transcription in the control of splicing is knowingwhich

aspectsof the splicingprocessoccurco- or posttranscriptionally.

Some of the first mechanistic insights into the cotranscrip-

tional nature of splicing came from chromatin immunoprecipita-

1254 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

tion studies in yeast. These experiments revealed that splicing

factors fail to associate with intronless genes but are recruited

to intron-containing genes concomitant with the transcription

of the splice sites they recognize (Gornemann et al., 2005;

Lacadie and Rosbash, 2005). The main exceptions were genes

containing short last exons, in which case U1 snRNP was

recruited cotranscriptionally, but U2 snRNP was recruited post-

transcriptionally (Tardiff et al., 2006). Similar approaches have

been used in human cells with similar results (Listerman et al.,

2006). These data paint a general picture in which the splicing

machinery is typically recruited to pre-mRNA in a cotranscrip-

tional manner.

Although splicing factors are cotranscriptionally recruited, it

does not necessarily follow that the splicing reaction itself

occurs cotranscriptionally. Recently, Vargas et al. used in situ

hybridization methods with single-molecule resolution and

found that constitutively spliced introns, which typically are

efficiently spliced, were removed cotranscriptionally (Vargas

et al., 2011). However, mutations that decreased the splicing

efficiency, for instance by sequestering splicing signals in

RNA secondary structures, caused introns to be posttranscrip-

tionally spliced. More interestingly, two alternatively spliced

introns examined were found to be posttranscriptionally

spliced. This study suggested that introns could be either co-

transcriptionally or posttranscriptionally spliced, in part de-

pending on the strength and type of surrounding cis-regulatory

elements.

The extent to which specific classes of splicing events occur

co- or posttranscriptionally has since been examined on a

genome-wide level. Several groups have analyzed RNA-Seq

data generated from total cellular RNA, total nuclear RNA, nucle-

oplasmic RNA, or chromatin-associated RNA (Ameur et al.,

2011; Bhatt et al., 2012; Khodor et al., 2012; Khodor et al.,

2011; Tilgner et al., 2012). Each group used a different method

to assess the extent of cotranscriptional splicing. Though the

precise frequency differed in each study, most introns appeared

to be cotranscriptionally spliced. The likelihood of cotranscrip-

tional splicing increases with increased distance of introns

from the 30 ends of genes (Khodor et al., 2012). Strikingly, the

set of posttranscriptionally spliced introns is strongly enriched

for alternatively spliced introns. Moreover, it was observed that

most human transcripts are cleaved and polyadenylated before

splicing of all introns is complete, yet these transcripts remain

associated with the chromatin until splicing is finished (Bhatt

et al., 2012).

Because most splicing events (constitutive and alternative)

occur cotranscriptionally, an important goal is to determine the

extent to which chromatin and transcription factors impact

them. Understanding such links necessitates considering the

possible contribution of each step in transcription, through

initiation, elongation, and termination, and therefore also how

transcription is impacted by different chromatin states.

Promoter-Directed Control of Splicing

Pioneering studies performed in the late 90’s employing trans-

fected minigene reporter experiments demonstrated that the

type of promoter used to drive transcription by RNA poly-

merase II (Pol II) can impact the level of AS of a downstream

exon (Cramer et al., 1997). Two nonexclusive models were

proposed to explain this effect (Figure 2). In the ‘‘recruitment

model,’’ a change in promoter architecture results in the recruit-

ment of one or more splicing factors to the transcription

machinery that in turn impact splicing of the nascent RNA. In

the ‘‘kinetic model,’’ the change in promoter architecture affects

the elongation rate of Pol II, such that there is more or less time

for splice sites or other splicing signals flanking the alternative

exon to be recognized by trans-acting factors (Kornblihtt,

2007). For example, if these splice sites are weak (i.e., they

deviate from consensus splice site sequences associated with

efficient recognition by the splicing machinery), rapid elongation

will expose distal, stronger splice sites such that exon skipping

occurs, as productive splicing complexes will associate with

the stronger splice sites first. If elongation is slow, there is

increased time for splicing factors to bind to the weak sites in

the nascent RNA and promote exon inclusion. Alternatively,

reduced Pol II elongation kinetics can also favor the recognition

of splicing silencer elements surrounding an alternative exon,

resulting in increased exon skipping.

Although the mechanistic basis of promoter-dependent

effects on AS has been investigated using model splicing

reporters (see below), it is unclear to what extent and under

what conditions natural switching of promoters may function in

the regulation of downstream AS events in vivo. The analysis

of large collections of full-length transcript sequences has

revealed weak correlations between the use of alternative tran-

script start sites and the splicing of downstream cassette exons

(Chern et al., 2008), although it was not determinedwhether such

correlations may reflect tissue-dependent effects that indepen-

dently result in the increased complexity of transcription start

site usage, and the increased complexity of AS. With the accu-

mulation of data sets from the modENCODE/ENCODE projects

and other studies that have yielded parallel genome-

wide surveys of multiple aspects of gene regulation, including

transcription factor occupancy, epigenetic modifications, long-

range chromatin interactions and transcriptome profiles, it

should in principle be possible to obtain higher resolution pre-

dictions of causative promoter-dependent effects on splicing

and other RNA processing steps.

Despite our incomplete understanding of promoter-depen-

dent effects on RNA processing in vivo, evidence from numerous

model systems indicates that the strength and composition

of a promoter can impact splicing outcomes. For example, the

recruitment of the multifunctional proteins PSF/p54nrb by

promoter-bound activators stimulates splicing of first introns

(Rosonina et al., 2005). Activation of hormone receptors by

cognate ligands has been linked to specific splicing outcomes

(Auboeuf et al., 2002), and the association of PGC-1, a transcrip-

tional coactivator that plays a major role in the regulation of

adaptive thermogenesis, alters splicing activity when it is bound

to a gene (Monsalve et al., 2000). Interestingly, PGC-1 contains

an RS domain that may function to recruit splicing factors to

PGC-1-activated promoters. In the above and additional ex-

amples, the type of promoter-bound activator may influence

splicing outcomes, in part by altering the composition and/or

the processivity of Pol II (David and Manley, 2011). Under-

standing such effects therefore entails knowledge of factors

that bridge activators and Pol II, and of components of Pol II

that in turn transmit information to the nascent RNA to impact

splicing.

A recent study suggests that the Mediator complex may be

involved in integrating and relaying information to direct splicing

decisions (Huang et al., 2012). Mediator is a large multisubunit

complex that functions as a general factor at the interface

between promoter-bound transcriptional activators and Pol II

(Malik and Roeder, 2010). In addition to its general role,

locus-specific functions have been ascribed to Mediator, where

changes in its composition can lead to differential outcomes

in transcription, and possibly RNA processing. Huang and

colleagues showed that the MED23 subunit of Mediator physi-

cally interacts with several splicing and polyadenylation factors,

most notably hnRNP L (Huang et al., 2012). Indeed, MED23

was required for regulating the AS of a subset of hnRNP

L targets. It will be of interest to determine how and to what

extent Mediator relays information to impact the splicing

machinery on hnRNP L-regulated targets, and whether it acts

similarly to regulate RNA processing through other RNA-

binding proteins.

The RNA Polymerase II CTD in Splicing Control

The C-terminal domain (CTD) of Pol II’s largest subunit impacts

different stages of mRNA biogenesis, including addition of

a protective cap structure on the 50-end, splicing and formation

of the mature 30-end. The CTD consists of a repeating heptad

amino acid sequence with the consensus Y1S2P3T4S5P6S7,

and is predicted to be unstructured in isolation of other factors

(Hsin and Manley, 2012). The CTD can be posttranslationally

modified by phosphorylation on each of the residues

Y1S2T4S5S7, and these changes play important and distinct roles

in transcription and RNA processing (Hsin and Manley, 2012).

Initial evidence for a role of the CTD in RNA processing came

from experiments employing expression of an alpha-amanitin

resistant mutant of Pol II that harbors a truncated CTD. Trunca-

tion to five repeats led to defects in capping, splicing, and 30-endprocessing of model pre-mRNA reporters (McCracken et al.,

1997b; McCracken et al., 1997a), and the CTD was later found

to affect AS outcomes (de la Mata and Kornblihtt, 2006;

Rosonina and Blencowe, 2004). The CTD promotes capping

and 30-end formation through direct interactions with sets of

factors dedicated to these processes, and increasing evidence

indicates that it also serves as a platform to recruit splicing

factors that may participate in commitment complex formation

and the regulation of AS (David and Manley, 2011; Hsin and

Manley, 2012).

Affinity chromatography identified splicing and dual splicing/

transcription-associated factors as CTD-binding proteins.

These include yeast Prp40, human TCERG1/CA150, p54nrb/

PSF proteins, SR proteins, and U2AF (Hsin and Manley, 2012).

Recent work supports an RNA-dependent interaction of U2AF

with the phosphorylated CTD to stimulate splicing in vitro

through an association with the core spliceosomal factor

PRP19C (David et al., 2011). Taken together with previous

work showing that a phosphorylated CTD polypeptide can stim-

ulate splicing in vitro (Hirose et al., 1999) and that the CTD is

more active in promoting splicing of a substrate that has the

capacity to form exon-definition interactions compared to

a substrate that cannot (Zeng and Berget, 2000), it is interesting

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1255

Figure 2. Models for Chromatin and Transcription Elongation-Mediated Modulation of Alternative Splicing(Top Left) Promoter recruitment model. Different promoters differentially recruit splicing factors to the transcription complex. At promoters which fail to recruita key splicing factor (shown as an SR protein), the regulated alternative exon (exon 2) will be skipped, whereas genes containing promoters that recruit the splicingfactor will include exon 2.(Top Right) Promoter-directed kinetic model. Different promoters assemble transcription complexes capable of different transcription elongation rates. Atpromoters that assemble fast transcription elongation complexes, the regulated alternative exon (exon 2) will be skipped, whereas genes containing promotersthat assemble slow elongation complexes will include exon 2. This model requires that the alternative exon contains weak 30 and/or 50 splice sites in order to beskipped when the gene is rapidly transcribed.(Bottom Left) Chromatin-mediated recruitment model. The splicing of an alternative exon can be regulated by the chromatin-mediated recruitment of a splicingrepressor. In cells that skip the exon, an adaptor protein associates with the nucleosome assembled at the alternative exon, which in turn recruits a splicingrepressor. In cells that include the alternative exon, the adaptor protein and/or repressor are not expressed, or the nucleosome at the regulated alternative exon is

(legend continued on next page)

1256 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

to consider that the CTDmight function as a platform to facilitate

exon definition and commitment complex formation (Figures 1

and 2). In this manner, the CTD may also serve to tether exons

separated by great intronic distances to promote cotranscrip-

tional splicing (Dye et al., 2006). It will be important to determine

whether the CTD plays such roles in vivo in future work.

RNA Polymerase II Elongation and the Control

of Alternative Splicing

Numerous studies employing model experimental systems

designed to alter the rate of Pol II elongation have provided

evidence supporting the aforementioned kinetic model (Korn-

blihtt, 2007; Luco et al., 2011). More recent work has applied

genome-wide approaches to understand the extent and func-

tional relevance of this mode of regulation. In one study,

UV-induced DNA damage was found to result in a hyperphos-

phorylated form of the CTD and reduced Pol II elongation

kinetics, and these changes were proposed to cause changes

in AS of genes that function in cell cycle control and apoptosis

(Munoz et al., 2009). Another study globally monitored AS

changes following treatment of cells with camptothecin and

5,6-dichloro-1-b-D-ribofuranosyl-benzimidazole (DRB), which

act through different mechanisms to inhibit Pol II elongation

(Ip et al., 2011). Concentrations of these drugs that partially

inhibit Pol II elongation preferentially affected AS and transcript

levels of genes encoding RNA splicing factors and other RNA-

binding protein (RBP) genes. Many of the induced AS changes

introduced premature termination codons (PTCs) that elicited

nonsense-mediated mRNA decay (NMD; see below), which

further contributed to reductions in transcript levels. These

results suggest that conditions globally impacting elongation

rates can lead to the AS-mediated downregulation of RNA

processing factors, such that the levels of these factors are

calibrated with the overall RNA processing ‘‘needs’’ of the cell.

This type of Pol II-coupled AS network appears to be highly

conserved, because amino acid starvation, which causes

reduced elongation and/or increased Pol II pausing, was also

found to affect the AS of transcripts from splicing factor genes,

including several that can elicit NMD, in C. elegans (Ip et al.,

2011).

Chromatin Structure Distinguishes Exons from Introns

Although recognition of splice sites fundamentally has to occur

through direct interactions with pre-mRNA, chromatin features

can shape decisions about splice site usage and exon selection.

The basic unit of chromatin structure is the nucleosome, which

comprises 147 base-pairs of DNA wrapped around a histone

octamer consisting of two copies each of histones H2A, H2B,

H3, and H4 (Luger et al., 1997). Chromatin function can be regu-

lated by substituting canonical histones with nonallelic variants

and through posttranslational modification of histone tail resi-

dues most notably by methylation and acetylation (Kouzarides,

2007; Talbert and Henikoff, 2010). These histone ‘‘marks’’ and

direct modifications of DNA, including the addition of

not modified and therefore cannot recruit the repressor. Similar to this model, a nactivator, as proposed for Psip1/Ledgf (Pradeepa et al., 2012).(Bottom Right) Chromatin-mediated kinetic model. The splicing of an alternativescription elongation. Unmodified nucleosomes can be transcribed rapidly, resultiassembled on exon 2 has an H3K9me3 mark, CBX3 interacts with the modified nsplicing of the regulated alternative exon.

5-methylcytosine, 5-hydroxymethylcytosine, and other deriva-

tives (Wu and Zhang, 2011), affect the functional state of chro-

matin by altering its compaction and by modulating the binding

of effector proteins. It is well established that these features

have nonuniform distribution along genes with unique signatures

marking promoters and gene bodies in a transcription-depen-

dent manner (Smolle and Workman, 2013). More recently, it

has become apparent that these chromatin features are also

differentially distributed with respect to exon-intron boundaries,

and that this differential marking participates in exon recognition.

Analysis of data sets from chromatin immunoprecipitation

high-throughput sequencing (ChIP-seq), and from micrococcal

nuclease digestion followed by sequencing revealed that nucle-

osomes in a range of organisms display increased occupancy

over exons relative to neighboring intronic sequence (Andersson

et al., 2009; Chodavarapu et al., 2010; Schwartz et al., 2009;

Spies et al., 2009; Tilgner et al., 2009; Wilhelm et al., 2011). Sug-

gesting a possible role in facilitating splicing, exons that have

weak splice sites and that are surrounded by relatively long

introns have greater levels of nucleosome occupancy than do

exons with strong splice sites or that are flanked by short introns

(Spies et al., 2009; Tilgner et al., 2009). To assess whether exon-

enriched nucleosomes might be compositionally—and therefore

functionally—distinct, a number of studies examined global

distributions of specific histone modifications with respect to

exon-intron boundaries (Andersson et al., 2009; Dhami et al.,

2010; Hon et al., 2009; Huff et al., 2010; Kolasinska-Zwierz

et al., 2009; Schwartz et al., 2009; Spies et al., 2009). Some of

these studies reached different conclusions as to which modifi-

cations show enrichment over exons and to what extent such

enrichment is a consequence of increased nucleosome occu-

pancy. Nevertheless, trimethylation of lysine 36 on histone H3

(H3K36me3) was shown in multiple studies to be enriched over

exons above background nucleosome levels (Andersson et al.,

2009; Huff et al., 2010; Spies et al., 2009). Exon-enriched nucle-

osomes may also differ in their histone variant composition. The

H2A variant, H2A.Bbd, which is associated with active, intron-

containing genes, is enriched in positioned nucleosomes

flanking both 50 and 30splice sites (Tolstorukov et al., 2012).

Such specific histone marks or variants could therefore play

a widespread role in splicing (see below).

Base pair composition affects physical properties of the DNA

and is not uniform across the genome. Exons are in general

associated with higher GC content, which is an important feature

governing nucleosome occupancy (Tillo and Hughes, 2009). A

recent study found differences in relative GC content between

exons and introns that may have evolved to contribute to splicing

(Amit et al., 2012). In a reconstructed ‘‘ancestral’’ state, genes

contained exons with a low GC content that were flanked by

short introns of an even lower GC content. These subsequently

diverged to yield two different types of gene architectures in

animal species. In one architectural state, genes retained low

ucleosome-associated adaptor protein may also function to recruit a splicing

exon can be regulated by a chromatin-mediated change in the rate of tran-ng in skipping of the regulated alternative exon. In cells where the nucleosomeucleosome, slows down the transcription elongation complex, and enhances

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1257

exonic GC content with lower GC content in introns but experi-

enced an increase in intron length. In the other state, genes

retained short intron length but saw an overall increase in GC

content that eliminated differential exon-intron composition

(Amit et al., 2012). Bioinformatic and experimental evidence

supports a role for differential GC content in promoting exon

recognition in the context of the first type of architecture (Amit

et al., 2012). However, to what extent differential GC content

between exons and introns influences exon recognition through

possible mechanisms associated with (modified) nucleosome

deposition is unclear.

Studies employing genome-wide bisulphite sequencing have

suggested a role for modified cytosines at exonic CpG dinucle-

otides in exon recognition and the regulation of AS. Modified

CpG dinucleotides are enriched within exons relative to introns

in both plants and animals (Chodavarapu et al., 2010; Feng

et al., 2010; Laurent et al., 2010) with characteristic patterns at

the 50 and 30 splice sites (Laurent et al., 2010). Moreover,

widespread differences in CpG methylation have been detected

betweenworker and queen bee genomes, and intriguingly, some

of these differential methylation patterns appear to correlate with

differential AS (Lyko et al., 2010). Highlighting a possible role of

DNA epigenetic marks in mediating tissue-specific differences,

in mammalian neuronal tissues hydroxymethylation rather than

methylation was found to have significant exonic enrichment

(Khare et al., 2012). The possible mechanisms by which such

modifications affect splicing await future work.

Chromatin-Dependent Recruitment of the Splicing

Machinery

Analogous to roles of promoter architecture and the Pol II

CTD, accumulating evidence suggests that chromatin structure

throughout a gene facilitates splicing factor recruitment to

nascent transcripts. It has been proposed that splicing factors

interact with chromatin directly, or indirectly through inter-

mediate ‘‘adaptor’’ proteins (Figure 2). H3K4me3, which marks

the promoters of actively-transcribed genes, binds specifically

to CHD1, a protein that associates with U2 snRNP. Indeed,

this interaction was shown to increase splicing efficiency (Sims

et al., 2007). Similarly, H3K36me3, which is enriched over exons,

was recently reported to interact with a short splice iso-

form of Psip1/Ledgf, which in turn associates with several

splicing factors including the SR protein SRSF1 (Pradeepa

et al., 2012). Supporting a possible role as a recruitment adaptor,

knockdown of Psip1 led to a change in SRSF1 localization and

affected AS.

The aforementioned H2A.Bbd histone variant appears to func-

tion in splicing through the recruitment of splicing components

(Tolstorukov et al., 2012). Mass spectrometry data revealed

that H2A.Bbd interacts with numerous components of the

spliceosome, and depletion of this histone variant led to the

widespread disruption of constitutive and alternative splicing.

Another recent study suggests that recruitment of splicing

components by chromatin may be effected through global

changes in histone hyperacetylation or changes in the levels of

the heterochromatin-associated protein HP1a (Schor et al.,

2012). These alterations result in the global redistribution of

numerous splicing factors from chromatin to nuclear speckle

domains, which are thought to predominantly represent sites

1258 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

of splicing factor storage (Schor et al., 2012) (see below). Collec-

tively, these studies point to characteristic patterns of chromatin

structure associated with active gene expression that may

have a widespread impact on the nuclear localization of the

splicing machinery, which in turn can impact splicing of nascent

transcripts.

Chromatin structure can be altered in highly specific ways

within genes, for example, in response to environmental and

developmental cues. Such ‘‘local’’ changes are thought to also

impact AS of proximal exons on nascent RNA through the action

of adaptor proteins that bridge chromatin marks and splicing

factors. The first example of this type of proposed mechanism

involves the mutually exclusive exons IIIb and IIIc in the FGFR2

gene. Switching from exon IIIb to exon IIIc alters the ligand

affinity of this receptor and represents an important step in

the epithelial to mesenchymal transition. In mesenchymal cells,

the region encompassing these exons is characterized by

elevated levels of H3K36me3 and low levels of H3K4me3 and

H3K27me3 (Luco et al., 2010). H3K36me3 modifications favor

the binding of MRG15, which promotes the recruitment of the

splicing regulator PTBP1 to nascent RNA, and as a consequence

represses the use of exon IIIb in these cells (Luco et al., 2010).

Consistent with a more widespread role for an MRG15-adaptor

mechanism to control AS, significantly overlapping subsets

of cassette exons were affected by individual knockdown of

MRG15 and PTBP1 (Luco et al., 2010). However, the affected

exons generally displayed modest changes in inclusion level

and were found to be surrounded by relatively weak PTBP1-

binding sites, suggesting that this adaptor mechanism may be

more important for augmenting or stabilizing patterns of AS

achieved by direct action of RNA-based regulators, rather than

acting to promote pronounced cell-type-dependent, switch-

like regulation of AS.

Chromatin Structure Affects Splicing by Influencing Pol

II Elongation

Specific features of chromatin structure, as well as chromatin-

associated regulators, can influence splice site choice by

impacting transcription elongation (Figure 2). SWI/SNF chro-

matin remodelling factors interact directly with Pol II (Neish

et al., 1998; Wilson et al., 1996), and with splicing factors

(Batsche et al., 2006), suggesting that these factorsmight impact

splicing in an elongation-dependent manner. Supporting this

view, the association of the ATP-dependent SWI/SNF-type

chromatin remodelling factor BRM with the human CD44 gene

coincides with a change in inclusion levels of alternative exons

in CD44 transcripts (Batsche et al., 2006). Increased occupancy

of Pol II with elevated S5 phosphorylation of the CTD (which is

associatedwith a paused form of Pol II) was detected specifically

over CD44 alternative exons, indicating that a reduced elonga-

tion rate or increased pausing of Pol II might be responsible for

the change in AS. The BRM ATPase activity required for chro-

matin remodeling was, however, not required for the change in

AS (Batsche et al., 2006).

Recent studies analyzing BRM in Drosophila suggest that

it acts together with other members of the SWI/SNF com-

plex to regulate AS and polyadenylation in a locus-specific

manner (Waldholm et al., 2011; Zraly and Dingwall, 2012). Devel-

opmentally regulated intron retention of the Eig71Eh pre-mRNA

Figure 3. Reverse-Coupling Mechanisms(A) Splicing enhances transcription-associated histone modification. Splicingof the first intron enhances transcription initiation and stabilizes promoter-associated marks, including H3K4me3 and H3K9ac, near the 50 splice site ofexon 1. Splicing may also facilitate a transition between the elongation-associated marks H3K79me2 and H3K36me3 at the 30 splice site of the firstintron. Internal exons are particularly enriched for H3K36me3-modifiednucleosomes, due in part to splicing-increased nucleosome occupancy andaction of the histone methyltransferase SETD2 associated with elongatingPol II. These marks may also serve to reinforce splicing patterns of nascentpre-mRNA.(B) The SR protein SRSF2/SC35, which regulates splicing of alternative exons,also enhances transcription elongation by recruiting P-TEFb. P-TEFb phos-phorylates the Pol II CTD at Serine 2, which enhances the rate of transcriptionelongation.(C) The Hu family of splicing regulators bind to AU-rich sequences withinintrons and repress the splicing of regulated alternative exons. Shown here,HuR interacts with and represses the activity of the histone deacetylase,HDAC2, which stabilizes nearby acetylated nucleosomes. Acetylated nucle-osomes may enhance the rate of transcription elongation, and consequently,promote the skipping of exons with weak splice sites.

required the SNR1/SNF5 subunit, which suppresses BRM

ATPase, and reduced elongation was correlated with more effi-

cient intron splicing (Zraly and Dingwall, 2012).

Covalent modifications of histones impinge on Pol II elonga-

tion in ways that impact AS (Figure 2). The heterochromatin

protein HP1g/CBX3, which binds di- and trimethylated histone

H3K9 (Bannister et al., 2001; Lachner et al., 2001), mediates

inclusion of alternative exons in CD44 transcripts in human cells

upon stimulation of the PKC pathway, concomitantly with an

increase in Pol II occupancy over the alternatively spliced region

(Saint-Andre et al., 2011). However, CBX3 may also play a more

direct role in splicing factor recruitment. Depletion of CBX3

in human cells resulted in the accumulation of unspliced

transcripts and loss of recruitment of the U1 snRNP-70 kDa

(SNRNP70) protein and other splicing factors to active chromatin

(Smallwood et al., 2012).

Intriguingly, components of the RNAi machinery in association

with CBX3 were recently shown to also regulate AS of CD44

transcripts. Specifically, the Argonaute proteins AGO1 and

AGO2 were found by ChIP-seq analysis to bind the alternative

exon-containing region of CD44 and were loaded onto this

region by short RNAs derived from CD44 antisense transcripts

(Ameyar-Zazoua et al., 2012). Recruitment of AGO1 and AGO2

to CD44 required Dicer and CBX3 and resulted in increased

histone H3K9 methylation over the variant exons. Recruitment

of AGO proteins to the CD44 gene thus appears to locally induce

a chromatin state that affects Pol II elongation and AS.

RNA-binding proteins bound to nascent RNA may also alter

chromatin composition in ways that impact elongation and

splicing (Figure 3). Hu-family proteins, which have well defined

roles in the control of mRNA stability, were recently shown to

regulate AS by binding to nascent RNA proximal to alternative

exons in a manner that induced local histone hyperacetylation

and increased Pol II elongation (Mukherjee et al., 2011; Zhou

et al., 2011). This activity was linked to the direct inhibition of

histone deacetylase 2 (HDAC2) by Hu proteins (Zhou et al.,

2011).

RNA Pol II elongation rates are also impacted by nucleotide

sequence composition. A/T-rich sequences, in particular, are

more difficult for Pol II to transcribe. A novel complex found to

be associated with human mRNPs, termed DBIRD, facilitates

Pol II elongation across A/T rich sequences (Close et al., 2012).

Depletion of this complex resulted in reduced Pol II elongation

and changes in the splicing of exons proximal to A/T-rich

sequences. It was therefore proposed that DBIRD acts at the

interface of RNA Pol II and mRNP complexes to control AS

(Close et al., 2012).

Finally, the zinc finger DNA-binding transcription factor and

chromatin organizer CTCF has been linked to the regulation of

AS of exon 5 of the receptor-linked protein tyrosine phosphatase

CD45, and of other transcripts, by locally affecting Pol II elonga-

tion (Shukla et al., 2011). Variable inclusion of CD45 exon 5 is

controlled by RNA-binding proteins during peripheral lym-

phocyte maturation (Motta-Mena et al., 2010). Intriguingly,

CTCF appears to maintain the inclusion of exon 5 at the terminal

stages of lymphocyte development by causing Pol II pausing

proximal to this exon (Shukla et al., 2011). CTCF binding is

inhibited by CpG methylation. Accordingly, increased methyla-

tion proximal to CD45 exon 5 led to reduced CTCF occupancy

and reduced exon inclusion (Shukla et al., 2011). Analysis of

AS changes genome-wide using RNA-Seq following depletion

of CTCF further revealed that this factor is likely to have

a more widespread role in regulating AS through altering Pol II

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1259

elongation kinetics. However, CTCF is known to mediate intra-

chromosomal interactions (Ohlsson et al., 2010), and it therefore

remains to be determined whether the changes in AS caused by

CTCF reflect a direct inhibition of Pol II elongation, or whether

these effects are a consequence of more complex topological

changes to chromatin architecture.

In the examples described above and others (Luco et al.,

2011), changes in AS can be achieved through a variety of

mechanisms that perturb Pol II elongation in a widespread or

locus-specific manner. In other cases, AS is affected through

mechanisms involving the differential recruitment of splicing

factors to transcription or chromatin components. It is currently

unclear to what extent these mechanisms are distinct or overlap

as the recruitment of splicing factors to a transcript in some

cases appears to affect elongation kinetics, and in other cases

altered elongation kinetics may affect the recruitment of splicing

components to chromatin or transcription factors associated

with nascent transcripts. For example, as summarized earlier,

regulation of variable exon inclusion in CD44 transcripts appears

to involve the concerted action of chromatin remodeling, inhibi-

tion of Pol II elongation, and the recruitment of splicing factors

and the RNAi machinery. Individual genes may therefore

possess a unique set of mechanistic principles that are governed

by the specific combinatorial interplay between cis elements of

the splicing code and genomic features, which together deter-

mine the formation and activity of chromatin features and

transcription complexes. The increased use of comparative

analyses of parallel data sets interrogating transcriptomic,

genomic, and chromatin features should nevertheless facilitate

a more detailed mechanistic understanding of common princi-

ples by which chromatin, transcription, and splicing are coupled

to coordinate the regulation of subsets of genes.

Regulation of Chromatin and Transcription

by the Splicing Machinery

In addition to the extensive set of interactions and mechanisms

by which chromatin and transcription components can impact

splicing, increasing evidence indicates that splicing can have

a major impact on chromatin organization and transcriptional

output. Early indications of this ‘‘reverse-coupling’’ were that

the efficient expression of transgene constructs required the

presence of an intron (Brinster et al., 1988). Such effects were

later shown to arise in part as a consequence of enhanced tran-

scription (Furger et al., 2002). Subsequent studies have demon-

strated several mechanisms by which the splicing of nascent

transcripts can impact chromatin organization and transcription.

For example, H3K4me3 and H3K9ac, both of which are associ-

ated with active genes and widely assumed to peak in proximity

to promoters together with increased Pol II occupancy, are in

fact concentrated over first exon-intron boundaries (Bieberstein

et al., 2012) (Figure 3A). In genes with long first exons, these

marks are reduced at promoters, whereas in genes with short

first exons, the marks are increased at promoters as are tran-

scription levels. Confirming a role for first intron splicing in estab-

lishing promoter proximal architecture, intron deletion reduced

H3K4me3 levels and transcriptional output (Bieberstein et al.,

2012). Taken together with previous observations of associa-

tions between U1 and Pol II (Damgaard et al., 2008), and

between U2 snRNP and H3K4me3 (Sims et al., 2007), a picture

1260 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

emerges in which first intron splicing serves to establish or

perhaps reinforce promoter proximal marks, that in turn recruit

general transcription factors and Pol II to enhance initiation.

The enrichment of H3K36me3 at exons, which is established

by the methyltransferase SETD2 as it travels with elongating

Pol II, also arises in part as a consequence of splicing (Figure 3A).

Global inhibition of splicing (via depletion of specific spliceosome

components and/or exposure to the inhibitor spliceostatin)

decreasedH3K36me3 levels at particular exons, but also broadly

altered its distributionwithin genebodies (deAlmeida et al., 2011;

Kim et al., 2011). To what degree these effects are direct remains

unclear, as global inhibition of splicing would also be expected to

perturb transcription, for example, by affecting the expression

and/or deposition of transcription and chromatin factors (Bieber-

stein et al., 2012). Nonetheless, a direct role also seems likely. For

example, reciprocal H3K79me2 and H3K36me3 histone marks

transition at first intronic 30 splice site-first internal exon bound-

aries, but not at the corresponding boundaries of pseudoexons

(Huff et al., 2010) (ENCODE Project Consortium, 2012), suggests

more direct roles of splicing-dependent transitions in chromatin

modifications (Figure 3A). Moreover, mass spectrometry data

further suggests that SETD2 may associate with exon definition

complexes (Schneider et al., 2010).

Splicing also impacts Pol II pausing and elongation. An asso-

ciation between snRNPs and the Pol II elongation factor TAT-

SF1 can stimulate elongation in vitro, and this activity was further

enhanced by the presence of splicing signals in RNA (Fong and

Zhou, 2001). Because TAT-SF1 interacts with the positive elon-

gation factor P-TEFb, which phosphorylates the S2 residues of

the CTD to increase Pol II processivity, it was proposed that

the assembly of splicing complexes on nascent RNA may facili-

tate Pol II elongation across a gene (Fong and Zhou, 2001).

Additional studies have reported roles for splicing factors in

elongation. Because this topic has been reviewed elsewhere

(Pandit et al., 2008), only a few examples will be highlighted

here. Of particular interest are SR and SR-like proteins, which

have long-established roles in splicing. The S. cerevisiae SR-like

protein Npl3, for example, regulates the splicing of a subset of

introns (Chen et al., 2010; Kress et al., 2008), but it also facilitates

elongation by acting as an antitermination factor (Dermody et al.,

2008). Specific mutations in Npl3 lead to defects in the transcrip-

tion elongation and termination of �30% of genes (Dermody

et al., 2008). Npl3 binds the S2 phosphorylated CTD (Lei et al.,

2001), bringing it into close proximity to nascent RNA. Phosphor-

ylation of Npl3 was found to negatively regulate its binding to the

CTD and RNA, suggesting that unphosphorylated Npl3 specifi-

cally promotes elongation in association with Pol II (Dermody

et al., 2008).

Depletion of the SR family protein SRSF2/SC35 increases

Pol II pausing, most likely as a consequence of defective recruit-

ment of P-TEFb and reduced S2 CTD phosphorylation (Lin et al.,

2008) (Figure 3B). It is interesting to consider that Npl3, SRSF2,

and possibly other RNA-binding proteins, may also facilitate

elongation in part by preventing the formation of DNA-RNA

hybrids (or R-loops) formed by nascent RNA during transcription

(Pandit et al., 2008). Finally, it is also conceivable that SR

proteins bound to nascent RNA indirectly promote CTD phos-

phorylation and/or histone modifications that facilitate

Figure 4. Splicing Impacts the Regulation of Multiple Downstream Steps in Gene Regulation(A) Coupling connections between splicing and 30-end formation, RNA stability, and mRNA export. Splicing and 30-end formation are coupled by interactionsbetween exon-bound SR proteins and the cleavage and polyadenylation factor CFIm, and between U2AF and both CFIm and PAP. Cryptic upstream adenylationsites (PAS) are suppressed by U1 snRNP (left). Splicing impacts RNA stability by interactions between SR proteins and the EJC, which in turn interacts with theUPF proteins involved in NMD (middle). Splicing influences mRNA export through the splicing-dependent recruitment of the TREX complex, which in turninteracts with the RNA export factor TAP.(B) Multitasking roles of RBPs in splicing and alternative polyadenylation, RNA export and RNA transport. Top: the Nova RNA-binding proteins have been shownto not only regulate alternative splicing, but also alternative polyadenylation (pA). Both of these processes are modulated in a position-dependent manner withsome binding locations promoting splicing and polyadenylation and other locations repressing these processes. The result of this regulation is the generation ofmRNAs with different exons and 30 UTR sequences. Bottom: Similarly, Mbnl RNA-binding proteins impact alternative splicing in a position-dependent mannerand bind to 30 UTRs, where they function to control subcellular mRNA localization.

transcription. In this regard, it was recently shown that Npl3

associates in an RNA-independent manner with Bre1, a ubiquitin

ligase with specificity for H2B (Moehle et al., 2012) that facilitates

transcription elongation in vitro (Pavri et al., 2006).

The studies summarized above emphasize important roles for

nascent RNA splicing and the factors that control splicing in

establishing chromatin architecture and in controlling transcrip-

tion. It is interesting to consider, therefore, that a major determi-

nant of gene-specific chromatin architecture emanates from

information provided by cis-acting elements comprising the

splicing code. The previously described case of the Hu family

of hnRNP proteins is illustrative of a mechanism through which

proteins bound to nascent RNA can ‘‘reach back’’ to alter prox-

imal chromatin and affect Pol II elongation (Zhou et al., 2011)

(Figure 3C). Notably, this mode of regulation also mediates

highly ‘‘local’’ changes in chromatin structure that in turn regu-

late the AS regulation of nearby exons. A more systematic inves-

tigation of the roles of splicing components in establishing

region-specific chromatin modifications and functions will be

important for understanding the crosstalk between chromatin

and splicing.

Integration of Splicing with 30-End Processing,Turnover, and TransportCoupling and Coordination of Splicing with 30-EndFormation

Numerous studies have demonstrated communication between

factors involved in the splicing of 30-terminal introns and factors

involved in 30-end cleavage and polyadenylation (CPA), and this

topic has been reviewed in detail elsewhere (Di Giammartino

et al., 2011; Proudfoot, 2011). Similar to the formation of exon-

definition complexes, it has been proposed that U2AF binding

to the 30 splice site of a terminal exon forms interactions with

Cleavage Factor I and the CTD of poly(A) polymerase to mutually

stimulate terminal intron splicing and CPA (Millevoi et al., 2002;

Millevoi et al., 2006) (Figure 4A). SR proteins have also been

implicated in terminal exon crosstalk (Dettwiler et al., 2004;

McCracken et al., 2002). In certain cases, competition between

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1261

binding of CPA factors and splicing factors can result in physio-

logically important changes in AS and transcript levels (Evsyu-

kova et al., 2013) (see below).

In addition to their roles in the control of large networks of

alternative exons, splicing regulators such as Nova and hnRNP

H1 function in the regulation of alternative polyadenylation

(APA) through direct binding to recognition sites clustered

around the CPA signals (Katz et al., 2010; Licatalosi et al.,

2008) (Figure 4B). Although these ‘‘moonlighting’’ roles in APA

regulation appear to be largely independent of the splicing of

proximal exons/introns, regulation of AS and APA by the same

RBPs presumably is important for globally coordinating these

processes in a cell type or condition-dependent manner. For

example, transcript profiling studies have shown that APA is

widespread, affecting at least 50% of transcripts from human

genes (Tian et al., 2005) and that it plays an important role in

controlling the presence of miRNA and RNA-binding protein

target sites in UTR sequences, and therefore mRNA expression

levels (Mayr and Bartel, 2009; Sandberg et al., 2008). Control of

APA and AS by an overlapping set of RBP regulators may

therefore constitute an effective mechanism for functionally

coordinating these steps in RNA processing.

In an analogous manner, U1 snRNP also has dual roles in

splicing and CPA. U1 snRNP is more abundant than other

spliceosomal snRNPs, and this observation hinted that it may

have additional functions in the nucleus. Indeed, recent studies

have shown that, through binding to cryptic 50 splice sites within

pre-mRNAs, U1 snRNP can inhibit premature 30-end formation

at potential CPA sites that are distributed along pre-mRNAs

(Berg et al., 2012) (Figure 4A). In situations where U1 snRNP

becomes limiting, for example during bursts of pre-mRNA

transcription upon activation of neurons or immune cells, where

the ratio of cryptic and bona-fide 50 splice sites may be in excess

of available U1 snRNP, premature CPA sites are activated

leading to transcript shortening (Berg et al., 2012). Furthermore,

reduced U1 snRNP to pre-mRNA ratios resulted in changes in

terminal exon usage, consistent with the mutual stimulation

between the splicing and CPA machineries in terminal exon

definition. The discovery of a role for U1 snRNP in suppressing

CPA has provided further insight into the mechanism by which

certain mutations in 30 UTRs cause disease. For example,

a mutation in the 30 UTR of the p14/ROBLD3 receptor gene

that is causally linked to immunodeficiency creates a 50 splicesite that does not activate splicing but suppresses CPA, leading

to reduced p14/ROBLD3 expression (Langemeier et al., 2012).

Splicing Modulates RNA Stability and Transport

The NMD pathway acts to prevent spurious expression of

incompletely processed or mutant transcripts (Rebbapragada

and Lykke-Andersen, 2009). Although the NMD pathway

appears to be present in some form in all eukaryotes, there are

nonetheless species-specific differences, particularly in the

way PTCs are recognized and in the nature of the degradation

pathways involved. In mammalian cells, PTC recognition relies

to a large extent on deposition of the exon junction complex

(EJC) 20–24 nt upstream of exon-exon junctions. The EJC

encompasses a stable tetrameric core consisting of eIF4AIII,

MAGOH, MLN51, and Y14 proteins, which is deposited on

mRNA during splicing (Tange et al., 2005). This core associates

1262 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

with a host of SR and SR-related proteins to form megadalton

size complexes that presumably function in mRNP compac-

tion as well as in facilitating coupling of splicing with downstream

steps in gene expression (Singh et al., 2012) (Figures 1 and 4A).

During the pioneer round of translation, EJCs are displaced by

the ribosome (Isken et al., 2008). However, when the ribosome

encounters a PTC more than 50–55 nt upstream of a terminal

exon-exon junction, EJC components associate with upstream

frame shift (UPF) proteins (Figure 4A) that trigger release of the

ribosome through interaction with release factors (eRFs). These

and other interactions ultimately lead to mRNA decay through

pathways that involve 50-end decapping, deadenylation, and

exoribonucleolytic enzymes (Schoenberg and Maquat, 2012).

Alternative splicing coupled to NMD controls the levels

of specific subsets of genes. It has been estimated that approx-

imately 10%–20% of AS events that have the potential to

introduce PTCs lead to substantial changes in overall total

steady-state transcript levels (Pan et al., 2006). In many cases,

these AS-coupled NMD events serve to auto- and cross-

regulate expression levels of regulatory and core factors

involved in splicing and other aspects of RNA metabolism

(Cuccurese et al., 2005; Lareau et al., 2007b; Mitrovich and

Anderson, 2000; Ni et al., 2007; Plocik and Guthrie, 2012;

Saltzman et al., 2008), but important roles in the regulation of

other classes of proteins have also been reported (Barash

et al., 2010; Lareau et al., 2007a).

It is important for a cell to prevent incompletely or aberrantly

processed transcripts from being translated, as such transcripts

may express truncated proteins with aberrant or dominant

negative functions that have harmful consequences. One safe-

guarding mechanism is to prevent release of such transcripts

from the nucleus. The TREX (transcription/export) complex is

a conserved multiprotein complex that links transcription elon-

gation with nuclearmRNA export (Katahira et al., 2009). Although

S. cerevisiae TREX is recruited to intronless transcripts (Strasser

et al., 2002), its mammalian counterpart is incorporated into

maturing mRNPs by the splicing machinery (Masuda et al.,

2005) and further requires binding of the 50 cap by the TREX

component Aly (Cheng et al., 2006). TREX thenmediates associ-

ation with the TAP nuclear export receptor to facilitate mRNA

export through the nuclear pore complex (Stutz et al., 2000;

Zhou et al., 2000) (Figure 4A). Natural intronless genes can

circumvent the necessity for splicing to recruit TREX through

sequence elements that directly mediate TREX- and TAP-

dependent export (Lei et al., 2011). However, transcripts from

some intron-containing yeast genes, for example the gene

encoding the nuclear export factor SUS1, require introns for

efficient nuclear mRNA export (Cuenca-Bono et al., 2011) (see

below).

Regulated intron retention has been harnessed to play impor-

tant regulatory roles in the control of transcript levels. For

example, coordinated regulation of a set of alternative retained

introns controls the expression of the neuron-specific genes

Stx1b, Vamp2, Sv2a, and Kif5a. The splicing regulator Ptbp1,

which is expressed widely in nonneural cells, represses splicing

of these introns, such that the unspliced transcripts are retained

in the nucleus where they are degraded by the exosome

(Yap et al., 2012). Inhibition of Ptbp1 expression by miR-124 in

Figure 5. Organization of the Splicing Components in the Cell NucleusMajor nuclear domains enriched in splicing and other factors in the mammalian cell nucleus are depicted with known and putative roles indicated. Gray areasindicate nucleoli.

neural cells results in splicing of these introns, allowing export

and translation of the resulting mature mRNAs. With the wealth

of available transcriptome profiling data, it can be expected

that many additional examples of regulated intron removal linked

to functions such as mRNA turnover and transport will soon

emerge.

Although the EJC appears to be seldom required for NMD in

Drosophila, it is important for the localization of developmen-

tally important transcripts. Localization of oskar mRNA to the

posterior pole of the oocyte requires the deposition of the

EJC core components together with an exon-exon junction-

spanning localization element formed by splicing of the first

intron (Ghosh et al., 2012). Changes in alternative splicing,

particularly in UTR regions, have been observed to differen-

tially regulate mRNA localization in mammalian cells (La Via

et al., 2013; Terenzi and Ladd, 2010) and likely represent

a more widely used mode of regulation than currently appreci-

ated. Similar to previously mentioned examples in which

specific RBPs have roles in both AS and APA, specific RBPs

that function in AS regulation can also function in mRNA local-

ization. Transcriptome profiling of cells and tissues deficient of

MBNL1 and MBNL2, coupled with analysis of the in vivo target

sites of these proteins, has revealed that they regulate large

networks of alternative exons involved in differentiation and

development (Charizanis et al., 2012; Wang et al., 2012)

(Figure 4B). A transcriptomic and proteomic analysis of subcel-

lular compartments further uncovered a widespread role for

MBNL proteins in the regulation of transcript localization,

translation, and protein secretion (Wang et al., 2012). These

studies underscore the importance of integrative analyses

that capture information from multiple aspects of mRNA pro-

cessing and expression when analyzing the functions of indi-

vidual RBPs. In particular, it is becoming increasingly evident

that most if not all RBPs in the cell multitask, and the extent

to which the multiple regulatory functions of RBPs arise

through physical (i.e., direct) coupling between processes, as

opposed to independently operating functions, will be

important to determine.

Dynamic Nuclear Organization in Splicing Control

The majority of the mechanisms described thus far in this review

invoke the formation and disruption of protein-protein and

protein-RNA interactions in splicing control. However, of critical

importance to any one of these mechanisms in vivo, is the local

availability of active splicing components relative to the re-

quirements for these factors presented by cognate cis-acting

elements in nascent RNA. Regulation of the availability of

splicing components provides a potentially powerful means by

which constitutive and AS events may be controlled. The highly

compartmentalized nature of the cell nucleus, which contains

several different types of nonmembranous substructures, or

‘‘bodies,’’ that concentrate RNA processing factors, provides

such a regulatory architecture. Among the domains that con-

centrate splicing and other RNA processing factors are inter-

chromatin granule clusters or ‘‘speckles,’’ paraspeckles, Cajal

Bodies (CBs) and nuclear stress bodies (Figure 5) (Biamonti

and Vourc’h, 2010; Machyna et al., 2013; Nakagawa and Hirose,

2012; Spector and Lamond, 2011).

Mammalian cell nuclei typically contain 20–50 speckle struc-

tures that concentrate snRNP and non-snRNP splicing factors,

including numerous SR family and SR-like proteins (Spector

and Lamond, 2011). Experiments employing transcriptional

inhibitors and inducible gene loci revealed that splicing factors

can shuttle between speckles and nearby sites of nascent

RNA transcription, and additional studies have shown that

this shuttling behavior can be controlled by specific kinases

and phosphatases that alter the posttranslational modification

status of SR proteins and other splicing factors. These and other

observations led to the proposal that speckles primarily repre-

sent storage sites for splicing factors (Spector and Lamond,

2011). However, more recent studies using antibodies that

specifically recognize the phosphorylated U2 snRNP protein

SF3b155 (P-SF3b155), which is found only in catalytically

activated or active spliceosomes, paint a more complex

picture (Girard et al., 2012). Immunolocalization using an

anti-P-SF3b155 antibody showed spliceosomes localized to

regions of decompacted chromatin at the periphery of—or

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1263

within—nuclear speckles (Girard et al., 2012). Inhibition of tran-

scription and splicing after SF3b155 phosphorylation further

revealed that posttranscriptional splicing occurs in nuclear

speckles. These results are consistent with results from earlier

studies employing simultaneous fluorescence in situ hybrid-

ization detection of unspliced and spliced transcripts, which

suggested that the introns of specific transcripts are spliced

within speckles (Lawrence et al., 1993).

Paraspeckles are structures that form at the periphery of

speckle domains and have been observed widely across

mammalian cells and tissues (Fox and Lamond, 2010; Naka-

gawa and Hirose, 2012). They have been implicated in the

regulation of gene expression by mediating the nuclear retention

of adenosine-to-inosine (A-to-I) edited transcripts (Fox and

Lamond, 2010). However, the recent discovery that these struc-

tures concentrate on the order of 40multifunctional RNA-binding

proteins suggests yet undiscovered roles in other aspects of

RNA processing (Naganuma et al., 2012).

Mammalian nuclei typically contain several Cajal bodies, and

these domains are thought to represent primary sites of spliceo-

somal and nonspliceosomal snRNP biogenesis, maturation, and

recycling (Machyna et al., 2013). The formation and size of CBs

relates to the transcriptional and metabolic activity of cells, and

these structures are prominent in rapidly proliferating cells.

Because the in vivo concentration of basal spliceosomal compo-

nents, including snRNPs, can impact specific subsets of AS

events (Park et al., 2004), in particular those that are predicted

to regulate levels of RNA processing factors (Saltzman et al.,

2011), it is interesting to consider that processes that control

the formation and activity of CBs could indirectly control AS of

multiple genes to globally coordinate levels of RNA processing

factors according to themetabolic requirements of the cell. Anal-

ogous to this proposed role for CBs, nuclear stress bodies are

structures that form specifically in response to a variety of stress

conditions including heat shock, oxidative stress, or exposure to

toxic materials (Biamonti and Vourc’h, 2010). These structures

are thought to mediate global changes in gene expression, in

part by sequestering splicing factors (Biamonti and Vourc’h,

2010).

An important facet of understanding the role of nuclear

domains in the control of splicing and other steps in gene regu-

lation is to determine how they are formed. Much in the way

nucleoli form around tandem repeats of rRNA genes, formation

of nuclear domains with connections to the splicing process

may be nucleated by—or depend on for integrity—specific

DNA or RNA sequences, including long (intergenic) noncoding

RNAs (lnc/lincRNAs). CBs have been detected at U1 and U2

snRNA gene loci (Smith et al., 1995), although they may

assemble via the association of multiple different protein and

nucleic acid components (Machyna et al., 2013), and stress

body formation is dependent on transcriptionally active, peri-

centric tandem repeats of satellite III sequences bound by heat

shock transcription factor 1 (HSF1) (Biamonti and Vourc’h,

2010).

Speckle domains concentrate MALAT1, a nuclear lncRNA that

appears to participate in controlling the phosphorylation state of

SR proteins (Tripathi et al., 2010). Depletion of human MALAT1

was also reported to alter the nuclear distribution of SRSF1

1264 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

and to lead to changes in SRSF1-dependent AS events (Tripathi

et al., 2010), although a more recent study did not observe such

effects (Zhang et al., 2012). Moreover, recent studies employing

Malat1 knockout mice did not reveal an essential role for this

lncRNA under normal laboratory conditions (Eißmann et al.,

2012; Nakagawa et al., 2012), whereas another study reported

that it is important for metastasis-associated properties of lung

cancer cells (Gutschner et al., 2013). NEAT1, another lncRNA,

is an integral structural component of paraspeckles (Clemson

et al., 2009; Naganuma et al., 2012). A change in the alternative

30-end processing of NEAT1 lncRNA by hnRNP K affects the

formation of these domains (Naganuma et al., 2012). Very

recently, a class of sno-lncRNAs transcribed from a genomic

region linked to Prader-Willi syndrome was shown to sequester

the RBFOX2 splicing regulator and to modulate AS (Yin et al.,

2012). As additional ncRNAs are identified and characterized, it

can be expected that many other examples of ncRNA-based

control of splicing factor availability and functional activity will

be discovered.

In addition to the aforementioned roles for DNA and RNA, it

has recently emerged that the prevalence of low complexity or

disordered protein regions in splicing and other RNA processing

factorsmay play an important role in the formation and regulation

of the activity of nuclear domains. Homotypic and heterotypic

interactions involving these domains and RNA have been shown

to form hydrogel-like structures, and it is intriguing to consider

that such structures act as malleable interfaces or ‘‘matrices’’

with which to dynamically control (i.e., by differential phosphor-

ylation or other posttranslational modifications) the accessibility,

assembly, and activity, of splicing and other highly integrated

regulatory complexes in the cell nucleus (Han et al., 2012;

Kato et al., 2012).

Conclusions and Future Perspectives

During the past several years remarkable strides have been

made in our understanding of how splicing is dynamically

integrated with other layers of gene regulation and within the

context of subnuclear structure and organization. Advance-

ments in high-throughput technologies and computational

approaches, together with focused biochemical, molecular,

and cell biological methods, have powered the discovery and

characterization of the global principles by which splicing forms

a nexus of extensive crosstalk between gene expression

processes. This crosstalk temporally coordinates and enhances,

and in some cases represses, the kinetics of physically coupled

steps in RNA metabolism, but it also serves to coordinately

regulate different steps in the transcription, processing, export,

stability, and translation of mRNA.

Of key importance in future studies will be to determine

the specific conditions and mechanisms by which chromatin-

and transcription-associated components control splicing

outcomes, and vice versa. Current models often propose

networks of physical interactions between these processes.

However, it is unclear to what extent regulatory mechanisms

may rely on increased local concentrations of factors (i.e.,

through associations with chromatin and or other nuclear

domains) that provide kinetic advantages, which in turn promote

‘‘coupled’’ effects. Regardless of the specific mechanisms by

which crosstalk impacts splicing and coupled processes, it is

exciting to consider that entirely new functional connections

await discovery. For example, the role of splicing in the deposi-

tion of specific chromatin marks such as H3K36me3 could

impact additional chromatin mark-regulated functions, such as

DNA replication, repair, andmethylation (Wagner and Carpenter,

2012). The plethora of poorly characterized histone lysine meth-

ylation ‘‘readers’’ such as the tudor, chromodomain, PWWP, and

other ‘‘royal family’’ domain-containing proteins are candidates

for mediating possible new splicing-dependent regulation

involving chromatin marks and their binding to reader proteins

(Yap and Zhou, 2010).

Another important area of future investigation is to establish

the extent to which nucleic-acid-binding proteins multitask to

coordinate different aspects of biology. Although this review

focuses on a few examples of multitasking RBPs, it is telling

that almost every recent study employing in vivo mapping of

binding sites of splicing regulators or other RBPs has uncovered

previously unknown, additional functions of these proteins.

Moreover, other in vivo crosslinking studies using polyadeny-

lated RNA as bait to comprehensively identify RBPs, point to

a much more extensive multitasking world in which transcription

factors and proteins associated with other diverse cellular func-

tions, including metabolism, may have unsuspected functions in

association with RNA (Baltz et al., 2012; Castello et al., 2012). In

this regard, it should be noted that among the largest group of

uncharacterized nucleic-acid-binding factors are C2H2 and

other zinc-finger domain proteins, defined examples of which

can regulate gene expression through binding RNA.

Increasing examples of pivotal roles for switch-like AS events

is providing a perspective in which a relatively small number of

regulated exons can act to rewire entire programs of gene regu-

lation by modifying core domains of proteins that dictate the

activities of regulators of chromatin, transcription, and other

steps in gene regulation (Irimia and Blencowe, 2012). Numerous

other AS events remodel protein interaction and signaling

networks that are important for establishing cell type-specific

functions (Babu et al., 2011; Ellis et al., 2012; Weatheritt and

Gibson, 2012). Such AS events are often found in disordered

domains of proteins that are subject to phosphorylation and

other types of posttranslational modifications. Interestingly,

these domains are often found in splicing factors and other

nuclear gene expression regulators, with the RS-repeat domains

of SR proteins and the CTD of Pol II representing notable exam-

ples. A very important area of future investigation will be to

understand how these and other protein domains contribute to

the assembly and disassembly of higher-order nuclear struc-

tures that function to organize and possibly catalyze splicing

and other nuclear reactions (Han et al., 2012; Kato et al.,

2012). Also central to this understanding will be to discover

and characterize ncRNAs that participate in the dynamic integra-

tion of splicing with other nuclear processes.

ACKNOWLEDGMENTS

We thank members of the Graveley and Blencowe laboratories for helpful dis-

cussions. B.R.G. acknowledges support fromNIH grants R01GM067842, R01

GM095296, U54 HG007005, and U54 HG006994. B.J.B. acknowledges

funding from the Canadian Institutes of Health Research, Canadian Cancer

Society, Natural Sciences and Engineering Research Council of Canada

(NSERC), and the Ontario Research Fund. U.B. was supported by European

Molecular Biology Organization and Human Frontier Science Program Fellow-

ships, S.G. was supported by an NSERCStudentship, and A.P. was supported

by an NRSA Fellowship.

REFERENCES

Ameur, A., Zaghlool, A., Halvardson, J., Wetterbom, A., Gyllensten, U.,

Cavelier, L., and Feuk, L. (2011). Total RNA sequencing reveals nascent

transcription and widespread co-transcriptional splicing in the human brain.

Nat. Struct. Mol. Biol. 18, 1435–1440.

Ameyar-Zazoua, M., Rachez, C., Souidi, M., Robin, P., Fritsch, L., Young, R.,

Morozova, N., Fenouil, R., Descostes, N., Andrau, J.-C., et al. (2012). Argo-

naute proteins couple chromatin silencing to alternative splicing. Nat. Struct.

Mol. Biol. 19, 998–1004.

Amit, M., Donyo, M., Hollander, D., Goren, A., Kim, E., Gelfman, S., Lev-Maor,

G., Burstein, D., Schwartz, S., Postolsky, B., et al. (2012). Differential GC

content between exons and introns establishes distinct strategies of splice-

site recognition. Cell Rep 1, 543–556.

Andersson, R., Enroth, S., Rada-Iglesias, A., Wadelius, C., and Komorowski, J.

(2009). Nucleosomes are well positioned in exons and carry characteristic

histone modifications. Genome Res. 19, 1732–1741.

Auboeuf, D., Honig, A., Berget, S.M., and O’Malley, B.W. (2002). Coordinate

regulation of transcription and splicing by steroid receptor coregulators.

Science 298, 416–419.

Babu, M.M., van der Lee, R., de Groot, N.S., and Gsponer, J. (2011). Intrinsi-

cally disordered proteins: regulation and disease. Curr. Opin. Struct. Biol. 21,

432–440.

Baltz, A.G., Munschauer, M., Schwanhausser, B., Vasile, A., Murakawa, Y.,

Schueler, M., Youngs, N., Penfold-Brown, D., Drew, K., Milek, M., et al.

(2012). The mRNA-bound proteome and its global occupancy profile on

protein-coding transcripts. Mol. Cell 46, 674–690.

Bannister, A.J., Zegerman, P., Partridge, J.F., Miska, E.A., Thomas, J.O.,

Allshire, R.C., and Kouzarides, T. (2001). Selective recognition of methylated

lysine 9 on histone H3 by the HP1 chromo domain. Nature 410, 120–124.

Barash, Y., Calarco, J.A., Gao,W., Pan, Q., Wang, X., Shai, O., Blencowe, B.J.,

and Frey, B.J. (2010). Deciphering the splicing code. Nature 465, 53–59.

Barbosa-Morais, N.L., Irimia, M., Pan, Q., Xiong, H.Y., Gueroussov, S., Lee,

L.J., Slobodeniuc, V., Kutter, C., Watt, S., Colak, R., et al. (2012). The

evolutionary landscape of alternative splicing in vertebrate species. Science

338, 1587–1593.

Batsche, E., Yaniv, M., and Muchardt, C. (2006). The human SWI/SNF subunit

Brm is a regulator of alternative splicing. Nat. Struct. Mol. Biol. 13, 22–29.

Berg, M.G., Singh, L.N., Younis, I., Liu, Q., Pinto, A.M., Kaida, D., Zhang, Z.,

Cho, S., Sherrill-Mix, S., Wan, L., and Dreyfuss, G. (2012). U1 snRNP deter-

mines mRNA length and regulates isoform expression. Cell 150, 53–64.

Bhatt, D.M., Pandya-Jones, A., Tong, A.-J., Barozzi, I., Lissner, M.M., Natoli,

G., Black, D.L., and Smale, S.T. (2012). Transcript dynamics of proinflamma-

tory genes revealed by sequence analysis of subcellular RNA fractions. Cell

150, 279–290.

Biamonti, G., and Vourc’h, C. (2010). Nuclear stress bodies. Cold Spring Harb.

Perspect. Biol. 2, a000695.

Bieberstein, N.I., Carrillo Oesterreich, F., Straube, K., and Neugebauer, K.M.

(2012). First exon length controls active chromatin signatures and transcrip-

tion. Cell Rep 2, 62–68.

Brinster, R.L., Allen, J.M., Behringer, R.R., Gelinas, R.E., and Palmiter, R.D.

(1988). Introns increase transcriptional efficiency in transgenic mice. Proc.

Natl. Acad. Sci. USA 85, 836–840.

Buljan, M., Chalancon, G., Eustermann, S., Wagner, G.P., Fuxreiter, M.,

Bateman, A., and Babu, M.M. (2012). Tissue-specific splicing of disordered

segments that embed binding motifs rewires protein interaction networks.

Mol. Cell 46, 871–883.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1265

Castello, A., Fischer, B., Eichelbaum, K., Horos, R., Beckmann, B.M., Strein,

C., Davey, N.E., Humphreys, D.T., Preiss, T., Steinmetz, L.M., et al. (2012).

Insights into RNA biology from an atlas of mammalian mRNA-binding proteins.

Cell 149, 1393–1406.

Charizanis, K., Lee, K.Y., Batra, R., Goodwin, M., Zhang, C., Yuan, Y., Shiue,

L., Cline, M., Scotti, M.M., Xia, G., et al. (2012). Muscleblind-like 2-mediated

alternative splicing in the developing brain and dysregulation in myotonic

dystrophy. Neuron 75, 437–450.

Chen, M., and Manley, J.L. (2009). Mechanisms of alternative splicing regula-

tion: insights from molecular and genomics approaches. Nat. Rev. Mol. Cell

Biol. 10, 741–754.

Chen, Y.C., Milliman, E.J., Goulet, I., Cote, J., Jackson, C.A., Vollbracht, J.A.,

and Yu, M.C. (2010). Protein arginine methylation facilitates cotranscriptional

recruitment of pre-mRNA splicing factors. Mol. Cell. Biol. 30, 5245–5256.

Cheng, H., Dufu, K., Lee, C.-S., Hsu, J.L., Dias, A., and Reed, R. (2006). Human

mRNA export machinery recruited to the 50 end ofmRNA. Cell 127, 1389–1400.

Chern, T.-M., Paul, N., van Nimwegen, E., and Zavolan, M. (2008). Computa-

tional analysis of full-length cDNAs reveals frequent coupling between tran-

scriptional and splicing programs. DNA Res. 15, 63–72.

Chodavarapu, R.K., Feng, S., Bernatavichute, Y.V., Chen, P.-Y., Stroud, H.,

Yu, Y., Hetzel, J.A., Kuo, F., Kim, J., Cokus, S.J., et al. (2010). Relationship

between nucleosome positioning and DNA methylation. Nature 466, 388–392.

Clemson, C.M., Hutchinson, J.N., Sara, S.A., Ensminger, A.W., Fox, A.H.,

Chess, A., and Lawrence, J.B. (2009). An architectural role for a nuclear

noncoding RNA: NEAT1 RNA is essential for the structure of paraspeckles.

Mol. Cell 33, 717–726.

Close, P., East, P., Dirac-Svejstrup, A.B., Hartmann, H., Heron, M., Maslen, S.,

Chariot, A., Soding, J., Skehel, M., and Svejstrup, J.Q. (2012). DBIRD complex

integrates alternative mRNA splicing with RNA polymerase II transcript

elongation. Nature 484, 386–389.

Cramer, P., Pesce, C.G., Baralle, F.E., and Kornblihtt, A.R. (1997). Functional

association between promoter structure and transcript alternative splicing.

Proc. Natl. Acad. Sci. USA 94, 11456–11460.

Cuccurese, M., Russo, G., Russo, A., and Pietropaolo, C. (2005). Alternative

splicing and nonsense-mediated mRNA decay regulate mammalian ribosomal

gene expression. Nucleic Acids Res. 33, 5965–5977.

Cuenca-Bono, B., Garcıa-Molinero, V., Pascual-Garcıa, P., Dopazo, H., Llopis,

A., Vilardell, J., and Rodrıguez-Navarro, S. (2011). SUS1 introns are required

for efficient mRNA nuclear export in yeast. Nucleic Acids Res. 39, 8599–8611.

Damgaard, C.K., Kahns, S., Lykke-Andersen, S., Nielsen, A.L., Jensen, T.H.,

and Kjems, J. (2008). A 50 splice site enhances the recruitment of basal tran-

scription initiation factors in vivo. Mol. Cell 29, 271–278.

David, C.J., and Manley, J.L. (2011). The RNA polymerase C-terminal domain:

a new role in spliceosome assembly. Transcription 2, 221–225.

David, C.J., Boyne, A.R., Millhouse, S.R., and Manley, J.L. (2011). The RNA

polymerase II C-terminal domain promotes splicing activation through re-

cruitment of a U2AF65-Prp19 complex. Genes Dev. 25, 972–983.

de Almeida, S.F., and Carmo-Fonseca, M. (2012). Design principles of inter-

connections between chromatin and pre-mRNA splicing. Trends Biochem.

Sci. 37, 248–253.

de Almeida, S.F., Grosso, A.R., Koch, F., Fenouil, R., Carvalho, S., Andrade, J.,

Levezinho, H., Gut, M., Eick, D., Gut, I., et al. (2011). Splicing enhances recruit-

ment of methyltransferase HYPB/Setd2 and methylation of histone H3 Lys36.

Nat. Struct. Mol. Biol. 18, 977–983.

de la Mata, M., and Kornblihtt, A.R. (2006). RNA polymerase II C-terminal

domain mediates regulation of alternative splicing by SRp20. Nat. Struct.

Mol. Biol. 13, 973–980.

Dermody, J.L., Dreyfuss, J.M., Villen, J., Ogundipe, B., Gygi, S.P., Park, P.J.,

Ponticelli, A.S., Moore, C.L., Buratowski, S., and Bucheli, M.E. (2008).

Unphosphorylated SR-like protein Npl3 stimulates RNA polymerase II elonga-

tion. PLoS ONE 3, e3273.

Dettwiler, S., Aringhieri, C., Cardinale, S., Keller, W., and Barabino, S.M.

(2004). Distinct sequence motifs within the 68-kDa subunit of cleavage factor

1266 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Immediate RNA binding, protein-protein interactions, and subcellular localiza-

tion. J. Biol. Chem. 279, 35788–35797.

Dhami, P., Saffrey, P., Bruce, A.W., Dillon, S.C., Chiang, K., Bonhoure, N.,

Koch, C.M., Bye, J., James, K., Foad, N.S., et al. (2010). Complex exon-intron

marking by histone modifications is not determined solely by nucleosome

distribution. PLoS ONE 5, e12339.

Di Giammartino, D.C., Nishida, K., and Manley, J.L. (2011). Mechanisms and

consequences of alternative polyadenylation. Mol. Cell 43, 853–866.

Dye, M.J., Gromak, N., and Proudfoot, N.J. (2006). Exon tethering in transcrip-

tion by RNA polymerase II. Mol. Cell 21, 849–859.

Eißmann, M., Gutschner, T., Hammerle, M., Gunther, S., Caudron-Herger, M.,

Groß, M., Schirmacher, P., Rippe, K., Braun, T., Zornig, M., and Diederichs, S.

(2012). Loss of the abundant nuclear non-coding RNA MALAT1 is compatible

with life and development. RNA Biol. 9, 1076–1087.

Ellis, J.D., Barrios-Rodiles, M., Colak, R., Irimia, M., Kim, T., Calarco, J.A.,

Wang, X., Pan, Q., O’Hanlon, D., Kim, P.M., et al. (2012). Tissue-specific

alternative splicing remodels protein-protein interaction networks. Mol. Cell

46, 884–892.

ENCODE Project Consortium, Dunham, I., Kundaje, A., Aldred, S.F., Collins,

P.J., Davis, C.A., Doyle, F., Epstein, C.B., Frietze, S., Harrow, J., Kaul, R.,

et al. (2012). An integrated encyclopedia of DNA elements in the human

genome. Nature 489, 57–74.

Evsyukova, I., Bradrick, S.S., Gregory, S.G., and Garcia-Blanco, M.A. (2013).

Cleavage and polyadenylation specificity factor 1 (CPSF1) regulates alterna-

tive splicing of interleukin 7 receptor (IL7R) exon 6. RNA 19, 103–115.

Feng, S., Cokus, S.J., Zhang, X., Chen, P.Y., Bostick, M., Goll, M.G., Hetzel, J.,

Jain, J., Strauss, S.H., Halpern, M.E., et al. (2010). Conservation and

divergence of methylation patterning in plants and animals. Proc. Natl.

Acad. Sci. USA 107, 8689–8694.

Fong, Y.W., and Zhou, Q. (2001). Stimulatory effect of splicing factors on

transcriptional elongation. Nature 414, 929–933.

Fox, A.H., and Lamond, A.I. (2010). Paraspeckles. Cold Spring Harb. Perspect.

Biol. 2, a000687.

Furger, A., O’Sullivan, J.M., Binnie, A., Lee, B.A., and Proudfoot, N.J. (2002).

Promoter proximal splice sites enhance transcription. Genes Dev. 16,

2792–2799.

Ghosh, S., Marchand, V., Gaspar, I., and Ephrussi, A. (2012). Control of RNP

motility and localization by a splicing-dependent structure in oskar mRNA.

Nat. Struct. Mol. Biol. 19, 441–449.

Girard, C., Will, C.L., Peng, J., Makarov, E.M., Kastner, B., Lemm, I., Urlaub,

H., Hartmuth, K., and Luhrmann, R. (2012). Post-transcriptional spliceosomes

are retained in nuclear speckles until splicing completion. Nat Commun 3, 994.

Gornemann, J., Kotovic, K.M., Hujer, K., and Neugebauer, K.M. (2005).

Cotranscriptional spliceosome assembly occurs in a stepwise fashion and

requires the cap binding complex. Mol. Cell 19, 53–63.

Graveley, B.R. (2005). Mutually exclusive splicing of the insect Dscam

pre-mRNA directed by competing intronic RNA secondary structures. Cell

123, 65–73.

Gutschner, T., Hammerle, M., Eißmann, M., Hsu, J., Kim, Y., Hung, G.,

Revenko, A.S., Arun, G., Stentrup, M., Groß, M., et al. (2013). The noncoding

RNAMALAT1 is a critical regulator of the metastasis phenotype of lung cancer

cells. Cancer Res. 73, 1180–1189.

Han, T.W., Kato, M., Xie, S., Wu, L.C., Mirzaei, H., Pei, J., Chen, M., Xie, Y.,

Allen, J., Xiao, G., and McKnight, S.L. (2012). Cell-free formation of RNA

granules: bound RNAs identify features and components of cellular assem-

blies. Cell 149, 768–779.

Hirose, Y., Tacke, R., and Manley, J.L. (1999). Phosphorylated RNA poly-

merase II stimulates pre-mRNA splicing. Genes Dev. 13, 1234–1239.

Hon, G., Wang, W., and Ren, B. (2009). Discovery and annotation of functional

chromatin signatures in the human genome. PLoS Comput. Biol. 5, e1000566.

Hoskins, A.A., and Moore, M.J. (2012). The spliceosome: a flexible, reversible

macromolecular machine. Trends Biochem. Sci. 37, 179–188.

Hoskins, A.A., Friedman, L.J., Gallagher, S.S., Crawford, D.J., Anderson, E.G.,

Wombacher, R., Ramirez, N., Cornish, V.W., Gelles, J., and Moore, M.J.

(2011). Ordered and dynamic assembly of single spliceosomes. Science

331, 1289–1295.

Hsin, J.P., and Manley, J.L. (2012). The RNA polymerase II CTD coordinates

transcription and RNA processing. Genes Dev. 26, 2119–2137.

Huang, Y., Li, W., Yao, X., Lin, Q.-J., Yin, J.-W., Liang, Y., Heiner, M., Tian, B.,

Hui, J., and Wang, G. (2012). Mediator complex regulates alternative mRNA

processing via the MED23 subunit. Mol. Cell 45, 459–469.

Huff, J.T., Plocik, A.M., Guthrie, C., and Yamamoto, K.R. (2010). Reciprocal

intronic and exonic histone modification regions in humans. Nat. Struct. Mol.

Biol. 17, 1495–1499.

Ip, J.Y., Schmidt, D., Pan, Q., Ramani, A.K., Fraser, A.G., Odom, D.T., and

Blencowe, B.J. (2011). Global impact of RNA polymerase II elongation

inhibition on alternative splicing regulation. Genome Res. 21, 390–401.

Irimia, M., and Blencowe, B.J. (2012). Alternative splicing: decoding an expan-

sive regulatory layer. Curr. Opin. Cell Biol. 24, 323–332.

Isken, O., Kim, Y.K., Hosoda, N., Mayeur, G.L., Hershey, J.W., and Maquat,

L.E. (2008). Upf1 phosphorylation triggers translational repression during

nonsense-mediated mRNA decay. Cell 133, 314–327.

Kalsotra, A., and Cooper, T.A. (2011). Functional consequences of develop-

mentally regulated alternative splicing. Nat. Rev. Genet. 12, 715–729.

Katahira, J., Inoue, H., Hurt, E., and Yoneda, Y. (2009). Adaptor Aly and

co-adaptor Thoc5 function in the Tap-p15-mediated nuclear export of

HSP70 mRNA. EMBO J. 28, 556–567.

Kato, M., Han, T.W., Xie, S., Shi, K., Du, X., Wu, L.C., Mirzaei, H., Goldsmith,

E.J., Longgood, J., Pei, J., et al. (2012). Cell-free formation of RNA granules:

low complexity sequence domains form dynamic fibers within hydrogels.

Cell 149, 753–767.

Katz, Y., Wang, E.T., Airoldi, E.M., and Burge, C.B. (2010). Analysis and design

of RNA sequencing experiments for identifying isoform regulation. Nat.

Methods 7, 1009–1015.

Khare, T., Pai, S., Koncevicius, K., Pal, M., Kriukiene, E., Liutkeviciute, Z.,

Irimia, M., Jia, P., Ptak, C., Xia, M., et al. (2012). 5-hmC in the brain is abundant

in synaptic genes and shows differences at the exon-intron boundary. Nat.

Struct. Mol. Biol. 19, 1037–1043.

Khodor, Y.L., Rodriguez, J., Abruzzi, K.C., Tang, C.-H.A., Marr, M.T., 2nd, and

Rosbash, M. (2011). Nascent-seq indicates widespread cotranscriptional

pre-mRNA splicing in Drosophila. Genes Dev. 25, 2502–2512.

Khodor, Y.L., Menet, J.S., Tolan, M., and Rosbash, M. (2012). Cotranscrip-

tional splicing efficiency differs dramatically between Drosophila and mouse.

RNA 18, 2174–2186.

Kim, S., Kim, H., Fong, N., Erickson, B., and Bentley, D.L. (2011). Pre-mRNA

splicing is a determinant of histone H3K36 methylation. Proc. Natl. Acad.

Sci. USA 108, 13564–13569.

Kolasinska-Zwierz, P., Down, T., Latorre, I., Liu, T., Liu, X.S., and Ahringer, J.

(2009). Differential chromatin marking of introns and expressed exons by

H3K36me3. Nat. Genet. 41, 376–381.

Kornblihtt, A.R. (2007). Coupling transcription and alternative splicing. Adv.

Exp. Med. Biol. 623, 175–189.

Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128,

693–705.

Kress, T.L., Krogan, N.J., and Guthrie, C. (2008). A single SR-like protein, Npl3,

promotes pre-mRNA splicing in budding yeast. Mol. Cell 32, 727–734.

La Via, L., Bonini, D., Russo, I., Orlandi, C., Barlati, S., and Barbon, A. (2013).

Modulation of dendritic AMPA receptor mRNA trafficking by RNA splicing and

editing. Nucleic Acids Res. 41, 617–631.

Lacadie, S.A., and Rosbash, M. (2005). Cotranscriptional spliceosome

assembly dynamics and the role of U1 snRNA:5’ss base pairing in yeast.

Mol. Cell 19, 65–75.

Lachner, M., O’Carroll, D., Rea, S., Mechtler, K., and Jenuwein, T. (2001).

Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins.

Nature 410, 116–120.

Langemeier, J., Schrom, E.-M., Rabner, A., Radtke, M., Zychlinski, D., Sabo-

rowski, A., Bohn, G., Mandel-Gutfreund, Y., Bodem, J., Klein, C., and Bohne,

J. (2012). A complex immunodeficiency is based on U1 snRNP-mediated

poly(A) site suppression. EMBO J. 31, 4035–4044.

Lareau, L.F., Brooks, A.N., Soergel, D.A., Meng, Q., and Brenner, S.E. (2007a).

The coupling of alternative splicing and nonsense-mediated mRNA decay.

Adv. Exp. Med. Biol. 623, 190–211.

Lareau, L.F., Inada, M., Green, R.E., Wengrod, J.C., and Brenner, S.E. (2007b).

Unproductive splicing of SR genes associatedwith highly conserved and ultra-

conserved DNA elements. Nature 446, 926–929.

Laurent, L., Wong, E., Li, G., Huynh, T., Tsirigos, A., Ong, C.T., Low, H.M., Kin

Sung, K.W., Rigoutsos, I., Loring, J., andWei, C.L. (2010). Dynamic changes in

the human methylome during differentiation. Genome Res. 20, 320–331.

Lawrence, J.B., Carter, K.C., and Xing, X. (1993). Probing functional organiza-

tion within the nucleus: is genome structure integrated with RNA metabolism?

Cold Spring Harb. Symp. Quant. Biol. 58, 807–818.

Lei, E.P., Krebber, H., and Silver, P.A. (2001). Messenger RNAs are recruited

for nuclear export during transcription. Genes Dev. 15, 1771–1782.

Lei, H., Dias, A.P., and Reed, R. (2011). Export and stability of naturally intron-

less mRNAs require specific coding region sequences and the TREX mRNA

export complex. Proc. Natl. Acad. Sci. USA 108, 17985–17990.

Licatalosi, D.D., and Darnell, R.B. (2010). RNA processing and its regulation:

global insights into biological networks. Nat. Rev. Genet. 11, 75–87.

Licatalosi, D.D., Mele, A., Fak, J.J., Ule, J., Kayikci, M., Chi, S.W., Clark, T.A.,

Schweitzer, A.C., Blume, J.E., Wang, X., et al. (2008). HITS-CLIP yields

genome-wide insights into brain alternative RNA processing. Nature 456,

464–469.

Lin, S., Coutinho-Mansfield, G., Wang, D., Pandit, S., and Fu, X.D. (2008). The

splicing factor SC35 has an active role in transcriptional elongation. Nat.

Struct. Mol. Biol. 15, 819–826.

Listerman, I., Sapra, A.K., and Neugebauer, K.M. (2006). Cotranscriptional

coupling of splicing factor recruitment and precursor messenger RNA splicing

in mammalian cells. Nat. Struct. Mol. Biol. 13, 815–822.

Long, J.C., and Caceres, J.F. (2009). The SR protein family of splicing factors:

master regulators of gene expression. Biochem. J. 417, 15–27.

Luco, R.F., Pan, Q., Tominaga, K., Blencowe, B.J., Pereira-Smith, O.M., and

Misteli, T. (2010). Regulation of alternative splicing by histone modifications.

Science 327, 996–1000.

Luco, R.F., Allo, M., Schor, I.E., Kornblihtt, A.R., and Misteli, T. (2011). Epige-

netics in alternative pre-mRNA splicing. Cell 144, 16–26.

Luger, K., Mader, A.W., Richmond, R.K., Sargent, D.F., and Richmond, T.J.

(1997). Crystal structure of the nucleosome core particle at 2.8 A resolution.

Nature 389, 251–260.

Lyko, F., Foret, S., Kucharski, R., Wolf, S., Falckenhayn, C., and Maleszka, R.

(2010). The honey bee epigenomes: differential methylation of brain DNA in

queens and workers. PLoS Biol. 8, e1000506.

Machyna, M., Heyn, P., and Neugebauer, K.M. (2013). Cajal bodies: where

form meets function. Wiley Interdiscip Rev RNA 4, 17–34.

Malik, S., and Roeder, R.G. (2010). The metazoan Mediator co-activator

complex as an integrative hub for transcriptional regulation. Nat. Rev. Genet.

11, 761–772.

Martinez-Contreras, R., Cloutier, P., Shkreta, L., Fisette, J.F., Revil, T., and

Chabot, B. (2007). hnRNP proteins and splicing control. Adv. Exp. Med. Biol.

623, 123–147.

Masuda, S., Das, R., Cheng, H., Hurt, E., Dorman, N., and Reed, R. (2005).

Recruitment of the human TREX complex to mRNA during splicing. Genes

Dev. 19, 1512–1517.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1267

Mayr, C., and Bartel, D.P. (2009). Widespread shortening of 3’UTRs by alterna-

tive cleavage and polyadenylation activates oncogenes in cancer cells. Cell

138, 673–684.

McCracken, S., Fong, N., Yankulov, K., Ballantyne, S., Pan, G., Greenblatt, J.,

Patterson, S.D., Wickens, M., and Bentley, D.L. (1997a). The C-terminal

domain of RNA polymerase II couples mRNA processing to transcription.

Nature 385, 357–361.

McCracken, S., Fong, N., Rosonina, E., Yankulov, K., Brothers, G., Siderovski,

D., Hessel, A., Foster, S., Shuman, S., and Bentley, D.L. (1997b). 50-Cappingenzymes are targeted to pre-mRNA by binding to the phosphorylated

carboxy-terminal domain of RNA polymerase II. Genes Dev. 11, 3306–3318.

McCracken, S., Lambermon, M., and Blencowe, B.J. (2002). SRm160 splicing

coactivator promotes transcript 30-end cleavage. Mol. Cell. Biol. 22, 148–160.

McManus, C.J., and Graveley, B.R. (2011). RNA structure and the mecha-

nisms of alternative splicing. Curr. Opin. Genet. Dev. 21, 373–379.

Millevoi, S., Geraghty, F., Idowu, B., Tam, J.L., Antoniou, M., and Vagner, S.

(2002). A novel function for the U2AF 65 splicing factor in promoting

pre-mRNA 30-end processing. EMBO Rep. 3, 869–874.

Millevoi, S., Loulergue, C., Dettwiler, S., Karaa, S.Z., Keller, W., Antoniou, M.,

and Vagner, S. (2006). An interaction between U2AF 65 and CF I(m) links the

splicing and 30 end processing machineries. EMBO J. 25, 4854–4864.

Mitrovich, Q.M., and Anderson, P. (2000). Unproductively spliced ribosomal

protein mRNAs are natural targets of mRNA surveillance in C. elegans. Genes

Dev. 14, 2173–2184.

Moehle, E.A., Ryan, C.J., Krogan, N.J., Kress, T.L., and Guthrie, C. (2012). The

yeast SR-like protein Npl3 links chromatin modification to mRNA processing.

PLoS Genet. 8, e1003101.

Monsalve, M., Wu, Z., Adelmant, G., Puigserver, P., Fan, M., and Spiegelman,

B.M. (2000). Direct coupling of transcription and mRNA processing through

the thermogenic coactivator PGC-1. Mol. Cell 6, 307–316.

Moore, M.J., and Proudfoot, N.J. (2009). Pre-mRNA processing reaches back

to transcription and ahead to translation. Cell 136, 688–700.

Motta-Mena, L.B., Heyd, F., and Lynch, K.W. (2010). Context-dependent regu-

latory mechanism of the splicing factor hnRNP L. Mol. Cell 37, 223–234.

Mukherjee, N., Corcoran, D.L., Nusbaum, J.D., Reid, D.W., Georgiev, S.,

Hafner, M., Ascano, M., Jr., Tuschl, T., Ohler, U., and Keene, J.D. (2011). Inte-

grative regulatory mapping indicates that the RNA-binding protein HuR

couples pre-mRNA processing and mRNA stability. Mol. Cell 43, 327–339.

Munoz,M.J., Perez Santangelo, M.S., Paronetto,M.P., de laMata, M., Pelisch,

F., Boireau, S., Glover-Cutter, K., Ben-Dov, C., Blaustein, M., Lozano, J.J.,

et al. (2009). DNA damage regulates alternative splicing through inhibition of

RNA polymerase II elongation. Cell 137, 708–720.

Naganuma, T., Nakagawa, S., Tanigawa, A., Sasaki, Y.F., Goshima, N., and

Hirose, T. (2012). Alternative 30-end processing of long noncoding RNA

initiates construction of nuclear paraspeckles. EMBO J. 31, 4020–4034.

Nakagawa, S., and Hirose, T. (2012). Paraspeckle nuclear bodies—useful

uselessness? Cell. Mol. Life Sci. 69, 3027–3036.

Nakagawa, S., Ip, J.Y., Shioi, G., Tripathi, V., Zong, X., Hirose, T., and

Prasanth, K.V. (2012). Malat1 is not an essential component of nuclear

speckles in mice. RNA 18, 1487–1499.

Neish, A.S., Anderson, S.F., Schlegel, B.P., Wei, W., and Parvin, J.D. (1998).

Factors associated with the mammalian RNA polymerase II holoenzyme. Nu-

cleic Acids Res. 26, 847–853.

Ni, J.Z., Grate, L., Donohue, J.P., Preston, C., Nobida, N., O’Brien, G., Shiue,

L., Clark, T.A., Blume, J.E., and Ares, M., Jr. (2007). Ultraconserved elements

are associated with homeostatic control of splicing regulators by alternative

splicing and nonsense-mediated decay. Genes Dev. 21, 708–718.

Nilsen, T.W., and Graveley, B.R. (2010). Expansion of the eukaryotic proteome

by alternative splicing. Nature 463, 457–463.

Ohlsson, R., Bartkuhn, M., and Renkawitz, R. (2010). CTCF shapes chromatin

by multiple mechanisms: the impact of 20 years of CTCF research on under-

standing the workings of chromatin. Chromosoma 119, 351–360.

1268 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Pan, Q., Saltzman, A.L., Kim, Y.K., Misquitta, C., Shai, O., Maquat, L.E., Frey,

B.J., and Blencowe, B.J. (2006). Quantitative microarray profiling provides

evidence against widespread coupling of alternative splicing with nonsense-

mediated mRNA decay to control gene expression. Genes Dev. 20, 153–158.

Pan, Q., Shai, O., Lee, L.J., Frey, B.J., and Blencowe, B.J. (2008). Deep

surveying of alternative splicing complexity in the human transcriptome by

high-throughput sequencing. Nat. Genet. 40, 1413–1415.

Pandit, S., Wang, D., and Fu, X.D. (2008). Functional integration of transcrip-

tional and RNA processing machineries. Curr. Opin. Cell Biol. 20, 260–265.

Park, J.W., Parisky, K., Celotto, A.M., Reenan, R.A., and Graveley, B.R. (2004).

Identification of alternative splicing regulators by RNA interference in

Drosophila. Proc. Natl. Acad. Sci. USA 101, 15974–15979.

Pavri, R., Zhu, B., Li, G., Trojer, P., Mandal, S., Shilatifard, A., and Reinberg, D.

(2006). Histone H2B monoubiquitination functions cooperatively with FACT to

regulate elongation by RNA polymerase II. Cell 125, 703–717.

Plocik, A.M., and Guthrie, C. (2012). Diverse forms of RPS9 splicing are part of

an evolving autoregulatory circuit. PLoS Genet. 8, e1002620.

Pradeepa, M.M., Sutherland, H.G., Ule, J., Grimes, G.R., and Bickmore, W.A.

(2012). Psip1/Ledgf p52 binds methylated histone H3K36 and splicing factors

and contributes to the regulation of alternative splicing. PLoS Genet. 8,

e1002717.

Proudfoot, N.J. (2011). Ending the message: poly(A) signals then and now.

Genes Dev. 25, 1770–1782.

Rebbapragada, I., and Lykke-Andersen, J. (2009). Execution of nonsense-

mediated mRNA decay: what defines a substrate? Curr. Opin. Cell Biol. 21,

394–402.

Rosonina, E., and Blencowe, B.J. (2004). Analysis of the requirement for RNA

polymerase II CTD heptapeptide repeats in pre-mRNA splicing and 30-endcleavage. RNA 10, 581–589.

Rosonina, E., Ip, J.Y., Calarco, J.A., Bakowski, M.A., Emili, A., McCracken, S.,

Tucker, P., Ingles, C.J., and Blencowe, B.J. (2005). Role for PSF in mediating

transcriptional activator-dependent stimulation of pre-mRNA processing

in vivo. Mol. Cell. Biol. 25, 6734–6746.

Saint-Andre, V., Batsche, E., Rachez, C., and Muchardt, C. (2011). Histone H3

lysine 9 trimethylation and HP1g favor inclusion of alternative exons. Nat.

Struct. Mol. Biol. 18, 337–344.

Saltzman, A.L., Kim, Y.K., Pan, Q., Fagnani, M.M., Maquat, L.E., and

Blencowe, B.J. (2008). Regulation of multiple core spliceosomal proteins by

alternative splicing-coupled nonsense-mediated mRNA decay. Mol. Cell.

Biol. 28, 4320–4330.

Saltzman, A.L., Pan, Q., and Blencowe, B.J. (2011). Regulation of alternative

splicing by the core spliceosomal machinery. Genes Dev. 25, 373–384.

Sandberg, R., Neilson, J.R., Sarma, A., Sharp, P.A., and Burge, C.B. (2008).

Proliferating cells express mRNAs with shortened 30 untranslated regions

and fewer microRNA target sites. Science 320, 1643–1647.

Schneider, M., Will, C.L., Anokhina, M., Tazi, J., Urlaub, H., and Luhrmann, R.

(2010). Exon definition complexes contain the tri-snRNP and can be directly

converted into B-like precatalytic splicing complexes. Mol. Cell 38, 223–235.

Schoenberg, D.R., and Maquat, L.E. (2012). Regulation of cytoplasmic mRNA

decay. Nat. Rev. Genet. 13, 246–259.

Schor, I.E., Lleres, D., Risso, G.J., Pawellek, A., Ule, J., Lamond, A.I., and

Kornblihtt, A.R. (2012). Perturbation of chromatin structure globally affects

localization and recruitment of splicing factors. PLoS ONE 7, e48084.

Schwartz, S., Meshorer, E., and Ast, G. (2009). Chromatin organization marks

exon-intron structure. Nat. Struct. Mol. Biol. 16, 990–995.

Shukla, S., Kavak, E., Gregory, M., Imashimizu,M., Shutinoski, B., Kashlev,M.,

Oberdoerffer, P., Sandberg, R., and Oberdoerffer, S. (2011). CTCF-

promoted RNA polymerase II pausing links DNA methylation to splicing.

Nature 479, 74–79.

Sims, R.J., 3rd, Millhouse, S., Chen, C.-F., Lewis, B.A., Erdjument-Bromage,

H., Tempst, P., Manley, J.L., and Reinberg, D. (2007). Recognition of

trimethylated histone H3 lysine 4 facilitates the recruitment of transcription

postinitiation factors and pre-mRNA splicing. Mol. Cell 28, 665–676.

Singh, G., Kucukural, A., Cenik, C., Leszyk, J.D., Shaffer, S.A., Weng, Z., and

Moore, M.J. (2012). The cellular EJC interactome reveals higher-order mRNP

structure and an EJC-SR protein nexus. Cell 151, 750–764.

Smallwood, A., Hon, G.C., Jin, F., Henry, R.E., Espinosa, J.M., and Ren, B.

(2012). CBX3 regulates efficient RNA processing genome-wide. Genome

Res. 22, 1426–1436.

Smith, K.P., Carter, K.C., Johnson, C.V., and Lawrence, J.B. (1995). U2 and U1

snRNA gene loci associate with coiled bodies. J. Cell. Biochem. 59, 473–485.

Smolle, M., andWorkman, J.L. (2013). Transcription-associated histone modi-

fications and cryptic transcription. Biochim. Biophys. Acta 1829, 84–97.

Spector, D.L., and Lamond, A.I. (2011). Nuclear speckles. Cold Spring Harb.

Perspect. Biol. 3.

Spies, N., Nielsen, C.B., Padgett, R.A., and Burge, C.B. (2009). Biased

chromatin signatures around polyadenylation sites and exons. Mol. Cell 36,

245–254.

Strasser, K., Masuda, S., Mason, P., Pfannstiel, J., Oppizzi, M., Rodriguez-

Navarro, S., Rondon, A.G., Aguilera, A., Struhl, K., Reed, R., and Hurt, E.

(2002). TREX is a conserved complex coupling transcription with messenger

RNA export. Nature 417, 304–308.

Stutz, F., Bachi, A., Doerks, T., Braun, I.C., Seraphin, B., Wilm, M., Bork, P.,

and Izaurralde, E. (2000). REF, an evolutionary conserved family of hnRNP-

like proteins, interacts with TAP/Mex67p and participates in mRNA nuclear

export. RNA 6, 638–650.

Talbert, P.B., and Henikoff, S. (2010). Histone variants—ancient wrap artists of

the epigenome. Nat. Rev. Mol. Cell Biol. 11, 264–275.

Tange, T.O., Shibuya, T., Jurica, M.S., and Moore, M.J. (2005). Biochemical

analysis of the EJC reveals two new factors and a stable tetrameric protein

core. RNA 11, 1869–1883.

Tardiff, D.F., Lacadie, S.A., and Rosbash, M. (2006). A genome-wide analysis

indicates that yeast pre-mRNA splicing is predominantly posttranscriptional.

Mol. Cell 24, 917–929.

Terenzi, F., and Ladd, A.N. (2010). Conserved developmental alternative

splicing of muscleblind-like (MBNL) transcripts regulates MBNL localization

and activity. RNA Biol. 7, 43–55.

Tian, B., Hu, J., Zhang, H., and Lutz, C.S. (2005). A large-scale analysis of

mRNA polyadenylation of human and mouse genes. Nucleic Acids Res. 33,

201–212.

Tilgner, H., Nikolaou, C., Althammer, S., Sammeth, M., Beato,M., Valcarcel, J.,

and Guigo, R. (2009). Nucleosome positioning as a determinant of exon

recognition. Nat. Struct. Mol. Biol. 16, 996–1001.

Tilgner, H., Knowles, D.G., Johnson, R., Davis, C.A., Chakrabortty, S., Djebali,

S., Curado, J., Snyder, M., Gingeras, T.R., and Guigo, R. (2012). Deep

sequencing of subcellular RNA fractions shows splicing to be predominantly

co-transcriptional in the human genome but inefficient for lncRNAs. Genome

Res. 22, 1616–1625.

Tillo, D., and Hughes, T.R. (2009). G+C content dominates intrinsic nucleo-

some occupancy. BMC Bioinformatics 10, 442.

Tolstorukov, M.Y., Goldman, J.A., Gilbert, C., Ogryzko, V., Kingston, R.E., and

Park, P.J. (2012). Histone variant H2A.Bbd is associated with active transcrip-

tion and mRNA processing in human cells. Mol. Cell 47, 596–607.

Tripathi, V., Ellis, J.D., Shen, Z., Song, D.Y., Pan, Q., Watt, A.T., Freier, S.M.,

Bennett, C.F., Sharma, A., Bubulya, P.A., et al. (2010). The nuclear-retained

noncoding RNA MALAT1 regulates alternative splicing by modulating SR

splicing factor phosphorylation. Mol. Cell 39, 925–938.

Vargas, D.Y., Shah, K., Batish, M., Levandoski, M., Sinha, S., Marras, S.A.,

Schedl, P., and Tyagi, S. (2011). Single-molecule imaging of transcriptionally

coupled and uncoupled splicing. Cell 147, 1054–1065.

Wagner, E.J., and Carpenter, P.B. (2012). Understanding the language of

Lys36 methylation at histone H3. Nat. Rev. Mol. Cell Biol. 13, 115–126.

Wahl, M.C., Will, C.L., and Luhrmann, R. (2009). The spliceosome: design

principles of a dynamic RNP machine. Cell 136, 701–718.

Waldholm, J., Wang, Z., Brodin, D., Tyagi, A., Yu, S., Theopold, U., Farrants,

A.K., and Visa, N. (2011). SWI/SNF regulates the alternative processing of

a specific subset of pre-mRNAs in Drosophila melanogaster. BMC Mol. Biol.

12, 46.

Wang, Z., and Burge, C.B. (2008). Splicing regulation: from a parts list of

regulatory elements to an integrated splicing code. RNA 14, 802–813.

Wang, E.T., Sandberg, R., Luo, S., Khrebtukova, I., Zhang, L., Mayr, C.,

Kingsmore, S.F., Schroth, G.P., and Burge, C.B. (2008). Alternative isoform

regulation in human tissue transcriptomes. Nature 456, 470–476.

Wang, E.T., Cody, N.A.L., Jog, S., Biancolella, M., Wang, T.T., Treacy, D.J.,

Luo, S., Schroth, G.P., Housman, D.E., Reddy, S., et al. (2012).

Transcriptome-wide regulation of pre-mRNA splicing and mRNA localization

by muscleblind proteins. Cell 150, 710–724.

Weatheritt, R.J., and Gibson, T.J. (2012). Linear motifs: lost in (pre)translation.

Trends Biochem. Sci. 37, 333–341.

Wilhelm, B.T., Marguerat, S., Aligianni, S., Codlin, S., Watt, S., and Bahler, J.

(2011). Differential patterns of intronic and exonic DNA regions with respect

to RNA polymerase II occupancy, nucleosome density and H3K36me3

marking in fission yeast. Genome Biol. 12, R82.

Wilson, C.J., Chao, D.M., Imbalzano, A.N., Schnitzler, G.R., Kingston, R.E.,

and Young, R.A. (1996). RNA polymerase II holoenzyme contains SWI/SNF

regulators involved in chromatin remodeling. Cell 84, 235–244.

Witten, J.T., and Ule, J. (2011). Understanding splicing regulation through RNA

splicing maps. Trends Genet. 27, 89–97.

Wu, H., and Zhang, Y. (2011). Mechanisms and functions of Tet protein-

mediated 5-methylcytosine oxidation. Genes Dev. 25, 2436–2452.

Yap, K.L., and Zhou, M.M. (2010). Keeping it in the family: diverse histone

recognition by conserved structural folds. Crit. Rev. Biochem. Mol. Biol. 45,

488–505.

Yap, K., Lim, Z.Q., Khandelia, P., Friedman, B., and Makeyev, E.V. (2012).

Coordinated regulation of neuronal mRNA steady-state levels through

developmentally controlled intron retention. Genes Dev. 26, 1209–1223.

Yin, Q.-F., Yang, L., Zhang, Y., Xiang, J.-F., Wu, Y.-W., Carmichael, G.G., and

Chen, L.-L. (2012). Long noncoding RNAs with snoRNA ends. Mol. Cell 48,

219–230.

Zeng, C., and Berget, S.M. (2000). Participation of the C-terminal domain of

RNA polymerase II in exon definition during pre-mRNA splicing. Mol. Cell.

Biol. 20, 8290–8301.

Zhang, C., Frias, M.A., Mele, A., Ruggiu, M., Eom, T., Marney, C.B., Wang, H.,

Licatalosi, D.D., Fak, J.J., and Darnell, R.B. (2010). Integrative modeling

defines the Nova splicing-regulatory network and its combinatorial controls.

Science 329, 439–443.

Zhang, B., Arun, G., Mao, Y.S., Lazar, Z., Hung, G., Bhattacharjee, G., Xiao, X.,

Booth, C.J., Wu, J., Zhang, C., and Spector, D.L. (2012). The lncRNAMalat1 is

dispensable for mouse development but its transcription plays a cis-regulatory

role in the adult. Cell Rep 2, 111–123.

Zhou, Z., Luo, M.J., Straesser, K., Katahira, J., Hurt, E., and Reed, R. (2000).

The protein Aly links pre-messenger-RNA splicing to nuclear export in meta-

zoans. Nature 407, 401–405.

Zhou, H.L., Hinman, M.N., Barron, V.A., Geng, C., Zhou, G., Luo, G., Siegel,

R.E., and Lou, H. (2011). Hu proteins regulate alternative splicing by inducing

localized histone hyperacetylation in an RNA-dependent manner. Proc. Natl.

Acad. Sci. USA 108, E627–E635.

Zraly, C.B., and Dingwall, A.K. (2012). The chromatin remodeling and mRNA

splicing functions of the Brahma (SWI/SNF) complex are mediated by the

SNR1/SNF5 regulatory subunit. Nucleic Acids Res. 40, 5975–5987.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1269

Leading Edge

Review

Genome Architecture: DomainOrganization of Interphase Chromosomes

Wendy A. Bickmore1,* and Bas van Steensel2,*1MRC Human Genetics Unit, MRC Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh EH42XU, UK2Division of Gene Regulation, Netherlands Cancer Institute, 1066 CX Amsterdam, the Netherlands*Correspondence: wendy.bickmore@igmm.ed.ac.uk (W.A.B.), b.v.steensel@nki.nl (B.v.S.)

http://dx.doi.org/10.1016/j.cell.2013.02.001

The architecture of interphase chromosomes is important for the regulation of gene expression andgenome maintenance. Chromosomes are linearly segmented into hundreds of domains withdifferent protein compositions. Furthermore, the spatial organization of chromosomes isnonrandom and is characterized by many local and long-range contacts among genes and othersequence elements. A variety of genome-wide mapping techniques have made it possible to chartthese properties at high resolution. Combined with microscopy and computational modeling, theresults begin to yield a more coherent picture that integrates linear and three-dimensional (3D)views of chromosome organization in relation to gene regulation and other nuclear functions.

IntroductionThe idea that chromosomes are segmented into domains with

distinct functional properties goes back to the initial microscopy

observations of heterochromatin and euchromatin and the band-

ing patterns of mitotic chromosomes and polytene interphase

chromosomes upon staining with particular dyes. Later, immu-

nofluorescence microscopy with antibodies against specific

chromatin proteins led to the notion that differential protein

composition may underlie this segmentation. The development

of chromatin immunoprecipitation (ChIP) subsequently enabled

the locations of specific proteins or histone modifications to be

mapped along the genome with much higher resolution. This

provided direct evidence that some proteins can associate

with long genomic regions, such as the Polycomb protein at

the homeotic bithorax locus in Drosophila (Orlando and Paro,

1993). Further refinement of ChIP and the developments of the

complementary mapping technique DamID (van Steensel et al.,

2001) and methods for the mapping of various other properties

of chromatin (see below) have led to the generation of numerous

high-resolution genome-wide maps that identify various chro-

matin features with a domain-like organization.

Parallel to these developments, the three-dimensional (3D)

folding of chromosomes has been investigated by fluores-

cence in situ hybridization (FISH). Extensive microscopy

studies have revealed a high degree of nonrandom positioning

of loci within the nucleus and within chromosomal territories

(Cremer et al., 2006). More recently, chromosome conforma-

tion capture (3C) techniques have begun to offer detailed views

of the associations among linearly distant genomic loci that

can be captured by formaldehyde crosslinking in the nucleus,

and from which aspects of 3D chromosomal folding have

been inferred.

Excitingly, these linear and 3D views of interphase chromo-

some architecture are now beginning to converge, revealing

that the chromatin modifications of genomic regions and the

1270 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

overall 3D organization are linked. Here, we will discuss current

insights in the relationships between the linear and 3D domain

architecture of interphase chromosomes, with emphasis on

results obtained in mammals and in Drosophila.

Linear Views of Chromatin DomainsDNA-Sequence Organization

The genomes ofmost species show striking nonrandompatterns

in their sequence composition. Mammalian and bird genomes

are a patchwork of long DNA stretches (>100 kb, Figure 1A)

termed isochores, which differ in A/T content (Costantini et al.,

2006). Gene density is also highly nonhomogeneous along the

genome and roughly corresponds to isochore patterning, with

high gene density overlapping with A/T-poor regions. Interest-

ingly, genes embedded in gene-dense regions tend to be more

active than those in gene-poor regions, resulting in megabase-

sized domains of alternating high and low transcriptional activity

(Caron et al., 2001). Most transposons and virus-derived

elements are also nonevenly distributed. For example, of the

most abundant repetitive elements in the human genome,

SINE elements tend to be located in gene-dense regions,

whereas LINE elements are more abundant in gene-poor

regions. As will become clear below, the nonrandom distribution

of DNA-sequence features is closely linked to chromatin-domain

organization, although it is unknown how the former helps to

establish the latter. Various satellite repeats constitute very

prominent domains, but it is beyond the scope of this Review

to discuss them here.

Polycomb Domains

Perhaps the best studied chromatin domains are those formed

in Drosophila by the Polycomb group (PcG) proteins, known

for their role in maintaining silencing of gene clusters during

development. These proteins form multisubunit complexes

of which the most prominent are Polycomb-repressive com-

plexes 1 and 2 (PRC1 and PRC2). PRC2 contains a histone

Figure 1. Chromatin Types(A) Size distribution (in base pairs) of some genome features (blue) and various types of chromatin domains in human fibroblasts (red). Numbers on the righthandside indicate approximate genome-wide counts of the respective domains. Note that the sizes and counts of chromatin domains can vary between cell types andcan depend much on the algorithm used to define the domains. (Data are from http://genome.ucsc.edu; Costantini et al., 2006; Guelen et al., 2008; Lister et al.,2009; Hawkins et al., 2010; Dixon et al., 2012; Pope et al., 2012.)(B) Cartoon model of a chromosomal fiber, illustrating its segmentation into domains of distinct chromatin types, each consisting of a specific combination ofproteins and histone modifications (indicated by colors).(C) Classification of chromatin types. This example shows bindingmaps of 12 chromatin proteins along a part of chromosome 2L inDrosophilaKc cells. Computeralgorithms are used to search for recurrent combinations of proteins (chromatin ‘‘types’’ or ‘‘states’’) and to subsequently define linear chromosomal domainscovered by these types (highlighted in different colors; the classification in this example was based on 53 protein profiles). Note that some proteins are present ina single chromatin type, whereas others can be shared among multiple types. Adapted from Filion and van Steensel (2010).

methyltransferase that catalyzes trimethylation of H3K27

(H3K27me3), whereas PRC1 includes Polycomb (Pc), which

binds to this histone mark (Morey and Helin, 2010).

In Drosophila, PcG proteins and H3K27me3 form a few

hundred domains, of about 10 to 150 kb in size, that are scat-

tered along the genome. These domains often cover multiple

genes (Tolhuis et al., 2006; Schwartz et al., 2010), most of which

are transcriptionally inactive. Specialized Polycomb response

elements (PREs) of several hundred base pairs in size can act

as nucleation sites from which the PcG complexes spread later-

ally to form the domains (Morey and Helin, 2010). Genes within

a PcG domain are generally not coregulated throughout devel-

opment, although exceptions to this rule may occur (Tolhuis

et al., 2006). Rather, PcG domains are dynamic structures that,

depending on the cell type, can be partially or entirely cleared

of PcG proteins to accommodate expression of one or several

of the underlying genes (Schwartz et al., 2010).

In mouse and human, most available evidence so far points to

the existence of only a few large PcGdomains that covermultiple

neighboring genes, primarily at the Hox gene clusters (Bernstein

et al., 2006) and on the inactive X chromosome in female cells

(Marks et al., 2009). Otherwise, PcG domains are relatively small

(�10 kb) (Figure 1A) and usually overlap with individual CpG-rich

promoter regions (Lee et al., 2006; Ku et al., 2008). However, one

ChIP study (Pauler et al., 2009) suggests that besides the small

H3K27me3 peaks, there may be hundreds of larger regions

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1271

(average �40 kb) with milder but significant H3K27me3 enrich-

ment in mouse embryonic fibroblasts.

Like in Drosophila, mammalian PcG target genes are highly

enriched in genes that have regulatory functions in development

(Lee et al., 2006). In embryonic stem cells (ESCs), many of these

sites are also marked by H3K4me3, a state that is referred to as

‘‘bivalent.’’ Like H3K27me3/PRC2, this H3K4me3 is required for

differentiation rather than the stem cell state per se (Jiang et al.,

2011). Upon ESC differentiation, the bivalent state typically

resolves into either an active H3K4me3-only or a repressed

H3K27me3-only state (Bernstein et al., 2006). In case of the

latter, the H3K27me3 domains expand 2- to 3-fold in size once

ESCs differentiate (Hawkins et al., 2010).

Interestingly, ectopically integrated unmethylated CpG-rich

elements can recruit PRC2 (Mendenhall et al., 2010), indicating

that they harbor the necessary sequence information for setting

up a small PcG domain. However, most CpG islands lack

H3K27me3, so additional local cues must modulate this ability.

Specific transcription factors seem to have a role in PcG recruit-

ment (Arnold et al., 2013), whereas activating signals might

counteract the formation of a PcG domain at CpG islands (Men-

denhall et al., 2010).

H3K9me2/3 Chromatin

H3K9me2 and H3K9me3 are abundant histone marks produced

by several enzymes; in mammals, these include G9a/GLP,

SETDB1, and SUV39H1. Various proteins bind specifically to

methylated H3K9, each with its own preference for mono-, di-,

or trimethylation. The most extensively studied among these

are theHP1 proteins. Together with H3K9me2/3, they cover peri-

centric and telomeric regions of many species, where they have

structural roles, suppress recombination, and silence transpos-

able elements (Zeng et al., 2010). In Drosophila, H3K9me2 and

HP1 additionally associate with a few hundred genes throughout

the genome, where they are often confined to individual tran-

scription units. Although H3K9me2 and H3K9me3 are widely

believed to be repressive marks, there are many examples of

transcriptionally active genes that are covered by these marks

and by HP1 (Kwon and Workman, 2011).

In mammals, there is now evidence for extensive formation of

chromatin domains by H3K9me2/3 and HP1s. As well as binding

to methylated H3K9, ectopic recruitment of HP1a to a specific

site is sufficient in itself to establish an H3K9me3 domain of

about 10 kb (Hathaway et al., 2012).Thousands of endogenous

H3K9me2 domains were found along the genome in mouse cells

and tissues (Wen et al., 2009). These domains, referred to as

large organized chromatin K9 modifications (LOCKs), have

a size of roughly 100 kb and together cover up to 45% of the

genome, depending on the cell type. It was claimed that mouse

ESCs (mESCs) mostly lack these domains, but the statistical

significance of this observation was questioned (Filion and van

Steensel, 2010), and an independent study found only marginal

differences in the H3K9me2 domain patterns between mESCs

and differentiated cells (Lienert et al., 2011). In human ESCs

(hESCs), H3K9me3 forms more than ten thousand small

domains (median size 7 kb). These tend to be about 2-fold larger

in fibroblasts (Hawkins et al., 2010) (Figure 1A), which could point

to the ability of this chromatin type to spread in cis. Analysis of

H3K9me2 in human lymphocytes suggests a global pattern of

1272 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Mb-sized domains (Ryba et al., 2010). No study so far has care-

fully compared H3K9me2 and H3K9me3 patterns, so it is unclear

to what extent these two marks overlap.

A striking example of the H3K9me3 chromatin type is formed

by the clusters of KRAB-ZNF genes, which encode the largest

mammalian family of transcription factors. These clusters,

some of which are more than 1 Mb in size, are extensively

covered by H3K9me3, CBX1/HP1b, SETDB1, and SUV39H1

(Vogel et al., 2006; Frietze et al., 2010). However, these compo-

nents are not homogenously distributed within the domains, as

they show higher occupancy near the 30 ends of the KRAB-

ZNF genes. Also present at these positions is the repressor

protein KAP1, which is known to interact with HP1 proteins

and SETDB1 (Frietze et al., 2010; Groner et al., 2010). Artificial

targeting of KAP1 to genomic loci can trigger recruitment and

cis-spreading of HP1b and H3K9me3, often over several tens

of kb (Groner et al., 2010), pointing to a role of KAP1 in the nucle-

ation of at least some H3K9me3 domains.

Replication-Timing Domains

Timing of DNA replication during S phase is tightly controlled.

High-resolution genome mapping shows �100 kb to several

Mb-sized (Figure 1A) alternating segments of early and late repli-

cating DNA (Hiratani et al., 2008; Schwaiger et al., 2009). Such

domains are thought to contain clusters of synchronously firing

origins. A finer temporal sampling through S phase, however,

suggests that a binary early versus late classification may be

overly simplistic and should eventually be replaced with

a more continuous scale (Hansen et al., 2010). Nevertheless,

early-replicating domains generally contain active genes,

whereas late-replicating domains are mostly transcriptionally

silent (Hiratani et al., 2008; Schwaiger et al., 2009).

For about 50% of the genome, the timing of replication is

dependent on cell type and usually linked to the transcription

status of the region (Schwaiger et al., 2009; Hansen et al.,

2010; Ryba et al., 2010). Human late-replicating regions are

depleted of various active histone marks and enriched for

H3K9me2, but they are not enriched for H3K27me3, indicating

the different characteristics of inactive domains that are estab-

lished by distinct mechanisms (Ryba et al., 2010). Late-repli-

cating domains also correlate with lamina-associated domains

(LADs), which are regions that are preferentially located at the

nuclear periphery (see below). During differentiation, there is

a consolidation of replication-timing domains to fewer larger

domains (Ryba et al., 2010).

Intriguing insights into the influence of genome sequence and

genomic context on replication-timing domains come from the

behavior of a human chromosome 21 in mouse cells (Pope

et al., 2012), but how the regional coordination of replication

timing is established is still largely unclear. Depletion of HP1 in

Drosophila was found to cause both premature as well as de-

layed replication of large HP1-bound domains, depending on

the genomic location (Schwaiger et al., 2010). In mESCs, dele-

tion of G9a leads to substantial global loss of H3K9me2 but

does not detectably affect replication timing (Yokochi et al.,

2009), indicating that late replication is not a direct result of the

presence of H3K9me2. However, H3K9me3 is still present in

these cells and may be redundant with H3K9me2 in the regula-

tion of replication timing. Systematic studies are needed to

determine the roles of chromatin components in the regional

timing of replication.

DNA Methylation

DNA cytosine methylation in mammals occurs predominantly at

CG dinucleotides, although bisulfite sequencing of the genome

fromhESCs andmouse brain has also detected substantial cyto-

sine methylation in other sequence contexts (Lister et al., 2009;

Xie et al., 2012). CG methylation in hESCs and other cells in vivo

is ubiquitous, with nearly 80% of the CG dinucleotides

throughout the genome showing a methylation frequency above

80%. This contrasts strongly with the mostly unmethylated state

of focal CpG islands (approximately 1 kb in size, Figure 1A). In

general, there is a strong anticorrelation between Polycomb

and DNA methylation in such cells. In contrast, cultured human

somatic cell lines exhibit numerous so-called partiallymethylated

domains (PMDs) that exhibit lower CG methylation frequencies

and have a median size of�50 kb (Lister et al., 2009) (Figure 1A).

Genes in PMDs tend to be downregulated and, in some cases,

show increased levels of H3K27me3. Highly similar hypomethy-

lated PMDs were also observed in human colon cancers, but

not in the adjacent normal tissue (Hansen et al., 2011; Berman

et al., 2012). Furthermore, it was noted that these PMDs tend to

overlap with LADs and domains of H3K9me2 enrichment (Han-

sen et al., 2011; Berman et al., 2012), although the latter domains

weremapped in different cell types. Therefore, in somemamma-

liancell linesand tumors,but apparently not innormal cells in vivo,

the globally high DNA methylation is interrupted by domains of

reduced methylation levels that tend to coincide with repressive

domains marked by H3K9me2 and nuclear lamina interactions.

Inferring 3D Organization from Linear Maps: LADs

and NADs

A special class of chromatin domains is formed by genomic

regions that interact with relatively fixed nuclear structures. In

particular, the nuclear lamina (NL) has been implicated in the

anchoring of chromosomal domains. The NL covers the nucleo-

plasmic side of the inner nuclear membrane (INM) and consists

of lamin proteins that form long polymers (Prokocimer et al.,

2009). Maps of genome–NL interactions provide insight into

the overall spatial organization of interphase chromosomes. In

mammalian cells, DamID was used to identify about 1,300

LADs that contact lamin B1 (Guelen et al., 2008). These domains

are large, about 100 kb–10 Mb (median size of �0.5 Mb,

Figure 1A), and collectively cover nearly 40% of the genome. In

part, the NL interaction pattern is cell type specific (Peric-

Hupkes et al., 2010). A domain pattern of interactions with the

NL was also observed in flies and worms (Pickersgill et al.,

2006; Gerstein et al., 2010; van Bemmel et al., 2010).

The majority of genes located in LADs are transcriptionally

inactive, indicating that the NL constitutes a repressive environ-

ment (Pickersgill et al., 2006; Guelen et al., 2008; Gerstein et al.,

2010; Peric-Hupkes et al., 2010; van Bemmel et al., 2010).

Indeed, artificial tethering of a locus to the NL or INM can lead

to reduced gene expression (Finlan et al., 2008; Reddy et al.,

2008; Dialynas et al., 2010), although not in all instances (Ku-

maran and Spector, 2008).

How LADs are targeted to the NL is still poorly understood. In

part, it may involve NL-associated DNA-binding proteins that

recognize specific sequence features. A recent study suggested

that (GA)n repeats can target certain human LADs to the NL (Zullo

et al., 2012). However, a systematic genome-wide survey of

repeats did not find (GA)n repeats to be enriched in LADs (Guelen

et al., 2008), so this mechanism must be relatively rare. In

mammals, constitutive LADs (i.e., LADs shared among cell

types) show a striking overlap with A/T-rich isochores, suggest-

ing a role for long stretches of A/T-rich DNA in NL targeting (Meu-

leman et al., 2012). In C. elegans, it was found that two H3K9

methyltransferases, MET-2 and SET-25, together promote the

peripheral localization and silencing of a transgene repeat (Tow-

bin et al., 2012). A combined knockout of the two enzymes also

caused a partial loss of NL interactions genome-wide. Thus, both

DNA sequences and chromatin modifications can drive NL

contacts of LADs.

The nucleolus appears to be another platform for organizing

the genome. Two groups have identified human DNA sequences

in the chromatin associated with purified nucleoli (Nemeth et al.,

2010; van Koningsbruggen et al., 2010). Besides the expected

ribosomal DNA (rDNA) loci, a domain-like interaction pattern

was observed across all chromosomes. These nucleolus-asso-

ciated domains (NADs) preferentially contain repressed genes

and show enrichment for repressive histone marks, in particular

H3K9me3.

Surprisingly, LAD and NAD patterns in human cells overlap

substantially (Nemeth et al., 2010; van Koningsbruggen et al.,

2010), although comparisons so far are based on data from

different cell types. It is possible that LADs and NADs in part

consist of the same type of repressive chromatin that distributes

between the NL and nucleoli in a random manner. This model is

supported by microscopy observations that some chromosomal

regions associated with a nucleolus in amother cell can be repo-

sitioned to the nuclear periphery in the daughter cells after

mitosis (Thomson et al., 2004; van Koningsbruggen et al., 2010).

Nuclear pore proteins (Nups) also interact with specific

genomic loci. One study in Drosophila found many 10–500 kb

domains to be bound by Nups (Vaquerizas et al., 2010), but

a parallel study reported mostly narrow peaks of binding (Kal-

verda et al., 2010). Because Nups freely roam through the nucle-

oplasm, most of these binding events occur in the nuclear

interior (Capelson et al., 2010; Kalverda et al., 2010) and thus

provide only limited information on the spatial organization of

chromosomes.

Integrative Approaches to Classify Chromatin Domains

The overview above illustrates that chromatin domains are prev-

alent along metazoan genomes. In addition, the distribution

patterns of different markers often correlate, indicating that

multiple chromatin components work together in the same

genomic regions. This raises a number of important questions.

Is the genome segmented into domains of a limited number of

types (or ‘‘states’’), defined by recurrent combinations of DNA

sequences, proteins, and histone marks? What are these chro-

matin types and what are their functions? Several laboratories

have begun to address these questions by collecting large sets

of genome-wide chromatin maps and using computational

approaches to identify chromatin types and to study their

domain patterns (Figures 1B and 1C).

A survey of 53 broadly selected chromatin proteins in

Drosophila Kc cells defined five principal chromatin types that

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1273

segment the genome into domains that frequently span multiple

neighboring genes and that consist of specific combinations of

proteins (Filion et al., 2010). Some proteins mark only a single

chromatin type, whereas others are shared among multiple

types. The five chromatin types are the HP1 and Polycomb chro-

matin types; two distinct types of transcriptionally active chro-

matin that harbor different sets of genes; and BLACK chromatin,

a strongly repressive chromatin type that lacks HP1 or Polycomb

proteins. Even though BLACK chromatin covers nearly two-

thirds of all repressed genes, it is still largely uncharacterized.

Among the proteins that make up BLACK chromatin are histone

H1, the AT-hook protein D1, and SU(UR), which is a regulator of

late replication. Also present is LAM, the sole B-type Lamin in

Drosophila, indicating that BLACK chromatin is preferentially

located at the nuclear periphery.

In another study in male Drosophila S2 cells, integration of 18

histone modification maps yielded a segmentation into nine

chromatin types (Kharchenko et al., 2011). One of these is linked

to the dosage-compensation complex that is specific to themale

X chromosome, whereas the remaining eight states appear to

represent a somewhat finer subdivision of the five states

mentioned above. It is noteworthy that the prevailing state in

S2 cells lacks all of the mapped histone marks and may largely

correspond to the BLACK state in Kc cells. For a complete

picture, it is thus essential to include nonhistone proteins in inte-

grative approaches such as these.

Similar systematic mapping efforts in human cell lines have led

to classifications of a highly variable number of chromatin states,

ranging from 6 to as many as 51 (Ernst and Kellis, 2010; Ram

et al., 2011; Dunham et al., 2012). However, mostly small, sub-

genic domains were identified, not the large LADs, NADs,

LOCKs, or replication-timing domains—because good markers

of these latter domain types were mostly missing from these

studies. Because there are still hundreds of chromatin compo-

nents that have not been mapped, and algorithms to categorize

chromatin types are still being refined, the classification of chro-

matin domains should presently be regarded as ‘‘work-in-prog-

ress.’’ Nevertheless, the results of these efforts provide a useful

framework to think about the diversity of chromatin domains and

their functions.

Insulators and Linear Chromatin Domains

Insulator elements were originally identified based on their ability

to block activation of a promoter by a distant enhancer when in-

serted between these elements. This enhancer-blocking activity

is mediated by proteins/complexes that bind specifically to insu-

lator element sequences (Vogelmann et al., 2011). Later it was

recognized that these proteins may also help to delimit chro-

matin domains. In Drosophila, the five known insulator-binding

proteins—SU(HW), dCTCF, GAF, BEAF32, and ZW-5—each

have several thousand partially overlapping focal binding sites

in the genome (Negre et al., 2010; van Bemmel et al., 2010;

Schwartz et al., 2012). Interestingly, most exhibit a significant

enrichment at the borders of H3K27me3 domains, suggestive

of a boundary function to limit this chromatin type (Negre et al.,

2010). Indeed, depletion of dCTCF and its cofactor CP190 cause

some expansion of certain H3K27me3 domains (Bartkuhn et al.,

2009). SU(HW) binding is enriched at LAD borders, but its deple-

tion has only subtle effects on NL interactions (van Bemmel et al.,

1274 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

2010). A recent extensive survey indicates that most binding

sites for insulator-binding proteins in Drosophila may not act as

insulators or chromatin-blocking sites but may serve other func-

tions depending on the combination of proteins that is present

(Schwartz et al., 2012).

The only well-documented mammalian insulator-binding

protein to date is CTCF. It has about 15-30,000 focal binding

sites (Cuddapah et al., 2009; Handoko et al., 2011) and exhibits

a sharp enrichment at the borders of LADs (Guelen et al., 2008). It

also frequently demarcates domains of specific histone marks

such as H3K27me3 and H2AK5ac (Cuddapah et al., 2009; Han-

doko et al., 2011), and it may also contribute to the boundaries of

some topological domains (see below). This suggests that CTCF

helps to insulate chromatin domains, but conclusive evidence for

this has been difficult to obtain because depletion of CTCF is

lethal to many cell types. Furthermore, as will be discussed

below, insulator-binding proteins appear to be tightly linked to

the 3D organization of chromatin, and this may well hold a clue-

to their mechanisms of action. Disruption of individual CTCF-

binding sites will be needed to directly demonstrate their

function.

Domains in 3DLinear genomicmaps of chromatin features provide new insights

into the relationships between primary DNA sequence, some

chromatin structures, and gene expression, but understanding

genome function in vivo requires a consideration of the 3D struc-

ture of chromosomes in the nucleus. Historically, insights into

this aspect of genome biology have come from visual

approaches—mainly immunofluorescence and FISH. But this

is now being complemented by high-throughput assays that

use crosslinking and intramolecular ligation assays to query

the spatial relationships of different genomic loci. Collectively

these techniques have been termed ‘‘3C’’-based approaches

(see de Wit and de Laat, 2012 for a recent review of methodolo-

gies). Broadly, these methods fall into two categories: those that

‘‘look outward’’ from a particular sequence of interest to see

what other sequences in the genome can be captured together

with the original ‘‘bait’’ locus, and those that query all possible

combinations of ligated fragments—either within a defined

genomic region or across a whole genome. The former (one

against all) methodology is exemplified by the 4C technique (Si-

monis et al., 2006) and has been particularly useful for investi-

gating the associations of specific genes with putative long-

range regulatory elements (Noordermeer et al., 2008; Montavon

et al., 2011). Of the ‘‘all-against-all’’ methods, 5C is designed to

investigate chromatin conformation at high resolution within

a defined genomic region of up to a few Mb in size (Dostie

et al., 2006; Nora et al., 2012). This is exactly the size range

over which much long-range gene regulation seems to be able

to operate inmammalian cells (Williamson et al., 2011). However,

probing the spatial associations of entire large genomes of

complex metazoans is currently beyond the scope of 5C. Hi-C

is the method of choice for this latter kind of analysis (Lieber-

man-Aiden et al., 2009; Dixon et al., 2012; Kalhor et al., 2012;

Sexton et al., 2012; Zhang et al., 2012).

The first Hi-C studies of metazoan genomes gave relatively

coarse-grained (�1 Mb resolution) views of genome topology,

Figure 2. Radial Organization in the Nucleus, within and between Chromosomes(A) mESC nucleus hybridized by FISH with a paint for mouse chromosome 2 (green) and a probe (red) for just the exome of mouse chromosome 2 (Boyle et al.,2011). This demonstrates the looping-out of the gene-dense regions from the core chromosome territory and their disposition away from the nuclear peripheryand toward the interior of the nucleus.(B) Human lymphoblastoid cell nucleus hybridized by FISH with paint for the gene-rich human chromosome 19 (red) and gene-poor chromosome 18 (green)reveals the radial organization of chromosomes in the nucleus.(C) Courtesy of Frank Alber (University of Southern California). A density contour plot for the localization probability (red = max, green = min) of human chro-mosomes 1, 11, 14, 15, 16, 17, 19, 20, 21, and 22 in lymphoblastoid cell nuclei modeled from Hi-C data recapitulates the clustering of gene-rich chromosomes inthe center of the nucleus, along with the rDNA-containing acrocentric chromosomes (Kalhor et al., 2012).

the effective resolution of the analyses being limited not only by

the choice of restriction enzyme used to fragment the chromatin

(e.g., 4 bp or 6 bp cutter) but also by the library complexity gener-

ated after the PCR amplification of the ligated fragments and by

the depth of sequencing (Lieberman-Aiden et al., 2009). In an all-

versus-all assay such as Hi-C, a 10-fold increase in resolution

requires a 102-fold increase in the sequencing depth of a library

that is of sufficient complexity. Rapid increases in sequence

depth are now allowing the development of higher-resolution

topological genome maps (Dixon et al., 2012).

The Territory of the Chromosome

The dominant feature apparent in Hi-C analyses of metazoan

genomes, irrespective of cell type, is that each chromosome is

largely an individual territory—most of the captured associations

are in cis rather than in trans (Lieberman-Aiden et al., 2009; Kal-

hor et al., 2012; Sexton et al., 2012; Zhang et al., 2012), and this is

consistent with the appearance of chromosomes in the nucleus

as detected by FISH (Figures 2A and 2B). Both 4C (Tolhuis et al.,

2011) and Hi-C analyses (Hou et al., 2012; Kalhor et al., 2012;

Sexton et al., 2012) suggest that the centromere forms some

kind of barrier that attenuates associations between sequences

located on the two opposite arms (Figure 3) of the same chromo-

some—consistent with the visually separate appearance of the

p and q arms of human metacentric chromosomes (Dietzel

et al., 1998).

Even though associations in cis seem to dominate in most 3C

studies, robust associations of some sequences in trans are also

consistently seen in large-scale studies. Such sequences tend to

be in chromosomal regions characterized by high local gene

density, high transcriptional activity, and high density of DNase

I-hypersensitive sites (DHS) (Simonis et al., 2006; Lieberman-Ai-

den et al., 2009; Hou et al., 2012; Kalhor et al., 2012; Sexton

et al., 2012). The loci that have the highest probability of making

crosslinkable interactions in trans are those that also visibly loop

outside of the core of the visible chromosome territory to the

greatest extent (Kalhor et al., 2012). The infiltration of a looped-

out activated genomic region—the human major histocompati-

bility complex—into the territories of other chromosomes has

been directly visualized by high-resolution FISH (Branco and

Pombo, 2006). The whole-scale extent to which gene-dense

chromosomal regions decorate the outside of their own chromo-

some territories, beyond the limits of the core territory detected

by FISH with traditional chromosome ‘‘paints’’ and toward the

nuclear interior (Figure 2A), has now been revealed using custom

FISH probes composed of high-complexity oligonucleotide

pools targeted to the exonic regions of an entire chromosome

(Boyle et al., 2011). This calls for a reconsideration of what is

understood by the term chromosome territory to encapsulate

the core condensed part of the territory (that is visible with stan-

dard chromosome paints) and the surrounding territory ‘‘corona’’

that is only detectable by FISH with either locus-specific probes

orwith probes targeted at gene-dense chromatin. In this context,

looping-out from chromosome territories is actually looping-in to

the corona of one’s own territory and perhaps also looping-in to

the corona of neighboring territories.

The ability of a particular locus to locate outside of its own

chromosome territory core, and so have an increased proba-

bility of intermingling with sequences from other chromosomes,

is not just dependent on that locus’s own transcriptional activity

or chromatin state but rather is influenced by its local linear chro-

mosome context. This is best exemplified by the different

behaviors of the a-globin locus in mouse or human erythroid

cells. In human primary erythroid cells, the a-globin gene cluster

is within a large decondensed chromatin domain that often lies

outside of its own chromosome territory core (Brown et al.,

2006) and in spatial proximity to other active erythroid gene

loci located on other chromosomes (Brown et al., 2008). Due

to a break in conserved synteny, the mouse a-globin locus is

embedded in a genomic context different from that of its human

ortholog, its local chromatin environment is more condensed,

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1275

Figure 3. Long-Range Interactions among

Polycomb Domains in Drosophila(A) Interaction profile of the Polycomb target geneAnt-B (position indicated by red triangle) asdetermined by 4C. Data are shown as a domaino-gram, which visualizes the statistical significance(purple: moderately significant; red: highly signifi-cant) of the detected interactions for each chro-mosomal position. Vertical axis represents theresolution at which the significancewas calculated(see Tolhuis et al., 2011 for details). Note that theinteractions mostly overlap with Polycombdomains (blue boxes) yet are restricted to onechromosome arm.(B) Effect of a chromosomal inversion that joins thetwo red segments and the two blue segments.Data are plotted in the same order as in (A) for easycomparison. Note that Ant-B now interacts withseveral loci that were previously on the otherchromosome arm and no longer with loci that aremoved toward the other arm.(C) Cartoon interpretation of (A), illustrating thestochastic contacts of Ant-B (red triangle) withvarious Polycomb domains on the same chromo-some arm.

and it makes infrequent associations in trans with other highly

expressed genes in murine erythroid cells. Strikingly, when the

endogenous mouse a-globin locus was replaced with 120 kb

of the human sequence, the nuclear organization properties of

the ‘‘humanized’’ a-globin locus in mouse erythroid cells—i.e.,

few trans-associations and a localization within the chromo-

some territory core—were those characteristic of mouse and

not human a-globin (Brown et al., 2008). Thus, even for a 120

kb locus, the broader genomic context matters for its spatial

organization.

Clustering of Active Regions

Systematic FISH analysis across a multimegabase region of the

mouse genome demonstrated that within a single chromosome,

multiple gene-rich segments within the linear genome sequence

have a tendency to cluster together in the nuclear space (Shop-

land et al., 2006). This clustering was not seen for the intervening

gene-poor domains. 4CandHi-C techniques have confirmed this

tendency for active gene-dense domains to be able to associate

with each other on a global scale (Simonis et al., 2006; Lieber-

man-Aiden et al., 2009; Hakim et al., 2011; Splinter et al., 2011;

Yaffe and Tanay, 2011; Hou et al., 2012; Kalhor et al., 2012;

Sexton et al., 2012; Zhang et al., 2012). The majority of these

associations are intrachromosomal, but some interchromosomal

associations are also captured. This has been confirmed by

FISH—spatial proximity of loci associating in cis is generally

seen in a much higher proportion of nuclei than for loci involved

in 3C interchromosomal associations. Although these associa-

tions are occurring between active genomic regions, they are

not dependent on ongoing transcription (Palstra et al., 2008). It

may be that some other chromatin or functional feature of these

regions is responsible, or that transcription hasa role in theestab-

1276 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

lishment but not themaintenance of chro-

mosomal interaction networks.

More detailed analyses of genome-

wide chromatin profiles have identified

DHS as the most prominent chromatin feature enriched in the

active domains with a propensity to form long-range and inter-

chromosomal 3C associations with each other (Hakim et al.,

2011; Yaffe and Tanay, 2011). Whether it is the transcription

factors and chromatin-binding proteins responsible for gener-

ating the DHS or a feature of the underlying DNA topology itself

that is critical for the high-frequency of formaldehyde cross-

linked associations captured between these domains remains

to be determined.

Sowhat is the functional significance of the clustering of active

domains in the nucleus? At one extreme, it could be amanifesta-

tion of gene relocalization to specific transcription factories that

are specialized in particular transcriptional pathways and driven

by discrete transcription factors (Sutherland and Bickmore,

2009; Schoenfelder et al., 2010). However, 4C indicates that

associations between active domains are rather similar between

tissue types and that the associations captured by the active

b-globin locus in erythroid cells are to other generally transcrip-

tionally active regions, rather than to regions with erythroid- or

otherwise tissue-specific expression (Simonis et al., 2006).

Moreover, the transcriptional changes that are rapidly induced

in cells in response to ligand-activated glucocorticoid receptor

(GR) occur without any detectable dramatic changes in nuclear

organization (Hakim et al., 2011). One possibility is that the clus-

tering of active genes is a reflection of their congregation around

splicing factor-enriched nuclear speckles (Brown et al., 2006).

Consistent with the view of transcriptional responses playing

out against the background of a pre-established generic spatial

organization, recent Hi-C analysis has suggested that the asso-

ciations between active genomic regions are largely indiscrimi-

nate—i.e., there are no preferred pairs of associated domains,

beyond those that are due to sharing of the same chromosome

territory or due to the known spatial clustering of small gene-

rich chromosomes to the center of the nucleus as a consequence

of radial nuclear organization (Boyle et al., 2001; Cremer et al.,

2001; Kalhor et al., 2012) (Figures 2B and 2C). This stochastic

self-association of very active genomic regions may in turn

contribute toward radial nuclear organization.

The functional consequences of spatial encounters between

loci from different chromosomes are revealed by the behavior

of a human b-globin enhancer (the LCR) integrated into an

ectopic site in the mouse genome. Even in places of erythropoi-

esis (fetal liver) where the LCR normally acts to enhance b-globin

expression, this elementdid not change the spectrumof4Casso-

ciations made by the site of integration. However, it did increase

the efficiency of pre-existing contacts likely by enhancing looping

out of the surrounding region from the chromosome territory core

(Noordermeer et al., 2011a). There was no functional conse-

quence of the integrated LCR on the expression of other mouse

genes, except for those immediately in cis to the integrated

LCR and one other gene located elsewhere. The endogenous

mouse b-globin gene Hbb-bh1 is normally expressed only at

earlier—embryonic—stages of development dependent on its

own LCR, but it was upregulated in fetal livers of human LCR

transgenic animals. Almost a third of the cells with detectable

cytoplasmic Hbb-bh1 messenger RNA (mRNA) showed spatial

colocalization of Hbb-bh1 with the ectopic human LCR in trans,

far above the frequency seen in the cell population as a whole.

This elegant experiment demonstrates that there can be func-

tional consequences on gene expression for colocalization in

trans, but that due to the stochastic nature of these interactions

and the constraint placed on them by their surrounding genomic

context, they are unlikely to have a deterministic role in pathways

of developmental gene regulation. However, they could con-

tribute to variation in gene expression levels between cells of

a population that could then be acted upon and exploited, for

example, by external signaling pathways or environmental cues.

Clustering of Inactive Regions

Whereas active regions tend to associate with other active

regions, do inactive regions then preferentially keep company

with other inactive regions—and for what purpose?

Although FISH analysis along a mouse chromosome did not

find clustering of inactive domains with each other in the same

way that was seen for the active regions (Shopland et al.,

2006), both 4C (Simonis et al., 2006) and Hi-C (Lieberman-Aiden

et al., 2009; Sexton et al., 2012) studies revealed the preferential

capture of inactive loci with other inactive regions of the genome.

A refinement of the computational analysis of Hi-C data from

mammalian cells unmasked further subtleties in this pattern,

with clusters of inactive regions partitioning into those that are

close to centromeres or located on relatively short chromosome

arms and those that are more distally located on larger chromo-

somes (Yaffe and Tanay, 2011; Imakaev et al., 2012). This seems

unlikely to reflect a fundamental difference in the nature of

these inactive chromatin domains themselves but rather some

restriction that is placed on their ability to encounter each

other—dictated by overall chromosome topology or centromere

behavior. A Hi-C analysis of theDrosophila embryowas also able

to detect the known spatial clustering of telomeres and of

centromeres with each other and with the heterochromatic 4th

chromosome (Sexton et al., 2012). This is consistent with the

known ability of heterochromatin clustering to drive changes in

nuclear organization (Csink and Henikoff, 1996) and indicates

that H3K9-methylated genomic regions have a tendency to

self-associate.

Inactive regions are more constrained from associating over

long genomic distances than are active regions—their Hi-C

contacts being restricted to regions from the same chromosome

and, moreover, the same chromosome arm. This probably

reflects the fact that inactive regions have less freedomofmotion

in the nucleus than active domains and are restricted to life within

their own core chromosome territories and at the NL (Kalhor

et al., 2012). This has been directly visualized for one particular

example. Hox loci are maintained in a silent and compact chro-

matin state in mammalian ESCs by the Polycomb PRC2 and

PRC1 complexes (Eskeland et al., 2010), and they are entirely

locatedwithin their host chromosome territories. Upon their tran-

scriptional activation, Hox loci acquire a greater freedom of

movement in the nucleus, and the active alleles can then be

seen to adopt positions either inside or outside of their chromo-

some territory cores (Morey et al., 2009). This coincides with an

enhanced ability of Hox loci to be captured together with

sequences from other chromosomes in 3C-type experiments

(Wurtele and Chartrand, 2006).

Genome-wide Hi-C studies do not reveal any spatial segrega-

tion of different categories of inactive chromatin (e.g., H3K9me3

versus H3K27me3). However, in Drosophila, long-range 4C

contacts and spatial colocalization of silent Polycomb targets,

including Hox loci, have been demonstrated in embryos and

larvae (Bantignies et al., 2011; Tolhuis et al., 2011) (Figure 3A).

Silent, non-Polycomb target loci do not cluster with the Poly-

comb sites, indicating that this association is not simply driven

by lack of transcriptional activity, and indeed associations of Pol-

ycomb target loci were shown to be dependent on PcG proteins

themselves. This indicates that the spatial associations of some

inactive regions are directly linked to their particular mechanism

of epigenetic silencing.

The phenotypic consequences of spatial associations

between Polycomb targets are less clear. Some altered gene

expression, and a modest phenotypic consequence, could be

measured when associations were perturbed in one study (Ban-

tignies et al., 2011). However, major changes in spatial associa-

tions of PcG targets, caused by a chromosomal inversion (Fig-

ure 3B), were not accompanied by any detectable alteration of

gene expression and no gross phenotypic consequence in flies

homozygous for the inversion (Tolhuis et al., 2011).

Topologically Associating Domains

Recent Hi-C and 5C studies in fly and mammalian cells have

yielded data sets of unprecedented resolution and coverage

(Dixon et al., 2012; Hou et al., 2012; Nora et al., 2012; Sexton

et al., 2012). These studies indicate that crosslinked associations

are enriched locally within discrete domains that are �100 kb in

size inDrosophila and an average of�900 kb in mammalian cells

(Figures 1A and 4A). Further increases in sequencing depth may

lead to refinements of these estimates. In mammalian cells,

about 2,000 of these domains collectively tile most of the

genome (Dixon et al., 2012). High-resolution FISH was used to

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1277

Figure 4. Topologically Associating Domains(A) Matrix plot showing all pairwise interaction frequencies captured by 5C(color scale white—blue—black) among loci in an �4.5 Mb region on the Xchromosome in mESCs. TADs are the large, discrete blocks within which thepairwise contacts are relatively frequent. Gray area corresponds to a repetitiveregion that could not be probed.(B) H3K27me3 distribution (Marks et al., 2009) and Lamin B1 interactions(Peric-Hupkes et al., 2010) in the same region. Note the partial overlap withTADs.Images in (A) and (B) are courtesy of Elphege Nora (Institut Curie).

show that two loci within a single domain are visibly closer

together in the nucleus than loci separated by the same genomic

distance but located in adjacent domains. Additionally, hybrid-

ization signals from complex probe pools entirely located within

one domain intermingle with each other to a greater extent than

probe pools that span across domain boundaries (Nora et al.,

2012). Hence these sub-megabase-sized self-associating

domains have been called ‘‘topologically associating domains

(TADs),’’ or simply ‘‘topological domains.’’

These domains show a remarkable degree of alignment to the

distribution of some active and repressive (H3K9me3 and

H3K27me3) histone modifications along the genome and also

to LADs (Figure 4B). However, the stability of these local self-in-

teracting domains in cell types with very different patterns of

gene expression and epigenetic modifications—e.g., ESCs,

1278 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

adult fibroblasts, and brain—suggests that it is not transcription

or histone modifications that dictate the domain structure but

rather some inherent property linked to the underlying genome

sequence, for example, binding sites for ubiquitous sequence-

specific DNA-binding factors. That histone modifications act

downstream of topological domain structure is most graphically

illustrated by persistence of the domains in ESCs that lack

G9a or Eed, the histone-modifying activities responsible for

H3K9me2 and H3K27me3, respectively (Nora et al., 2012).

Some differences in Hi-C interactions are seen between cell

types, and these correspond to differentially regulated genes.

Interestingly, these facultative focal interactions tend to occur

within the confines of the stable domains. This suggests that

Hi-C- and 5C-defined domains may reflect the normal sphere

of influence over which long-range gene regulation can occur.

Similarly, expression changes during differentiation of X-linked

genes within the same 5C-defined domain are more correlated

than those of genes located in separate domains or for random

gene sets. Moreover, genetic deletion of a specific topological

boundary on the mouse X resulted in transcriptional misregula-

tion and the acquisition of new ectopic long-range 5C contacts

(Nora et al., 2012).

To define what might constitute the boundaries of TADs, their

positions have been aligned to maps of insulator proteins and

other epigenomic features. In Drosophila, domain borders were

enriched in DHS and particularly in binding sites of the insulator

component CP190 and of Chromator, a known regulator of chro-

mosome structure (Sexton et al., 2012). In mammalian cells,

domain boundaries were especially enriched in the promoters

of housekeeping genes and binding sites for CTCF. However,

only a minority of CTCF sites are associated with TAD bound-

aries. Other genomic elements enriched at mammalian TAD

boundaries include transfer RNA (tRNA) genes and interspersed

repeats of the SINE family, both of which are capable of confer-

ring insulator activity (Lunyak et al., 2007; Raab et al., 2012). One

feature that may unify these various elements is the presence of

nucleosome-free regions and disrupted chromatin fiber struc-

ture. Together with the binding of specific protein factors, this

might result in a rigid chromatin structure that constrains regions

on either side from intermingling. That elements located between

TADs are important for restraining associations between adja-

cent domains is demonstrated by the consequence of a deletion

that removes one of the TAD boundaries on the mouse X chro-

mosome (Nora et al., 2012). Loci from the two flanking TADs

gained interactions with each other, and the structure of one of

the original TADs was reconfigured as a result of this deletion.

However, the fact that the two TADs flanking the deletion did

not completely merge into one also indicates that TADs are gov-

erned by factors in addition to domain boundaries.

Chromatin Compaction and Chromatin Domains

It has long been appreciated that levels of chromatin compaction

vary across the genome. The appearance of puffs on Drosophila

polytene chromosomes is a graphic manifestation of the chro-

matin decompaction of active genomic regions, as is the

‘‘bloated’’ appearance of the hyperactive X chromosome in

male flies. In mammalian cells, FISH has also revealed differ-

ences in compaction from different regions of the genome. For

example, hybridization signal from a gene-rich chromosome

territory occupies a larger proportion of the nucleus than does

the signal from an equivalently (in Mb) sized gene-poor chromo-

some (Croft et al., 1999), and different degrees of compaction

have been inferred from the relationships of interprobe nuclear

versus genomic distances at G-band and R-band regions of

the genome (Yokota et al., 1997).

The first genome-wide attempt to document chromatin

compaction in mammalian cells used sucrose-gradient sedi-

mentation to assay the frictional properties of long chromatin

fragments (Gilbert et al., 2004). Slowly sedimenting fibers (high

frictional coefficient) were inferred to be in a more open structure

than fast sedimenting fibers of the same length, and this was

confirmed by FISH. Domains of open chromatin fiber structure

correspond to the most gene-dense active ‘‘R-band’’ regions

of the genome. Other studies have confirmed a more compact

and spherical chromatin structure of gene-poor regions

compared to gene-dense domains, and this structure is inde-

pendent of transcriptional activity (Goetze et al., 2007). In the

same way that interphase distances measured by FISH scale

with genomic distance, so all 3C-type data decaywith increasing

genomic distance. Consistent with the idea of a less compact

chromatin structure at active genomic regions, the calculated

scaling factor for Hi-C data from active regions of the Drosophila

genome is higher than that for repressed domains, and this was

suggested to also reflect a lower level of chromatin compaction

at active domains (Sexton et al., 2012).

There will be very many factors involved in modulating higher-

order chromatin compaction, but one specific determinant

known to be important for maintaining inactive domains in

a compact state are the PRCs. In mESCs, the PRC1 complex

was shown by FISH to be essential for maintaining a compact

chromatin state at silent Hox loci. The PRC2 complex and the

associated H3K27me3 are insufficient to maintain chromatin

compaction; PRC1 was required (Eskeland et al., 2010). These

observations are consistent with the very compact appearance

of Hox loci in polytene chromosomes from anterior parts of

Drosophila larvae where Hox genes are repressed and bound

by Polycomb (Marchetti et al., 2003).

This view of Polycomb-repressed regions as compact chro-

matin domains is consistent with some (Ferraiuolo et al., 2010;

Noordermeer et al., 2011b), but not all (Wang et al., 2011) inter-

pretations of 3C-type studies. Moreover, the compact structure

of active Hox loci inferred from 4C and 5C analysis of human

fibroblasts is at odds with the decompact appearance of active

Hox loci in mammalian cells and embryos as visualized by

FISH (Chambeyron et al., 2005; Morey et al., 2007), and with

the puffed appearance of active Hox loci from the posterior of

Drosophila larvae (Marchetti et al., 2003). More work needs to

be done to determine to what extent 3C profiles across a locus

can be translated into 3D chromatin conformation (Dostie and

Bickmore, 2012).

Computational Models of Genome Topology

Interpreting the rich data from large-scale 3C-type studies in

terms of chromosome folding requires advanced computer

modeling. Methods for this are still being developed and come

in two broad classes.

The first class of methods begins with theoretical analysis or

in silico modeling of idealized polymers in a confined space as

a model of chromosomes in a nucleus. By altering specific vari-

ables such as polymer stiffness, repulsion or ‘‘stickiness’’ of

fibers, the presence of looping interactions, etc., one can then

calculate defining parameters that describe the spatial folding

of chromosomes, such as the overall relationship between the

linear distance (in kb) of two loci and their contact frequency.

These theoretical parameters are then matched against those

determined from 5C or Hi-C data sets as well as FISH distance

measurements, and themodel yielding the best match is consid-

ered to be the most likely model to describe chromosome archi-

tecture.

Taking this approach, Lieberman-Aiden et al. (2009)

concluded from their Hi-C data that human interphase chromo-

somes may best be described by a model of a fractal globule—

a polymer model in which one region is topologically constrained

from passing across and entangling with another region and in

which the polymer crumples into globules on all scales (Mirny,

2011). However, this model is inconsistent with FISH data that

show that the linear relationship of physical distance (mean-

square interphase distance) to genomic separation plateaus at

distances > 1–2 Mb (Yokota et al., 1995; Gilbert et al., 2004; Ma-

teos-Langerak et al., 2009). FISH data seem more compatible

with �1 Mb equilibrium globule models in which the chromatin

fiber has a random-walk configuration but is confined within

a defined volume. The nature of those boundary walls is not

known, but they may be compatible with the boundaries

between TADs identified by high-resolution Hi-C and 5C (Dixon

et al., 2012; Nora et al., 2012). Another testable feature that is

fundamentally different between the fractal and equilibrium

globule models is the entanglement of the polymer chain in the

latter. However, the action of topoisomerase II should be suffi-

cient to deal with this, and indeed, simulations have suggested

that topoII activity may be sufficient to rapidly drive a fractal

globule into an equilibrium globule state during interphase

(Mirny, 2011). A recent ‘‘strings and binders switch’’ model

combines features of a random-walk polymer model with the

effects of interactions mediated by diffusible factors (e.g.,

proteins) bound to the chromatin (Barbieri et al., 2012). This

model nicely reproduces the plateauing of mean-squared inter-

phase distances at larger genomic separations, suggests that

high concentrations of bound factors can collapse the chromatin

into a compact state, and shows how interactions between

different types of bound factors can reproduce the spatial segre-

gation of genomic domains with different characteristics (e.g.,

active versus inactive).

The second class of computational methods aims to recon-

struct the actual trajectory of a chromosomal fiber inside the

nucleus, based on 5C or Hi-C data. Here, the major challenge

is that chromosome folding is stochastic and variable from cell

to cell. Appropriate modeling is hence done in a probabilistic

manner that yields ensembles of possible trajectories, rather

than a single ‘‘average’’ path. One such study adapted a method

for solving protein structures from NMR spectroscopy data to

model the most likely folding trajectories of the human 0.5 Mb

a-globin locus and a bacterial chromosome (Bau et al., 2011;

Umbarger et al., 2011). Another study modeled the coarse-

grained organization and positions of entire chromosomes in

thousands of cell nuclei based on Hi-C data (Kalhor et al.,

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1279

2012). This faithfully reproduced the statistical clustering of

gene-rich chromosomes in the center of the nucleus and the

looping out of highly transcribed regions from the bulk of their

chromosome territories—allowing contact between chromo-

somes (Figure 2). Further development of such computational

models will be indispensable to interpret the large amounts of

4C, 5C, and Hi-C data sets that are to be expected.

Challenges and Future DirectionsGenome-wide methodologies provide an unprecedented oppor-

tunity to describe both the linear and 3D compartmentalization of

genomes. However, presently available mapping techniques

require thousands to millions of cells and thus only provide pop-

ulation averages. Most Hi-C and microscopy evidence indicates

that specific long-range contacts occur in a small fraction of cells

at any given moment (Simonis et al., 2006; Lieberman-Aiden

et al., 2009; Schoenfelder et al., 2010). Accordingly, our under-

standing of the true nature of chromatin domains is probably

blurred by population-averaged data, and cartoon models as

in Figure 1B should be viewed only as basic working models.

For example, is a given chromatin domain occupied by its

cognate proteins in every cell of a population, or only in a subset

of cells? And in any single cell, is a chromatin domain covered in

its entirety by proteins, or only in part? An important future chal-

lengewill be to devise strategies that can generate genome-wide

data sets from single cells and so directly capture the stochastic

behavior of chromosomes.

Also lagging behind are experimental approaches that can

efficiently manipulate linear and 3D domains. These types of

interventionist approaches are key to determining the functional

significance of genome organization or whether the structures

are just reflective of genome functions. The overall similarity of

Hi-C maps generated from mammalian cell populations as

diverse as rapidly dividing pluripotent ESCs, terminally differen-

tiated cells (fibroblasts and lymphoblastoid cell lines), and nondi-

viding pro-B cells (Lieberman-Aiden et al., 2009; Dixon et al.,

2012; Zhang et al., 2012) suggests that the overall spatial orga-

nization of the mammalian genome is a fundamental state

upon which genome function is then largely played out. One

well-established consequence of 3D organization is its influence

on the spectrum of chromosomal translocations that occur as

a consequence of double-strand break repair by nonhomolo-

gous end-joining (Zhang et al., 2012). Conversely, translocations

perturb the normal spatial context of the participating chromo-

somes (Tolhuis et al., 2011) and in some instances do appear

to impact on gene expression from these chromosomes (Hare-

wood et al., 2010). Using mouse genetics to tailor-make specific

translocations might be a productive way to better explore the

functional consequences of some aspects of 3D organization.

Similarly, deleting specific boundaries between chromatin

domains and binding sites for proteins thought to be important

determinants of such domains provides a way to explore the

functional relevance of linear and 3D compartments (Nora

et al., 2012). Newmethods for the efficient editing of the genome

(e.g., Wood et al., 2011) will facilitate such approaches. Artificial

DNA-binding proteins can also be harnessed to create new

topological structures and study the functional consequences

(Deng et al., 2012).

1280 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Complementing these approaches is the integration of

reporter constructs into different types of chromatin domains,

which offers a direct readout out of the effects of distinct local

environments (Gierman et al., 2007). Because the interplay

between genomic elements is expected to be complex, such

strategies need to be scaled up in order to explore the wide

range of combinatorial possibilities and to obtain unbiased and

broadly interpretable results.

Despite these challenges, it appears that the pieces of the

chromosome puzzle are now coming together, and it will be

exciting to dissect the underlying molecular mechanisms and

elucidate how chromosome-domain organization contributes

to gene regulation and other nuclear functions.

ACKNOWLEDGMENTS

We thank members of the van Steensel lab; Duncan Sproul, Nick Gilbert, and

Colin Semple of IGMM for critical reading of the manuscript; an anonymous

reviewer for many thoughtful suggestions; and Frank Alber, Elphege Nora,

and Tyrone Ryba for help with figures. We are supported by ERC Advanced

grant 293662 and NWO-VICI (B.v.S.) and by the MRC and ERC Advanced

grant 249956 (W.A.B.).

REFERENCES

Arnold, P., Scholer, A., Pachkov, M., Balwierz, P.J., Jørgensen, H., Stadler,

M.B., van Nimwegen, E., and Schubeler, D. (2013). Modeling of epigenome

dynamics identifies transcription factors that mediate Polycomb targeting.

Genome Res. 23, 60–73.

Bantignies, F., Roure, V., Comet, I., Leblanc, B., Schuettengruber, B., Bonnet,

J., Tixier, V., Mas, A., and Cavalli, G. (2011). Polycomb-dependent regulatory

contacts between distant Hox loci in Drosophila. Cell 144, 214–226.

Barbieri, M., Chotalia, M., Fraser, J., Lavitas, L.M., Dostie, J., Pombo, A., and

Nicodemi, M. (2012). Complexity of chromatin folding is captured by the

strings and binders switch model. Proc. Natl. Acad. Sci. USA 109, 16173–

16178.

Bartkuhn, M., Straub, T., Herold, M., Herrmann, M., Rathke, C., Saumweber,

H., Gilfillan, G.D., Becker, P.B., and Renkawitz, R. (2009). Active promoters

and insulators are marked by the centrosomal protein 190. EMBO J. 28,

877–888.

Bau, D., Sanyal, A., Lajoie, B.R., Capriotti, E., Byron, M., Lawrence, J.B., Dek-

ker, J., and Marti-Renom, M.A. (2011). The three-dimensional folding of the

a-globin gene domain reveals formation of chromatin globules. Nat. Struct.

Mol. Biol. 18, 107–114.

Berman, B.P., Weisenberger, D.J., Aman, J.F., Hinoue, T., Ramjan, Z., Liu, Y.,

Noushmehr, H., Lange, C.P., van Dijk, C.M., Tollenaar, R.A., et al. (2012).

Regions of focal DNA hypermethylation and long-range hypomethylation in

colorectal cancer coincide with nuclear lamina-associated domains. Nat.

Genet. 44, 40–46.

Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry,

B., Meissner, A., Wernig, M., Plath, K., et al. (2006). A bivalent chromatin struc-

ture marks key developmental genes in embryonic stem cells. Cell 125,

315–326.

Boyle, S., Gilchrist, S., Bridger, J.M., Mahy, N.L., Ellis, J.A., and Bickmore,

W.A. (2001). The spatial organization of human chromosomes within the nuclei

of normal and emerin-mutant cells. Hum. Mol. Genet. 10, 211–219.

Boyle, S., Rodesch, M.J., Halvensleben, H.A., Jeddeloh, J.A., and Bickmore,

W.A. (2011). Fluorescence in situ hybridization with high-complexity repeat-

free oligonucleotide probes generated by massively parallel synthesis. Chro-

mosome Res. 19, 901–909.

Branco, M.R., and Pombo, A. (2006). Intermingling of chromosome territories

in interphase suggests role in translocations and transcription-dependent

associations. PLoS Biol. 4, e138.

Brown, J.M., Leach, J., Reittie, J.E., Atzberger, A., Lee-Prudhoe, J., Wood,

W.G., Higgs, D.R., Iborra, F.J., and Buckle, V.J. (2006). Coregulated human

globin genes are frequently in spatial proximity when active. J. Cell Biol. 172,

177–187.

Brown, J.M., Green, J., das Neves, R.P., Wallace, H.A., Smith, A.J., Hughes,

J., Gray, N., Taylor, S., Wood, W.G., Higgs, D.R., et al. (2008). Association

between active genes occurs at nuclear speckles and is modulated by chro-

matin environment. J. Cell Biol. 182, 1083–1097.

Capelson, M., Liang, Y., Schulte, R., Mair, W., Wagner, U., and Hetzer, M.W.

(2010). Chromatin-bound nuclear pore components regulate gene expression

in higher eukaryotes. Cell 140, 372–383.

Caron, H., van Schaik, B., van der Mee, M., Baas, F., Riggins, G., van Sluis, P.,

Hermus, M.C., van Asperen, R., Boon, K., Voute, P.A., et al. (2001). The human

transcriptome map: clustering of highly expressed genes in chromosomal

domains. Science 291, 1289–1292.

Chambeyron, S., Da Silva, N.R., Lawson, K.A., and Bickmore, W.A. (2005).

Nuclear re-organisation of the Hoxb complex during mouse embryonic devel-

opment. Development 132, 2215–2223.

Costantini, M., Clay, O., Auletta, F., and Bernardi, G. (2006). An isochore map

of human chromosomes. Genome Res. 16, 536–541.

Cremer, M., von Hase, J., Volm, T., Brero, A., Kreth, G., Walter, J., Fischer, C.,

Solovei, I., Cremer, C., and Cremer, T. (2001). Non-random radial higher-order

chromatin arrangements in nuclei of diploid human cells. Chromosome Res. 9,

541–567.

Cremer, T., Cremer, M., Dietzel, S., Muller, S., Solovei, I., and Fakan, S. (2006).

Chromosome territories—a functional nuclear landscape. Curr. Opin. Cell Biol.

18, 307–316.

Croft, J.A., Bridger, J.M., Boyle, S., Perry, P., Teague, P., and Bickmore, W.A.

(1999). Differences in the localization and morphology of chromosomes in the

human nucleus. J. Cell Biol. 145, 1119–1131.

Csink, A.K., and Henikoff, S. (1996). Genetic modification of heterochromatic

association and nuclear organization in Drosophila. Nature 381, 529–531.

Cuddapah, S., Jothi, R., Schones, D.E., Roh, T.Y., Cui, K., and Zhao, K. (2009).

Global analysis of the insulator binding protein CTCF in chromatin barrier

regions reveals demarcation of active and repressive domains. Genome

Res. 19, 24–32.

de Wit, E., and de Laat, W. (2012). A decade of 3C technologies: insights into

nuclear organization. Genes Dev. 26, 11–24.

Deng, W., Lee, J., Wang, H., Miller, J., Reik, A., Gregory, P.D., Dean, A., and

Blobel, G.A. (2012). Controlling long-range genomic interactions at a native

locus by targeted tethering of a looping factor. Cell 149, 1233–1244.

Dialynas, G., Speese, S., Budnik, V., Geyer, P.K., andWallrath, L.L. (2010). The

role of Drosophila Lamin C in muscle function and gene expression. Develop-

ment 137, 3067–3077.

Dietzel, S., Jauch, A., Kienle, D., Qu, G., Holtgreve-Grez, H., Eils, R., Munkel,

C., Bittner, M., Meltzer, P.S., Trent, J.M., and Cremer, T. (1998). Separate and

variably shaped chromosome arm domains are disclosed by chromosome

arm painting in human cell nuclei. Chromosome Res. 6, 25–33.

Dixon, J.R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., Hu, M., Liu, J.S., and

Ren, B. (2012). Topological domains in mammalian genomes identified by

analysis of chromatin interactions. Nature 485, 376–380.

Dostie, J., and Bickmore, W.A. (2012). Chromosome organization in the

nucleus - charting new territory across the Hi-Cs. Curr. Opin. Genet. Dev.

22, 125–131.

Dostie, J., Richmond, T.A., Arnaout, R.A., Selzer, R.R., Lee, W.L., Honan, T.A.,

Rubio, E.D., Krumm, A., Lamb, J., Nusbaum, C., et al. (2006). Chromosome

Conformation Capture Carbon Copy (5C): a massively parallel solution for

mapping interactions between genomic elements. Genome Res. 16, 1299–

1309.

Dunham, I., Kundaje, A., Aldred, S.F., Collins, P.J., Davis, C.A., Doyle, F., Ep-

stein, C.B., Frietze, S., Harrow, J., Kaul, R., et al.; ENCODE Project Consor-

tium. (2012). An integrated encyclopedia of DNA elements in the human

genome. Nature 489, 57–74.

Ernst, J., and Kellis, M. (2010). Discovery and characterization of chromatin

states for systematic annotation of the human genome. Nat. Biotechnol. 28,

817–825.

Eskeland, R., Leeb, M., Grimes, G.R., Kress, C., Boyle, S., Sproul, D., Gilbert,

N., Fan, Y., Skoultchi, A.I., Wutz, A., and Bickmore, W.A. (2010). Ring1B

compacts chromatin structure and represses gene expression independent

of histone ubiquitination. Mol. Cell 38, 452–464.

Ferraiuolo, M.A., Rousseau, M., Miyamoto, C., Shenker, S., Wang, X.Q., Na-

dler, M., Blanchette, M., and Dostie, J. (2010). The three-dimensional architec-

ture of Hox cluster silencing. Nucleic Acids Res. 38, 7472–7484.

Filion, G.J., and van Steensel, B. (2010). Reassessing the abundance of

H3K9me2 chromatin domains in embryonic stem cells. Nat. Genet. 42, 4,

author reply 5–6.

Filion, G.J., van Bemmel, J.G., Braunschweig, U., Talhout, W., Kind, J., Ward,

L.D., Brugman, W., de Castro, I.J., Kerkhoven, R.M., Bussemaker, H.J., and

van Steensel, B. (2010). Systematic protein location mapping reveals five prin-

cipal chromatin types in Drosophila cells. Cell 143, 212–224.

Finlan, L.E., Sproul, D., Thomson, I., Boyle, S., Kerr, E., Perry, P., Ylstra, B.,

Chubb, J.R., and Bickmore, W.A. (2008). Recruitment to the nuclear periphery

can alter expression of genes in human cells. PLoS Genet. 4, e1000039.

Frietze, S., O’Geen, H., Blahnik, K.R., Jin, V.X., and Farnham, P.J. (2010).

ZNF274 recruits the histone methyltransferase SETDB1 to the 30 ends of

ZNF genes. PLoS ONE 5, e15082.

Gerstein, M.B., Lu, Z.J., Van Nostrand, E.L., Cheng, C., Arshinoff, B.I., Liu, T.,

Yip, K.Y., Robilotto, R., Rechtsteiner, A., Ikegami, K., et al.; modENCODE

Consortium. (2010). Integrative analysis of the Caenorhabditis elegans

genome by the modENCODE project. Science 330, 1775–1787.

Gierman, H.J., Indemans, M.H., Koster, J., Goetze, S., Seppen, J., Geerts, D.,

van Driel, R., and Versteeg, R. (2007). Domain-wide regulation of gene expres-

sion in the human genome. Genome Res. 17, 1286–1295.

Gilbert, N., Boyle, S., Fiegler, H., Woodfine, K., Carter, N.P., and Bickmore,

W.A. (2004). Chromatin architecture of the human genome: gene-rich domains

are enriched in open chromatin fibers. Cell 118, 555–566.

Goetze, S., Mateos-Langerak, J., Gierman, H.J., de Leeuw, W., Giromus, O.,

Indemans, M.H., Koster, J., Ondrej, V., Versteeg, R., and van Driel, R.

(2007). The three-dimensional structure of human interphase chromosomes

is related to the transcriptome map. Mol. Cell. Biol. 27, 4475–4487.

Groner, A.C., Meylan, S., Ciuffi, A., Zangger, N., Ambrosini, G., Denervaud, N.,

Bucher, P., and Trono, D. (2010). KRAB-zinc finger proteins and KAP1 can

mediate long-range transcriptional repression through heterochromatin

spreading. PLoS Genet. 6, e1000869.

Guelen, L., Pagie, L., Brasset, E., Meuleman, W., Faza, M.B., Talhout, W., Eu-

ssen, B.H., de Klein, A., Wessels, L., de Laat, W., and van Steensel, B. (2008).

Domain organization of human chromosomes revealed by mapping of nuclear

lamina interactions. Nature 453, 948–951.

Hakim, O., Sung, M.H., Voss, T.C., Splinter, E., John, S., Sabo, P.J., Thurman,

R.E., Stamatoyannopoulos, J.A., de Laat, W., and Hager, G.L. (2011). Diverse

gene reprogramming events occur in the same spatial clusters of distal regu-

latory elements. Genome Res. 21, 697–706.

Handoko, L., Xu, H., Li, G., Ngan, C.Y., Chew, E., Schnapp, M., Lee, C.W., Ye,

C., Ping, J.L., Mulawadi, F., et al. (2011). CTCF-mediated functional chromatin

interactome in pluripotent cells. Nat. Genet. 43, 630–638.

Hansen, K.D., Timp, W., Bravo, H.C., Sabunciyan, S., Langmead, B., McDo-

nald, O.G., Wen, B., Wu, H., Liu, Y., Diep, D., et al. (2011). Increased methyla-

tion variation in epigenetic domains across cancer types. Nat. Genet. 43,

768–775.

Hansen, R.S., Thomas, S., Sandstrom, R., Canfield, T.K., Thurman, R.E.,

Weaver, M., Dorschner, M.O., Gartler, S.M., and Stamatoyannopoulos, J.A.

(2010). Sequencing newly replicated DNA reveals widespread plasticity in

human replication timing. Proc. Natl. Acad. Sci. USA 107, 139–144.

Harewood, L., Schutz, F., Boyle, S., Perry, P., Delorenzi, M., Bickmore, W.A.,

and Reymond, A. (2010). The effect of translocation-induced nuclear reorgani-

zation on gene expression. Genome Res. 20, 554–564.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1281

Hathaway, N.A., Bell, O., Hodges, C., Miller, E.L., Neel, D.S., and Crabtree,

G.R. (2012). Dynamics and memory of heterochromatin in living cells. Cell

149, 1447–1460.

Hawkins, R.D., Hon, G.C., Lee, L.K., Ngo, Q., Lister, R., Pelizzola, M., Edsall,

L.E., Kuan, S., Luu, Y., Klugman, S., et al. (2010). Distinct epigenomic land-

scapes of pluripotent and lineage-committed human cells. Cell Stem Cell 6,

479–491.

Hiratani, I., Ryba, T., Itoh, M., Yokochi, T., Schwaiger, M., Chang, C.W., Lyou,

Y., Townes, T.M., Schubeler, D., and Gilbert, D.M. (2008). Global reorganiza-

tion of replication domains during embryonic stem cell differentiation. PLoS

Biol. 6, e245.

Hou, C., Li, L., Qin, Z.S., and Corces, V.G. (2012). Gene density, transcription,

and insulators contribute to the partition of the Drosophila genome into phys-

ical domains. Mol. Cell 48, 471–484.

Imakaev, M., Fudenberg, G., McCord, R.P., Naumova, N., Goloborodko, A.,

Lajoie, B.R., Dekker, J., and Mirny, L.A. (2012). Iterative correction of Hi-C

data reveals hallmarks of chromosome organization. Nat. Methods 9,

999–1003.

Jiang, H., Shukla, A., Wang, X., Chen, W.Y., Bernstein, B.E., and Roeder, R.G.

(2011). Role for Dpy-30 in ES cell-fate specification by regulation of H3K4

methylation within bivalent domains. Cell 144, 513–525.

Kalhor, R., Tjong, H., Jayathilaka, N., Alber, F., and Chen, L. (2012). Genome

architectures revealed by tethered chromosome conformation capture and

population-based modeling. Nat. Biotechnol. 30, 90–98.

Kalverda, B., Pickersgill, H., Shloma, V.V., and Fornerod, M. (2010). Nucleo-

porins directly stimulate expression of developmental and cell-cycle genes

inside the nucleoplasm. Cell 140, 360–371.

Kharchenko, P.V., Alekseyenko, A.A., Schwartz, Y.B., Minoda, A., Riddle,

N.C., Ernst, J., Sabo, P.J., Larschan, E., Gorchakov, A.A., Gu, T., et al.

(2011). Comprehensive analysis of the chromatin landscape in Drosophila mel-

anogaster. Nature 471, 480–485.

Ku, M., Koche, R.P., Rheinbay, E., Mendenhall, E.M., Endoh, M., Mikkelsen,

T.S., Presser, A., Nusbaum, C., Xie, X., Chi, A.S., et al. (2008). Genomewide

analysis of PRC1 and PRC2 occupancy identifies two classes of bivalent

domains. PLoS Genet. 4, e1000242.

Kumaran, R.I., and Spector, D.L. (2008). A genetic locus targeted to the

nuclear periphery in living cells maintains its transcriptional competence. J.

Cell Biol. 180, 51–65.

Kwon, S.H., and Workman, J.L. (2011). The changing faces of HP1: From

heterochromatin formation and gene silencing to euchromatic gene expres-

sion: HP1 acts as a positive regulator of transcription. Bioessays 33, 280–289.

Lee, T.I., Jenner, R.G., Boyer, L.A., Guenther, M.G., Levine, S.S., Kumar, R.M.,

Chevalier, B., Johnstone, S.E., Cole, M.F., Isono, K., et al. (2006). Control of

developmental regulators by Polycomb in human embryonic stem cells. Cell

125, 301–313.

Lieberman-Aiden, E., van Berkum, N.L., Williams, L., Imakaev, M., Ragoczy,

T., Telling, A., Amit, I., Lajoie, B.R., Sabo, P.J., Dorschner, M.O., et al.

(2009). Comprehensive mapping of long-range interactions reveals folding

principles of the human genome. Science 326, 289–293.

Lienert, F., Mohn, F., Tiwari, V.K., Baubec, T., Roloff, T.C., Gaidatzis, D., Sta-

dler, M.B., and Schubeler, D. (2011). Genomic prevalence of heterochromatic

H3K9me2 and transcription do not discriminate pluripotent from terminally

differentiated cells. PLoS Genet. 7, e1002090.

Lister, R., Pelizzola, M., Dowen, R.H., Hawkins, R.D., Hon, G., Tonti-Filippini,

J., Nery, J.R., Lee, L., Ye, Z., Ngo, Q.M., et al. (2009). Human DNAmethylomes

at base resolution show widespread epigenomic differences. Nature 462,

315–322.

Lunyak, V.V., Prefontaine, G.G., Nunez, E., Cramer, T., Ju, B.G., Ohgi, K.A.,

Hutt, K., Roy, R., Garcıa-Dıaz, A., Zhu, X., et al. (2007). Developmentally regu-

lated activation of a SINE B2 repeat as a domain boundary in organogenesis.

Science 317, 248–251.

Marchetti, M., Fanti, L., Berloco, M., and Pimpinelli, S. (2003). Differential

expression of the Drosophila BX-C in polytene chromosomes in cells of larval

1282 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

fat bodies: a cytological approach to identifying in vivo targets of the homeotic

Ubx, Abd-A and Abd-B proteins. Development 130, 3683–3689.

Marks, H., Chow, J.C., Denissov, S., Francoijs, K.J., Brockdorff, N., Heard, E.,

and Stunnenberg, H.G. (2009). High-resolution analysis of epigenetic changes

associated with X inactivation. Genome Res. 19, 1361–1373.

Mateos-Langerak, J., Bohn, M., de Leeuw, W., Giromus, O., Manders, E.M.,

Verschure, P.J., Indemans, M.H., Gierman, H.J., Heermann, D.W., van Driel,

R., and Goetze, S. (2009). Spatially confined folding of chromatin in the inter-

phase nucleus. Proc. Natl. Acad. Sci. USA 106, 3812–3817.

Mendenhall, E.M., Koche, R.P., Truong, T., Zhou, V.W., Issac, B., Chi, A.S., Ku,

M., and Bernstein, B.E. (2010). GC-rich sequence elements recruit PRC2 in

mammalian ES cells. PLoS Genet. 6, e1001244.

Meuleman, W., Peric-Hupkes, D., Kind, J., Beaudry, J.B., Pagie, L., Kellis, M.,

Reinders, M., Wessels, L., and van Steensel, B. (2012). Constitutive nuclear

lamina-genome interactions are highly conserved and associated with A/T-

rich sequence. Genome Res. 23, 270–280.

Mirny, L.A. (2011). The fractal globule as a model of chromatin architecture in

the cell. Chromosome Res. 19, 37–51.

Montavon, T., Soshnikova, N., Mascrez, B., Joye, E., Thevenet, L., Splinter, E.,

de Laat, W., Spitz, F., and Duboule, D. (2011). A regulatory archipelago

controls Hox genes transcription in digits. Cell 147, 1132–1145.

Morey, L., and Helin, K. (2010). Polycomb group protein-mediated repression

of transcription. Trends Biochem. Sci. 35, 323–332.

Morey, C., Da Silva, N.R., Perry, P., and Bickmore, W.A. (2007). Nuclear reor-

ganisation and chromatin decondensation are conserved, but distinct, mech-

anisms linked to Hox gene activation. Development 134, 909–919.

Morey, C., Kress, C., and Bickmore, W.A. (2009). Lack of bystander activation

shows that localization exterior to chromosome territories is not sufficient to

up-regulate gene expression. Genome Res. 19, 1184–1194.

Negre, N., Brown, C.D., Shah, P.K., Kheradpour, P., Morrison, C.A., Henikoff,

J.G., Feng, X., Ahmad, K., Russell, S., White, R.A., et al. (2010). A comprehen-

sive map of insulator elements for the Drosophila genome. PLoS Genet. 6,

e1000814.

Nemeth, A., Conesa, A., Santoyo-Lopez, J., Medina, I., Montaner, D., Peterfia,

B., Solovei, I., Cremer, T., Dopazo, J., and Langst, G. (2010). Initial genomics of

the human nucleolus. PLoS Genet. 6, e1000889.

Noordermeer, D., Branco, M.R., Splinter, E., Klous, P., van Ijcken, W., Swage-

makers, S., Koutsourakis, M., van der Spek, P., Pombo, A., and de Laat, W.

(2008). Transcription and chromatin organization of a housekeeping gene

cluster containing an integrated beta-globin locus control region. PLoS Genet.

4, e1000016.

Noordermeer, D., deWit, E., Klous, P., van deWerken, H., Simonis, M., Lopez-

Jones, M., Eussen, B., de Klein, A., Singer, R.H., and de Laat, W. (2011a).

Variegated gene expression caused by cell-specific long-range DNA interac-

tions. Nat. Cell Biol. 13, 944–951.

Noordermeer, D., Leleu, M., Splinter, E., Rougemont, J., De Laat, W., and Du-

boule, D. (2011b). The dynamic architecture of Hox gene clusters. Science

334, 222–225.

Nora, E.P., Lajoie, B.R., Schulz, E.G., Giorgetti, L., Okamoto, I., Servant, N.,

Piolot, T., van Berkum, N.L., Meisig, J., Sedat, J., et al. (2012). Spatial partition-

ing of the regulatory landscape of the X-inactivation centre. Nature 485,

381–385.

Orlando, V., and Paro, R. (1993). Mapping Polycomb-repressed domains in the

bithorax complex using in vivo formaldehyde cross-linked chromatin. Cell 75,

1187–1198.

Palstra, R.J., Simonis, M., Klous, P., Brasset, E., Eijkelkamp, B., and de Laat,

W. (2008). Maintenance of long-range DNA interactions after inhibition of

ongoing RNA polymerase II transcription. PLoS One 3, e1661.

Pauler, F.M., Sloane, M.A., Huang, R., Regha, K., Koerner, M.V., Tamir, I.,

Sommer, A., Aszodi, A., Jenuwein, T., and Barlow, D.P. (2009). H3K27me3

forms BLOCs over silent genes and intergenic regions and specifies a histone

banding pattern on a mouse autosomal chromosome. Genome Res. 19,

221–233.

Peric-Hupkes, D., Meuleman, W., Pagie, L., Bruggeman, S.W., Solovei, I.,

Brugman, W., Graf, S., Flicek, P., Kerkhoven, R.M., van Lohuizen, M., et al.

(2010). Molecular maps of the reorganization of genome-nuclear lamina inter-

actions during differentiation. Mol. Cell 38, 603–613.

Pickersgill, H., Kalverda, B., de Wit, E., Talhout, W., Fornerod, M., and van

Steensel, B. (2006). Characterization of the Drosophila melanogaster genome

at the nuclear lamina. Nat. Genet. 38, 1005–1014.

Pope, B.D., Chandra, T., Buckley, Q., Hoare, M., Ryba, T., Wiseman, F.K.,

Kuta, A., Wilson, M.D., Odom, D.T., and Gilbert, D.M. (2012). Replication-

timing boundaries facilitate cell-type and species-specific regulation of a rear-

ranged human chromosome in mouse. Hum. Mol. Genet. 21, 4162–4170.

Prokocimer, M., Davidovich, M., Nissim-Rafinia, M., Wiesel-Motiuk, N., Bar,

D.Z., Barkan, R., Meshorer, E., and Gruenbaum, Y. (2009). Nuclear lamins:

key regulators of nuclear structure and activities. J. Cell. Mol. Med. 13,

1059–1085.

Raab, J.R., Chiu, J., Zhu, J., Katzman, S., Kurukuti, S., Wade, P.A., Haussler,

D., and Kamakaka, R.T. (2012). Human tRNA genes function as chromatin

insulators. EMBO J. 31, 330–350.

Ram, O., Goren, A., Amit, I., Shoresh, N., Yosef, N., Ernst, J., Kellis, M., Gym-

rek, M., Issner, R., Coyne, M., et al. (2011). Combinatorial patterning of chro-

matin regulators uncovered by genome-wide location analysis in human cells.

Cell 147, 1628–1639.

Reddy, K.L., Zullo, J.M., Bertolino, E., and Singh, H. (2008). Transcriptional

repression mediated by repositioning of genes to the nuclear lamina. Nature

452, 243–247.

Ryba, T., Hiratani, I., Lu, J., Itoh, M., Kulik, M., Zhang, J., Schulz, T.C., Robins,

A.J., Dalton, S., and Gilbert, D.M. (2010). Evolutionarily conserved replication

timing profiles predict long-range chromatin interactions and distinguish

closely related cell types. Genome Res. 20, 761–770.

Schoenfelder, S., Sexton, T., Chakalova, L., Cope, N.F., Horton, A., Andrews,

S., Kurukuti, S., Mitchell, J.A., Umlauf, D., Dimitrova, D.S., et al. (2010). Prefer-

ential associations between co-regulated genes reveal a transcriptional inter-

actome in erythroid cells. Nat. Genet. 42, 53–61.

Schwaiger, M., Stadler, M.B., Bell, O., Kohler, H., Oakeley, E.J., and Schub-

eler, D. (2009). Chromatin state marks cell-type- and gender-specific replica-

tion of the Drosophila genome. Genes Dev. 23, 589–601.

Schwaiger, M., Kohler, H., Oakeley, E.J., Stadler, M.B., and Schubeler, D.

(2010). Heterochromatin protein 1 (HP1) modulates replication timing of the

Drosophila genome. Genome Res. 20, 771–780.

Schwartz, Y.B., Kahn, T.G., Stenberg, P., Ohno, K., Bourgon, R., and Pirrotta,

V. (2010). Alternative epigenetic chromatin states of polycomb target genes.

PLoS Genet. 6, e1000805.

Schwartz, Y.B., Linder-Basso, D., Kharchenko, P.V., Tolstorukov, M.Y., Kim,

M., Li, H.B., Gorchakov, A.A., Minoda, A., Shanower, G., Alekseyenko, A.A.,

et al. (2012). Nature and function of insulator protein binding sites in the

Drosophila genome. Genome Res. 22, 2188–2198.

Sexton, T., Yaffe, E., Kenigsberg, E., Bantignies, F., Leblanc, B., Hoichman,

M., Parrinello, H., Tanay, A., and Cavalli, G. (2012). Three-dimensional folding

and functional organization principles of the Drosophila genome. Cell 148,

458–472.

Shopland, L.S., Lynch, C.R., Peterson, K.A., Thornton, K., Kepper, N., Hase,

Jv., Stein, S., Vincent, S., Molloy, K.R., Kreth, G., et al. (2006). Folding and

organization of a contiguous chromosome region according to the gene distri-

bution pattern in primary genomic sequence. J. Cell Biol. 174, 27–38.

Simonis, M., Klous, P., Splinter, E., Moshkin, Y., Willemsen, R., de Wit, E., van

Steensel, B., and de Laat, W. (2006). Nuclear organization of active and inac-

tive chromatin domains uncovered by chromosome conformation capture-on-

chip (4C). Nat. Genet. 38, 1348–1354.

Splinter, E., de Wit, E., Nora, E.P., Klous, P., van de Werken, H.J., Zhu, Y.,

Kaaij, L.J., van Ijcken, W., Gribnau, J., Heard, E., and de Laat, W. (2011).

The inactive X chromosome adopts a unique three-dimensional conformation

that is dependent on Xist RNA. Genes Dev. 25, 1371–1383.

Sutherland, H., and Bickmore, W.A. (2009). Transcription factories: gene

expression in unions? Nat. Rev. Genet. 10, 457–466.

Thomson, I., Gilchrist, S., Bickmore, W.A., and Chubb, J.R. (2004). The radial

positioning of chromatin is not inherited through mitosis but is established de

novo in early G1. Curr. Biol. 14, 166–172.

Tolhuis, B., deWit, E., Muijrers, I., Teunissen, H., Talhout, W., van Steensel, B.,

and van Lohuizen, M. (2006). Genome-wide profiling of PRC1 and PRC2 Poly-

comb chromatin binding in Drosophila melanogaster. Nat. Genet. 38, 694–699.

Tolhuis, B., Blom, M., Kerkhoven, R.M., Pagie, L., Teunissen, H., Nieuwland,

M., Simonis, M., de Laat, W., van Lohuizen, M., and van Steensel, B. (2011).

Interactions among Polycomb domains are guided by chromosome architec-

ture. PLoS Genet. 7, e1001343.

Towbin, B.D., Gonzalez-Aguilera, C., Sack, R., Gaidatzis, D., Kalck, V.,

Meister, P., Askjaer, P., and Gasser, S.M. (2012). Step-wise methylation of

histone H3K9 positions heterochromatin at the nuclear periphery. Cell 150,

934–947.

Umbarger, M.A., Toro, E., Wright, M.A., Porreca, G.J., Bau, D., Hong, S.H.,

Fero, M.J., Zhu, L.J., Marti-Renom, M.A., McAdams, H.H., et al. (2011). The

three-dimensional architecture of a bacterial genome and its alteration by

genetic perturbation. Mol. Cell 44, 252–264.

van Bemmel, J.G., Pagie, L., Braunschweig, U., Brugman, W., Meuleman, W.,

Kerkhoven, R.M., and van Steensel, B. (2010). The insulator protein SU(HW)

fine-tunes nuclear lamina interactions of the Drosophila genome. PLoS ONE

5, e15013.

van Koningsbruggen, S., Gierlinski, M., Schofield, P., Martin, D., Barton, G.J.,

Ariyurek, Y., den Dunnen, J.T., and Lamond, A.I. (2010). High-resolution

whole-genome sequencing reveals that specific chromatin domains from

most human chromosomes associate with nucleoli. Mol. Biol. Cell 21, 3735–

3748.

van Steensel, B., Delrow, J., and Henikoff, S. (2001). Chromatin profiling using

targeted DNA adenine methyltransferase. Nat. Genet. 27, 304–308.

Vaquerizas, J.M., Suyama, R., Kind, J., Miura, K., Luscombe, N.M., and Akh-

tar, A. (2010). Nuclear pore proteins nup153 and megator define transcription-

ally active regions in the Drosophila genome. PLoS Genet. 6, e1000846.

Vogel, M.J., Guelen, L., de Wit, E., Peric-Hupkes, D., Loden, M., Talhout, W.,

Feenstra, M., Abbas, B., Classen, A.K., and van Steensel, B. (2006). Human

heterochromatin proteins form large domains containing KRAB-ZNF genes.

Genome Res. 16, 1493–1504.

Vogelmann, J., Valeri, A., Guillou, E., Cuvier, O., and Nollmann, M. (2011).

Roles of chromatin insulator proteins in higher-order chromatin organization

and transcription regulation. Nucleus 2, 358–369.

Wang, K.C., Yang, Y.W., Liu, B., Sanyal, A., Corces-Zimmerman, R., Chen, Y.,

Lajoie, B.R., Protacio, A., Flynn, R.A., Gupta, R.A., et al. (2011). A long noncod-

ing RNA maintains active chromatin to coordinate homeotic gene expression.

Nature 472, 120–124.

Wen, B., Wu, H., Shinkai, Y., Irizarry, R.A., and Feinberg, A.P. (2009). Large

histone H3 lysine 9 dimethylated chromatin blocks distinguish differentiated

from embryonic stem cells. Nat. Genet. 41, 246–250.

Williamson, I., Hill, R.E., and Bickmore, W.A. (2011). Enhancers: from develop-

mental genetics to the genetics of common human disease. Dev. Cell 21,

17–19.

Wood, A.J., Lo, T.W., Zeitler, B., Pickle, C.S., Ralston, E.J., Lee, A.H., Amora,

R., Miller, J.C., Leung, E., Meng, X., et al. (2011). Targeted genome editing

across species using ZFNs and TALENs. Science 333, 307.

Wurtele, H., and Chartrand, P. (2006). Genome-wide scanning of HoxB1-asso-

ciated loci in mouse ES cells using an open-ended Chromosome Conforma-

tion Capture methodology. Chromosome Res. 14, 477–495.

Xie, W., Barr, C.L., Kim, A., Yue, F., Lee, A.Y., Eubanks, J., Dempster, E.L., and

Ren, B. (2012). Base-resolution analyses of sequence and parent-of-origin

dependent DNA methylation in the mouse genome. Cell 148, 816–831.

Yaffe, E., and Tanay, A. (2011). Probabilistic modeling of Hi-C contact maps

eliminates systematic biases to characterize global chromosomal architec-

ture. Nat. Genet. 43, 1059–1065.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1283

Yokochi, T., Poduch, K., Ryba, T., Lu, J., Hiratani, I., Tachibana,M., Shinkai, Y.,

and Gilbert, D.M. (2009). G9a selectively represses a class of late-replicating

genes at the nuclear periphery. Proc. Natl. Acad. Sci. USA 106, 19363–19368.

Yokota, H., Singer, M.J., van den Engh, G.J., and Trask, B.J. (1997). Regional

differences in the compaction of chromatin in human G0/G1 interphase nuclei.

Chromosome Res. 5, 157–166.

Yokota, H., van den Engh, G., Hearst, J.E., Sachs, R.K., and Trask, B.J. (1995).

Evidence for the organization of chromatin in megabase pair-sized loops ar-

ranged along a random walk path in the human G0/G1 interphase nucleus.

J. Cell Biol. 130, 1239–1249.

1284 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Zeng, W., Ball, A.R., Jr., and Yokomori, K. (2010). HP1: heterochromatin

binding proteins working the genome. Epigenetics 5, 287–292.

Zhang, Y., McCord, R.P., Ho, Y.J., Lajoie, B.R., Hildebrand, D.G., Simon, A.C.,

Becker, M.S., Alt, F.W., and Dekker, J. (2012). Spatial organization of the

mouse genome and its role in recurrent chromosomal translocations. Cell

148, 908–921.

Zullo, J.M., Demarco, I.A., Pique-Regi, R., Gaffney, D.J., Epstein, C.B., Spoo-

ner, C.J., Luperchio, T.R., Bernstein, B.E., Pritchard, J.K., Reddy, K.L., and

Singh, H. (2012). DNA sequence-dependent compartmentalization and

silencing of chromatin at the nuclear lamina. Cell 149, 1474–1487.

Leading Edge

Review

CTCF and Cohesin: Linking GeneRegulatory Elements with Their Targets

Matthias Merkenschlager1,* and Duncan T. Odom2,*1Lymphocyte Development Group, MRC Clinical Sciences Centre, Imperial College London, Du Cane Road, London W12 0NN, UK2Cancer Research UK Cambridge Institute, University of Cambridge, Robinson Way, Cambridge CB2 0RE, UK*Correspondence: matthias.merkenschlager@csc.mrc.ac.uk (M.M.), duncan.odom@cruk.cam.ac.uk (D.T.O.)

http://dx.doi.org/10.1016/j.cell.2013.02.029

Current epigenomics approaches have facilitated the genome-wide identification of regulatoryelements based on chromatin features and transcriptional regulator binding and have begun tomap long-range interactions between regulatory elements and their targets. Here, we focus onthe emerging roles of CTCF and the cohesin in coordinating long-range interactions between regu-latory elements. We discuss how species-specific transposable elementsmay influence such inter-actions by remodeling the CTCF binding repertoire and suggest that cohesin’s association withenhancers, promoters, and sites defined by CTCF binding has the potential to form developmen-tally regulated networks of long-range interactions that reflect and promote cell-type-specific tran-scriptional programs.

IntroductionMammalian genomes are vast and are composed of billions of

bases of DNA containing sufficient regulatory information to

create complex organisms with thousands of cell types and

considerable behavioral repertoires. Each of the twenty-odd

thousand genes in the human genome likely has many distinct

regulatory regions spread across tens to hundreds of kilobases

that operate in concert to accurately instruct when, where, and

how much of each gene to transcribe. Here, we focus on the

role of the sequence-specific DNA-binding protein CTCF

(CCCTC-binding factor) and the multiprotein cohesin complex

in orchestrating tissue-specific gene regulation in an evolu-

tionary context.

The Interplay between Regulatory Elements andChromatin Directs Tissue-Specific TranscriptionTissue-specific transcription of protein-coding genes is

controlled by one or more small regulatory regions that contain

sets of DNA-binding proteins, which occupy DNA in a combina-

torial fashion (Lee et al., 2002; Odom et al., 2006; see Box 1

for a brief overview of regulatory elements). The DNA itself is

coiled around nucleosomes that are composed of histone

octamers and convey regulatory information by their position

and in the form of posttranslational histone modifications

(Segal and Widom, 2009; Campos and Reinberg, 2009). Regula-

tory regions are combined by through-space interactions to

finalize both assembly and control of basal transcriptional

machineries (Lee et al., 2002; Sanyal et al., 2012; Handoko

et al., 2011; Li et al., 2012; Dixon et al., 2012; Noordermeer

et al., 2011).

How to Recognize Regulatory Elements

Regulatory elements can be identified genome wide using indi-

rect and direct means. The successful application of sequence

conservation among mammals has been instrumental in identi-

fying the complete protein-coding complement of mammals

(Church et al., 2009). More recently, the same strategy has

been taken to identify the sequences in the genome under selec-

tive pressure, presumably by DNA-binding proteins and other

noncovalent regulators (Lindblad-Toh et al., 2011). One notable

success of this strategy includes the identification of thousands

of highly conserved CTCF binding locations that appear shared

among most mammalian species. However, a considerable

fraction of the regulation of the genome seems to occur in

a highly species-specific manner due to the rapid evolution of

tissue-specific transcription factor binding, indicating that direct

comparison of genome sequences has inherent limitations

(Kunarso et al., 2010; Schmidt et al., 2010b).

Recent advances in sequencing technology have facilitated

the genome-wide identification of regulatory elements based

on chromatin accessibility, posttranslational histone modifica-

tions, and the binding of regulatory factors (Johnson et al.,

2007, reviewed in Noonan and McCallion, 2010; see Box 2

for a brief overview). These approaches have been integrated

across the human genome to generate a first pass encyclopedia

of the regulatory features that are functional in the human

genome (reviewed in Noonan and McCallion, 2010).

How Regulatory Elements Work

How do histone marks, chromatin state, and genome architec-

ture conspire to create gene expression programs? Local chro-

matin accessibility, transcription factor binding, and specific

chromatin modifications such as acetylation, methylation, phos-

phorylation, or ubiquitylation not only mark regulatory elements,

but they also actively contribute to the control of gene expres-

sion (Figure 1). Specifically, DNA methylation, histone modifica-

tions, and the proteins that interact with them affect the accessi-

bility of chromatin. These factors link chromatin marks to the

general transcription machinery (Thomas and Chiang, 2006),

including TBP-associated factors (TAFs) and Mediator, as well

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1285

Box 1. Types of Regulatory Elements in Mammalian Genomes

1. Promoters are located near genes and directly regulate tran-

scription.

2. Enhancers are located distal to protein-coding genes and may

require chromatin conformational placement near regulated loci.

3. Insulators for either gene expression or chromatin state divide

heterochromatin from euchromatin and/or active from inactive

gene expression domains.

4. Repressors can decrease the expression of nearby genes.

Box 2. Mapping Regulatory Elements and Their Interactions

DNase hypersensitivity reveals accessible sites that are not protected

by nucleosomes or DNA-binding proteins (Hesselberth et al., 2009) and

can be analyzed for transcription factor binding motifs, sequence

conservation, and the tissue-specific gene expression (Neph et al.,

2012).

The chromatin state and the binding of regulatory proteins can be

mapped by chromatin immunoprecipitation (ChIP) using antibodies

against covalent histone modifications, site-specific transcription

factors, and transcriptional cofactors (Bernstein et al., 2005; Suga-

numa and Workman, 2011; Visel et al., 2009; Odom et al., 2004). A

recent refinement is ChIP-exo, which can identify the exact bases to

which a factor is bound, as exemplified for CTCF (Rhee and Pugh,

2011).

Chromosome conformation capture interrogates chromatin interac-

tions based on crosslinked and ligated DNA (Dekker, 2008) on

a gene-specific, regional, or genome-wide scale (Dostie et al., 2006;

Lieberman-Aiden et al., 2009; Simonis et al., 2009). The output of

current chromosome conformation capture experiments represents

probability distributions within cell populations (Nora et al., 2012;

Sanyal et al., 2012).

FISH visualizes the proximity of specific sequences in individual cells

on a gene-specific level (Dostie and Bickmore, 2012) and on a genomic

scale (Boyle et al., 2011).

as RNA polymerase II (RNAPII) to regulate transcription initiation

and elongation (see Figure 1 and Box 3).

Classical models of transcriptional control begin with

transcription factors binding to DNA, recruiting nucleosome

remodeling complexes and histone modifying enzymes whose

products can then interact with basal machinery to drive tran-

scription. Although this is often true, in reality, these events are

mutually interdependent (Figure 2).

In summary, chromatin marks facilitate not only the cata-

loguing of genomic features, but more importantly, they also

link regulatory elements to downstream effectors of transcrip-

tional activation or repression.

Evolution and Gene Regulatory Strategies

In simple eukaryotes such as yeast, gene regulation is largely

controlled by elements immediately proximal to their target

genes (Borneman et al., 2006; Harbison et al., 2004; Lee et al.,

2002). More complex model organisms like Drosophila and

C. elegans have vastly larger genomes than yeast, yetmost regu-

latory regions remain relatively close to their target genes (He

et al., 2011; MacArthur et al., 2009). Population genetics anal-

yses suggest that large breeding population sizes drive the

condensation of regulatory and genic sequences (Lusk and

Eisen, 2010; Lynch, 2007). In effect, competitive pressures can

select for more efficient genome organization and utilization,

leading to subtle growth advantages that become dominant

only over thousands of generations in large breeding population

sizes.

In contrast, vertebrate species often have small breeding pop-

ulations and thus lack these genome-compressing effects.

Therefore, their genomes must have mechanisms to adapt to

the constant onslaught of selfish element expansion and genetic

drift. The resulting fragmentation of the genome has been

exploited by mammals to increase the possible regulatory

combinations of control elements in the interest of cell-type

and tissue-specific gene expression (Dunham et al., 2012).

Indeed, the organismal complexity found in mammals, and

more generally in vertebrates, may be a direct result of this frag-

mentation (Figure 3A).

The challenge that had to be overcome to create this opportu-

nity to expand organismal complexity, however, was how to

efficiently connect the now-diffuse regulatory sequences with

their targets.

Linking Regulatory Elements with Their Targets

Eukaryotes appear to have evolvedmolecular systems from pre-

existing cellular machineries to connect remote enhancers and

promoters. The components of these systems have been inves-

tigated to different degrees. The proteins that directly occupy

1286 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

DNA, such as CTCF and clusters of tissue-specific transcription

factors, create a protein landscape that can be more readily

handled biophysically than the substrate nucleic acid. Connect-

ing directly with these regulators is the multicomponent cohesin

complex, including the associated loading complex NIPBL-

MAU2. An interesting feature of the cohesin complex is its

ring-like shape with an internal diameter of �40 nm (Nasmyth

and Haering, 2009). This arrangement enables cohesin to handle

themolecular dimensions of chromatin fibers, as illustrated by its

ability to mediate the interactions of sister chromatids (reviewed

in Nasmyth and Haering, 2009; Skibbens, 2009). Additional

players in the process of organizing regulatory elements include

polycomb, which associates with repressive histone marks,

SATB1 and -2, and the nuclear lamina (Galande et al., 2007;

Morey and Helin, 2010; van Steensel, 2011).

Methods to interrogate chromatin interactions include chro-

mosome conformation capture (Dekker, 2008) and fluorescence

in situ hybridization (FISH) (Dostie and Bickmore, 2012) (see Box

3) and have shown that particular enhancers and promoters

interact in a nonrandom fashion. A depletion of contacts can

be used to suggest locations for insulating elements that divide

gene expression domains.

CTCF: The CCCTC-Binding FactorCTCF and cohesin are central players in regulating long-range

interactions. We briefly describe the scientific history of each,

the recent discovery that they colocalize and interact to control

long-range regulatory interactions, and what models may fit

our current understanding of their roles in tissue-specific tran-

scription.

Identification of CTCF as a Transcriptional Regulator

CTCFwas originally identified as a transcriptional regulator of the

c-myc oncogene (Baniahmad et al., 1990; Filippova et al., 1996;

Figure 1. Chromatin Modifications at Regulatory Elements—from Marks to FunctionModels for gene regulation have moved from an early focus on transcription factors and DNA to encompass the full context of chromatin (left). Regulatoryelements are marked by patterns of DNA methylation, histone marks, and interacting proteins that link chromatin modifications to the regulation of transcription(center). Regulatory elements are often separated by considerable distances in the linear sequence of metazoan genomes. Transcriptional control is thought toinvolve interactions between regulatory elements in three-dimensional nuclear space (right). To illustrate this, the figure depicts regulatory elements of theimprinted IGF2/H19 locus and their interactions as detailed in the section ‘‘CTCF and Cohesin Regulate Complex Loci.’’

Lobanenkov et al., 1990). This widely expressed 11 zinc finger

DNA-binding protein is conserved in most higher eukaryotes

(Klenova et al., 1993) and is essential for cellular function (Burcin

et al., 1997; Fedoriw et al., 2004).

One of the most interesting aspects of CTCF is that it appears

to be the unique, major DNA binding component that establishes

vertebrate insulators (Bell et al., 1999). CTCF can block enhancer

function when placed between enhancers and promoters in

reporter plasmids, and most—if not all—CTCF binding sites

can serve as ‘‘insulators’’ in such constructs (Giles et al., 2010;

Phillips and Corces, 2009). Demonstrating enhancer blocking

function of CTCF sites in their native chromatin context is

much more difficult. A well-studied case is the IGF2/H19 locus,

where CTCF binding controls the functional interaction of the

IGF2 and H19 promoters with a distal enhancer, as supported

by the analysis of natural (Beygo et al., 2013) and engineered

mutations (discussed in detail below).

The functions of other CTCF sites have been probed by the

deletion of specific CTCF sites from the mouse immunoglobulin

heavy chain locus (Guo et al., 2011) and the insertion of ectopic

CTCF sites into the T cell receptor b chain locus (Shrimali et al.,

2012). Such site-specific experiments are complemented by

loss-of-function approaches in which CTCF is genetically

deleted (Ribeiro de Almeida et al., 2011; Hirayama et al., 2012).

Correlative studies link the position of CTCF binding sites to

long-range interactions by chromatin conformation assays

(Dixon et al., 2012; Sanyal et al., 2012) and the analysis of chro-

matin features. CTCF binding is often found at transitions

between distinct chromatin states as marked by histone modifi-

cations (Cuddapah et al., 2009) or interactions with the nuclear

lamina (van Steensel, 2011), supporting the notion that, in addi-

tion to limiting the ‘‘reach’’ of regulatory elements, CTCF can

form ‘‘boundaries.’’ However, based on these data, only a minor

fraction of CTCF sites appears to demarcate chromatin bound-

aries in their native chromatin context in vivo (Dixon et al.,

2012; Schmidt et al., 2012b). This suggests that plasmid-based

reporter constructs may not accurately capture the native chro-

matin environment, which is crucial to integrate regulatory

inputs.

Remarkably, CTCF may also help regulate viral genomes

(Holdorf et al., 2011; Stedman et al., 2008; Tempera et al.,

2010). CTCF interacts with specific locations in numerous viral

genomes, including EBV and murine and human herpes viruses

(Stedman et al., 2008; Stevens et al., 2012). These interactions

are functional, and CTCF regulates both individual viral genes

as well as entire programs; for instance, viral latency is influ-

enced by CTCF (Hughes et al., 2012; Kang et al., 2011). In the

case of Kaposi’s sarcoma-associated herpes virus, cooperation

between CTCF and cohesin has been documented (Chen et al.,

2012; Kang et al., 2011; Stedman et al., 2008).

As a host protein that can directly control viral gene expres-

sion, CTCF links the mammalian host’s defenses and gene

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1287

Box 3. Chromatin Modifications: From Marks to Function

Enhancers display monomethylated lysine 4 (H3K4me1) together with

acetylated lysine 27 on histone H3 (H3K27ac) in the active state and

trimethylated lysine 27 on histone H3 (H3K27me3) in the repressed

state, whereas promoters are marked by trimethylation of histone H3

at lysine 4 (H4K4me3) (reviewed by Bannister and Kouzarides, 2011).

H3K4 methylation marks are established by the SET1 and mixed

lineage leukemia (MLL) family of histone methyltransferases. Among

the readers of di- and trimethylated H3K4 are PHD (plant homeodo-

main) finger proteins (Bannister and Kouzarides, 2011)—for example,

the TAF3 subunit of the general transcription machinery for RNAPII.

The targeting of TFIID connects the promoter mark H3K4me3 to the

initiation of transcription (Vermeulen et al., 2007). Other readers of

H3K4 methylation include the V(D)J recombinase subunit RAG2, chro-

matin modifiers, and remodeling factors.

In contrast to H3K4, H3K27 is methylated by the polycomb repressive

complex 2 (PRC2). This modification recruits PRC1, a polycomb

complex that blocks RNA polymerase and mediates transcriptional

repression (Morey and Helin, 2010). Trimethylated H3K9 is a mark of

inaccessible chromatin, or heterochromatin. Among the readers of

H3K9me3 is the chromodomain protein heterochromatin protein 1

(HP1), which propagates the formation of inaccessible chromatin

(Bannister and Kouzarides, 2011).

Most CG dinucleotides in mammalian genomes are targeted by DNA

metyltransferases that modify cytosine residues. CG-rich promoter-

proximal sequences (CpG islands) are specifically protected from

DNA methylation by the binding of Cfp1 (Thomson et al., 2010).

Readers of DNA methylation include methyl-binding proteins such as

MECP2 (mutated in Rett syndrome; Guy et al., 2007).

Histone acetylation is linked to the transcriptional machinery by bro-

modomain proteins such as the BET protein BRD4, which interact

with the Mediator complex and transcription elongation factors.

Mediator regulates transcription by bridging sequence-specific DNA-

binding proteins with RNAPII (Conaway and Conaway, 2011; Soutour-

ina et al., 2011), and elongation factors facilitate transcription by

promoting Pol II processivity (Yang et al., 2005; Jang et al., 2005).

regulatory elements with the regulation, function, and pathoge-

nicity of the virus genome. In the next section, we explore how

CTCF interacts with and potentially controls another form of

parasite: transposable elements.

Insights from the Genome-wide Analysis of CTCF

Binding

Early genome-scale mapping of CTCF binding in mammalian

cells revealed a large, information-rich motif and mostly tissue-

independent binding preferentially to gene-dense regions but

with little or no enrichment in promoters (Kim et al., 2007). A

substantial minority of CTCF sites may be cell-type specific,

particularly in cancer-derived cell lines in which differential

binding correlates with differential DNA methylation (Wang

et al., 2012; see legend of Figure 2 for discussion). High conser-

vation of CTCF binding was predicted and later demonstrated by

both comparative (Lindblad-Toh et al., 2011) and experimental

(Kunarso et al., 2010; Schmidt et al., 2012a) approaches.

A series of simultaneous papers reported that CTCF and cohe-

sin co-occupy the genome (Parelho et al., 2008; Rubio et al.,

2008; Stedman et al., 2008; Wendt et al., 2008). This observation

provided a functional link between an extremely high-affinity

DNA-binding protein (CTCF) and cohesin, a key component of

1288 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

chromatin. Cohesin had long been known to connect sister chro-

matids, a function that strongly suggested that cohesin may play

a similar role in connecting chromatin domain loops within one

chromosome, whichmight also be anchored by CTCF. The inter-

action of cohesin and CTCF may explain how CTCF acts in

specific locations as a domain boundary for chromatin states.

Evolution of CTCF Binding

A link between CTCF and SINE repeat elements was reported in

a seminal paper that dissected transcription factor binding

profiles based on their association with repetitive elements

(Bourque et al., 2008). Thousands of B2 SINE elements in the

mouse genome carry a CTCF binding motif, and a significant

proportion of these motifs are bound by CTCF in vivo. It was

postulated that embedding a long, complex CTCF motif into

such a repeat represented a powerful mechanism for rapidly

expanding the CTCF regulatory repertoire; such a mechanism

had been previously suggested for REST/NRSF, a repressor

that targets a similarly large and complex motif (Mortazavi

et al., 2006).

Most recently, by comparing the in vivo genomic occupancy of

CTCF in six mammalian species, it was discovered that the SINE

repeats currently active in at least three of four major mammalian

lineages carry a high-affinity CTCF site (Schmidt et al., 2012a).

Indeed, hundreds to many thousands of SINE-expanded CTCF

sites were identified in dog, gray short-tailed opossum, and rat,

as well as in mice (Figure 3A). The comparison of the sequences

surrounding the most ancient, highly conserved CTCF binding

sites reveals over a hundred fossilized SINE repeat sequences

in multiple species of mammals, separated by up to 180 MY of

evolution. Thus, the repeat-driven expansion of CTCF binding

sites is an ancientmechanismof genome evolution. Interestingly,

neither human nor macaque show evidence of recent repeat-

associated expansion of SINEs, suggesting that primates may

have escaped this mode of genome remodeling. In contrast,

rodents show massive SINE element expansion; however, the

relative transposon activity of the SINE B2 elements that carry

CTCF is considerably accelerated in mouse versus rat. In the

time since their most recent common ancestor, almost four times

more SINE B2 insertions carrying CTCF have occurred inmouse.

The comparison of the thousands of CTCF binding events

occurring in mouse at SINE B2 repeat elements with the similar

number of CTCF binding sites that are deeply shared in mamma-

lian evolution revealed that both types of CTCF binding function

as transcriptional and chromatin insulators (Schmidt et al.,

2012a).

CTCF Binding as a Potential Survival Strategy

for Expanding Repeats

Analysis of the mechanisms by which mammals epigenetically

silence repeats such as SINE elements suggests possible bene-

fits that repeat elements might obtain by carrying binding motifs

for a transcriptional regulator.

The first possible advantage is that CTCF binding may modu-

late DNA methylation, which could otherwise silence transpo-

sons (Wang et al., 2012). It has been suggested that mammalian

genomes defend against the large burden of transposons and

endogenous retroviruses by methylating cytosines in DNA

(Walsh and Bestor, 1999), which leads to comparatively rapid

decay of these sequences via a C to T transposition (Bird,

Figure 2. Interdependence of Genome Sequence, Chromatin, and TranscriptionDNA sequence and local chromatin structure direct transcription factor (TF) binding. Some TFs harness cofactors to remodel chromatin, and others require openchromatin. The pluripotency factors Oct4, Sox2, and Klf4 predominantly engage ‘‘closed’’ distal regulatory elements during somatic cell reprogramming, whereasMyc prefers open chromatin (Soufi et al., 2012). Lineage-specific TFs often rely on pre-existing permissive chromatin; Foxp3, T-bet, and RORgt are induced inspecialized T cell subsets and bind pre-existing regulatory elements (Ciofani et al., 2012; Samstein et al., 2012; Vahedi et al., 2012).TF binding can be indifferent to DNA methylation (Bell et al., 2011), but CTCF prefers hypomethylated DNA (Bell and Felsenfeld, 2000; Hark et al., 2000; Kanduriet al., 2000; Wang et al., 2012). This relationship is reciprocal, as CTCF can influence the methylation status of distal regulatory regions (Stadler et al., 2011),blurring cause and effect of preferential binding to hypomethylated DNA. The differential methylation of CG-rich sites can exclude CTCF, allowing for the parent-of-origin-specific (imprinted) regulation of IGF2/H19 (Bell and Felsenfeld, 2000; Hark et al., 2000; Kanduri et al., 2000). CTCFL, a paralog of CTCF, binds DNAirrespective of methylation (Nguyen et al., 2008).TF binding recruits chromatin remodelers and chromatin-modifying enzymes (‘‘writers’’ and ‘‘erasers’’) that modify histones or methylate DNA. Chromatinmodifications can limit recombination between repetitive regions of the genome and impact the activity of transposable elements that drive genome evolution,including the evolution of regulatory elements.DNA methylation and histone modifications are recognized by chromatin ‘‘readers’’ that link chromatin modifications to the transcription machinery (Box 3).Transcription alters the chromatin structure of transcribed regions; at lymphocyte receptor loci, transcription drives rearrangement by depositing H3K4me3,which is recognized by the PHD finger of Rag2, a component of the V(D)J recombination machinery.As the process of transcription itself, RNA transcripts can impact the chromatin landscape. Repeat-associated transcripts activate RNA interference mecha-nisms that modify chromatin and control transposable elements (Fedoroff, 2012). In mammals, this mechanism appears restricted to germ cells (Siomi and Siomi,2011). Long noncoding RNAs also regulate chromatin structure and gene expression, as exemplified by Xist, which mediates X chromosome inactivation infemale mammals (Brockdorff, 2011).Cohesin is recruited to active genes alongside TFs and the basal transcription machinery (Schmidt et al., 2010a; Kagey et al., 2010) and in turn can facilitate TFbinding to low-affinity sites (Faure et al., 2012).

1980). Indeed, a substantial fraction of cytosine methylation in

mammalian genomes is found in transposable elements; CTCF

binding occurs in hypomethylated regions, thus partially protect-

ing surrounding genetic sequences from methylation (see also

Cohen et al., 2011).

A second possible advantage is in modulating the RNAi-medi-

ated control of transposable elements in somatic cells or in the

germline (Fedoroff, 2012; Siomi and Siomi, 2011) (Figure 2).

Duplication either of entire genomes or of genomic regions

results in repeated genomic information and the danger of

illegitimate recombination, and RNAi may have facilitated the

expansion of higher eukaryotic genomes by limiting the danger

of illegitimate recombination (Fedoroff, 2012). Once controlled,

duplications can diversify to drive the evolution of genes, gene

regulatory elements, and the factors that bind them (for example,

Boris, a CTCF paralog active in germline cells [Loukinov et al.,

2002]).

Epigenetic mechanisms such as DNA methylation and RNAi

have facilitated the domestication of transposable elements,

which in turn has enabled the genomes of higher eukaryotes to

accommodate vast numbers of transposable elements. These

transposable elements have been repurposed to build centro-

meres and telomeres (Lopez-Flores and Garrido-Ramos,

2010), to remodel genome and regulatory architectures (Kunarso

et al., 2010; Schmidt et al., 2012b), and to rearrange immune

receptor loci (Schatz, 2004).

CohesinCohesin Functions in the Cell Cycle

A strong candidate for mediating long-range interactions

between regulatory elements is cohesin, a multiprotein complex

that provides cohesion between sister chromatids from the time

of DNA replication in S phase until cell division (Nasmyth and

Haering, 2009). This function of cohesin enables postreplicative

DNA repair and proper chromosome segregation through

mitosis and meiosis and hence the integrity of genomic informa-

tion passed on from mother to daughter cells and from one

generation of multicellular organisms to the next. Unsurprisingly,

this function of cohesin is essential, and cohesin is highly

conserved through evolution (Nasmyth and Haering, 2009). At

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1289

Figure 3. The Evolution of CTCF Binding

Sites(A) Dispersal of CTCF binding sites by the activityof transposable elements. Both CTCF binding andthe mammalian genome itself are remodeled byrepeat elements. The ACSL6 locus with CTCFbinding and exon-to-exon homology mappingis shown for mouse and human. Intron sizeshave been extensively remodeled due to repeatelement expansions; in mouse, this expansionincludes the introduction of a CTCF binding sitecarried within a mouse SINE B3 repeat.(B) Retention of CTCF binding sites. A conservedCTCF binding site upstream of the Ifng locus ismaintained in rodent genomes despite the near-complete deletion of the associated gene, Il26 (redelements), and contributes to long-range interac-tions (Sekimata et al., 2009; Hadjur et al., 2009).

the heart of cohesin (as well as of the highly related condensin

and Smc5/6 complexes) are heterodimers of SMC (structural

maintenance of chromosomes) proteins. The V-shaped Smc1-

Smc3 heterodimer is complemented by Rad21/Scc1 and

Scc3/SA1/SA2 subunits to form a ring-like structure large

enough to topologically embrace two chromatin fibers (Nasmyth

and Haering, 2009).

Consistent with its role in postreplicative DNA repair and chro-

mosome segregation, cohesin is enriched at sites of DNA

damage (Strom et al., 2004; Unal et al., 2004) and at centromeres

(Nasmyth and Haering, 2009). In higher eukaryotes, cohesin is

a major component of chromatin also in noncycling and even

in postmitotic cells. This points to a role for cohesin outside of

the cell cycle, and indeed, there is growing evidence that cohesin

contributes to the regulation of chromatin structure and gene

expression in interphase.

Cohesin’s Emerging Roles in Gene Regulation

Initial evidence for a role of cohesin in gene regulation came from

genetic studies in Drosophila, in which the expression of specific

1290 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

homeobox genes is dependent on the

dosage of the cohesin loading factor

Nipped-B (Rollins et al., 1999). Heterozy-

gous mutations in NIPBL, the human

homologofNipped-B,were subsequently

found to cause the developmental

disorder Cornelia de Lange syndrome

(Strachan, 2005), and similar develop-

mental abnormalities are associated with

mutations in cohesin subunits (Strachan,

2005), cohesin cofactors (Zhang et al.,

2009), and cohesin-modifying enzymes

(Vega et al., 2005; Deardorff et al., 2012).

Although cultured cells derived from

NIPBL heterozygous individuals do not

showclear defects in chromosomesegre-

gation (Strachan, 2005), a distinction

between cell-division-related and cell-

division-independent cohesin functions

is required to support a direct link

between cohesin and gene expression.

This was first demonstrated by depleting

cohesin from postmitotic cells in Drosophila (Pauli et al., 2008,

2010; Schuldiner et al., 2008) and later in noncycling mouse

thymocytes (Seitan et al., 2011). Cohesin-depleted Drosophila

neurons show defective axon pruning as a result of deregulated

ecdyson receptor expression (Pauli et al., 2008, 2010; Schuldiner

et al., 2008). Genetic ablation of the Rad21 cohesin subunit in

mouse thymocytes impairs the transcription and rearrangement

of the developmentally regulated T cell receptor a locus and

disrupted thymocyte differentiation (Seitan et al., 2011).

Recent studies uncovered two distinct types of cohesin sites

that might mediate cohesin’s roles in gene regulation. Strong

cohesin sites usually coincide with the binding of CTCF (Wendt

et al., 2008; Parelho et al., 2008; Stedman et al., 2008; Rubio

et al., 2008), whereas numerous and often weaker cohesin sites

map to active promoters and enhancers (Schmidt et al., 2010a;

Kagey et al., 2010; Faure et al., 2012). Here, cohesin is colocal-

izedwith its loading factor Nipbl, withMediator components, and

with tissue-specific transcription factors (Schmidt et al., 2010a;

Kagey et al., 2010).

Figure 4. Cohesion and CTCF Link Regulatory Elements at the Tcra

LocusCohesin binding sites flank major regulatory elements of Tcra, the TEApromoter, and the Ea enhancer. Cohesin strengthens promoter-enhancerinteractions over a genomic distance of 80 kb, facilitating Tcra transcriptionand rearrangement of coding sequences (Seitan et al., 2011). A CTCF-dependent insulator separates the Ea enhancer from the housekeepinggene Dad1 (Magdinier et al., 2004; Zhong and Krangel, 1999). Cohesindepletion increases the transcription of Dad1 at the expense of Tcra (Seitanet al., 2011).

Box 4. Limitations of Current Experimental Approaches toUnderstanding Cohesin’s Role in Gene Expression

Schmidt et al. (2010a) correlated the binding of transcription factors

with cohesin recruitment but did not explore the biochemical mecha-

nisms that mediate this colocalization. They found that cohesin deple-

tion affects gene expression, but the interpretation of these data is

complicated by global shifts in gene expression. The authors dealt

with this issue by focusing on estrogen-responsive genes, but many

other gene expression changes remain to be explained.

Kagey et al. (2010) deprived ES cells of cohesin, a complex that is

essential for successful chromosome segregation in mitosis and for

other aspects of chromosome biology in cycling cells such as DNA

replication and postreplicative DNA repair. This resulted in the misex-

pression of most ES-cell-expressed genes. However, ES cells are

rapidly cycling, making it difficult to discern whether loss of cohesin

brought about changes in gene expression as a result of specific

gene regulatory functions or the activation of DNA damage check-

points. The authors deliberately limited the scope of their analysis by

focusing on the effects of knocking down cohesin, cohesin-loading

factors, and mediator subunits and by combining gene expression

data with genomic binding data. Nevertheless, it is important to

remember that the loss of cohesin from cycling cells can trigger

damage responses that may radically alter the pattern of gene expres-

sion and antagonizes the expression of pluripotency factors (Lin et al.,

2005).

A study by Seitan et al. (2011) largely avoids cell-cycle-related issues

but does make the assumption that cohesin-dependent enhancer-

promoter interactions are the ‘‘cause’’—rather than a correlate—of

defective transcription in cohesin-depleted cells. Studies on postmi-

totic cells in Drosophila provide the clearest dissociation to date

between cohesin functions in cycling and noncycling cells (Pauli

et al., 2008, 2010; Schuldiner et al., 2008) but provide little mechanistic

insight into how cohesin affects gene expression.

Cohesin Functions inGeneRegulation andDevelopmentMediating Chromosomal Long-Range Interactions

The demonstration of long-range interactions between cohesin

binding sites (Hadjur et al., 2009; Nativio et al., 2009; Kagey

et al., 2010; Seitan et al., 2011) suggested that cohesin may

affect gene expression by this mechanism. CTCF had long

been thought to contribute to the spatial organization of the

genome (Wallace and Felsenfeld, 2007), but a dependence of

CTCF-based long-range interactions on cohesin was first

demonstrated for the mouse Ifng locus (Hadjur et al., 2009). A

CTCF binding site 60–70 kb upstream of the Ifng coding region

is conserved in many mammals and is selectively retained in

rodent genomes, despite the near-complete deletion of the

associated gene, Il26 (Figure 3B). This site is preserved despite

the insertion of a long interspersed nuclear element (LINE)

at +57–59 kb and a long terminal repeat (LTR)–LINE–LTR

at +73–87 kb (Schoenborn et al., 2007) and complex structural

rearrangements and segmental duplications that disrupt synteny

with human over a region of �50 kb upstream of the Ifng coding

region (Schoenborn et al., 2007; She et al., 2008). In both human

and mouse, this site contacts two other CTCF sites, one in the

first intron of Ifng and the other about 100 kb downstream of

the locus (Sekimata et al., 2009; Hadjur et al., 2009). These

long-range interactions occur selectively in T helper 1 cells,

which inducibly express Ifng. CTCF and cohesin are both

required for these interactions. The contribution of the upstream

CTCF binding site suggests that the selective retention of this

site, despite the deletion of the associated Il26 locus, is function-

ally relevant for the regulation of Ifng (Sekimata et al., 2009;

Hadjur et al., 2009).

Cohesin depletion is linked to disrupted promoter-enhancer

interactions in embryonic stem (ES) cells (Kagey et al., 2010)

and in thymocytes (Seitan et al., 2011). Interactions mapped in

ES cells involve relatively short distances (3–5 kb; Kagey et al.,

2010), whereas deletion of the cohesin subunit Rad21 in noncy-

cling mouse thymocytes distorted the chromatin architecture

of the developmentally regulated T cell receptor a locus Tcra

over at least 80 kb. Interestingly, cohesin binding sites flank

major promoter and enhancer elements of Tcra, and cohesin

strengthens long-range promoter-enhancer interactions (Fig-

ure 4). This correlates with transcription and rearrangement of

the locus and, ultimately, thymocyte differentiation (Seitan

et al., 2011). In another example, the imprinted H19/IGF2 locus,

CTCF-based, cohesin-mediated long-range interactions were

shown to disrupt enhancer-promoter contacts (Nativio et al.,

2009). It is tempting to think that the impact of cohesin on

gene regulation depends on the nature of gene regulatory

elements it connects at a specific locus.

Although these examples show correlations between gene

expression, long-range interactions, and cohesin binding, it

should be noted that the detailed causal relationships remain

to be worked out. It also remains to be explored how the mech-

anism of cohesin-mediated long-range interactions in cis relates

to the topological embrace thought to provide sister chromatid

cohesion in trans (Nasmyth and Haering, 2009). Limitations of

current experimental approaches to understanding cohesin’s

role in gene expression are discussed in Box 4.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1291

Strengthening Transcription Factor Binding

at Low-Affinity Motifs

In addition to its role in supporting long-range interactions,

cohesin may facilitate the binding of transcription factors to

suboptimal sequence motifs (Faure et al., 2012). A recent study

exhaustively compared the genomic binding of a large set of

tissue-specific transcription factors, cohesin, and RNAPII with

full annotation of chromatin state in mouse liver with the goal of

understanding the transcriptional interplay among these ele-

ments of regulation. Cohesin is found to stabilize the binding of

transcription factors to lower-affinity sequencemotifs—ahypoth-

esis confirmed by testing whether specific transcription factor

modules (identified based on their motif quality) are destabilized

in a mouse haploinsufficient for the cohesin subunit RAD21.

In summary, mounting evidence argues for multiple roles of

cohesin in gene regulation. In a few examples (Pauli et al.,

2008, 2010; Schuldiner et al., 2008; Seitan et al., 2011), the

impact of cohesin on gene expression has been dissociated

from cohesin’s essential functions in DNA repair, chromosome

segregation, and emerging functions in DNA replication (Tittel-

Elmer et al., 2012).

CTCF and Cohesin Regulate Complex LociThe interdependence of chromatin modifications, regulatory

elements, transcription factor binding, and promoter-enhancer

interactions is illustrated by the imprinted H19/IGF2 locus,

which provides a well-documented example of a mammalian

insulator (Figure 1). The IGF2/H19 imprinting control region

(ICR) comprises a cluster of CTCF sites, and imprinted H19/

IGF2 expression is regulated by the selective ICR methylation

in sperm, but not in ova. Thus, CTCF selectively binds the

unmethylated maternal allele, where it blocks the expression of

IGF2 (Bell and Felsenfeld, 2000; Hark et al., 2000; Kanduri

et al., 2000). The insulator function of CTCF at the maternal

IGF2/H19 allele is reflected in reduced long-range interactions

of a distal enhancer with the maternal IGF2 promoter (Murrell

et al., 2004). In contrast, methylation of the paternal ICR

precludes CTCF binding and abrogates insulator function so

that paternal IGF2 is expressed (Bell and Felsenfeld, 2000;

Hark et al., 2000; Kanduri et al., 2000; Figure 1). Maternally

inherited ICR microdeletions that remove a subset of CTCF sites

can result in the methylation of remaining sites and the loss of

imprinting in Beckwith-Wiedemann syndrome (Choufani et al.,

2010). The impact of such deletions correlates with the spatial

arrangement rather than the number of the remaining CTCF sites

(Beygo et al., 2013).

In addition to H19/IGF2, CTCF and cohesin regulate many

other complex loci, including the b-globin locus (Splinter et al.,

2006), proto-cadherin loci (Hirayama et al., 2012; Remeseiro

et al., 2012), lymphocyte receptor loci (Seitan et al., 2012), and

the X chromosome inactivation region (Spencer et al., 2011). It

is possible that complex loci are particularly dependent on

CTCF and cohesin.

Regulation of Multigene Cluster Loci

Conditional deletion of CTCF from postmitotic projection

neurons results in the misexpression of several hundred tran-

scripts, including the clustered protocadherin genes. Mice

lacking CTCF in a subset of their neurons have defects in func-

1292 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

tional somatosensory mapping and suffer from postnatal growth

retardation and abnormal behavior (Hirayama et al., 2012). A

different mouse model demonstrates that the cohesin subunit

SA1 positively regulates neuronal protocadherin gene expres-

sion (Remeseiro et al., 2012).

Lymphocyte receptor loci contain hundreds of coding

elements arranged over large genomic regions. To make func-

tional lymphocyte receptors, these regions must be rearranged

by a somatic recombination process mediated by the trans-

poson-derived recombinases Rag1 and Rag2 (Schatz, 2004).

Rag2 links chromatin structure to the somatic rearrangement

of lymphocyte receptor gene loci due to its selective interaction

with H3K4me3 (Matthews et al., 2007). Recruitment of Rag2 by

transcription-associated histone modifications explains why

the initiation of recombination requires transcriptional activity.

Regulation of this activity in a cell-type- and developmental-

stage-specific manner provides a mechanism for rearranging

each lymphocyte receptor locus at the appropriate time and in

the appropriate cell type (Stanhope-Baker et al., 1996). Interest-

ingly, the coordination of cell-type- and developmental-stage-

specific lymphocyte receptor locus transcription requires both

CTCF and cohesin (Degner et al., 2011; Seitan et al., 2011;

Ribeiro de Almeida et al., 2011; reviewed by Seitan et al.,

2012; Bossen et al., 2012). Furthermore, transcription and rear-

rangement of lymphocyte receptor loci are perturbed by the

deletion of endogenous CTCF sites (Guo et al., 2011) or the intro-

duction of ectopic CTCF sites (Shrimali et al., 2012).

CTCF Control of Noncoding RNA Transcription

The impact of CTCF and cohesin on the transcription and rear-

rangement of lymphocyte receptor gene loci is mediated in

part by long-range interactions and in part by antisense tran-

scription (Degner et al., 2011; Featherstone et al., 2010). This

theme is reiterated at the locus encoding ataxin-7, which is

flanked by a CAG/polyglutamine repeat. When expanded, this

repeat results in the neurodegenerative disorder spinocerebellar

ataxia. The ataxin-7 repeat and translation start site are flanked

by binding sites for CTCF, and CTCF promotes the transcription

of a noncoding, convergently transcribed antisense RNA, which

determines ataxin-7 promoter usage (Sopher et al., 2011).

The ribosomal DNA (rDNA) locus contains hundreds of copies

of rDNA genes, only some of which are actively transcribed. In

addition to rDNA gene promoters, rDNA transcription is regu-

lated by spacer promoters that give rise to noncoding RNAs

and are regulated by CTCF (van de Nobelen et al., 2010).

In mammals, X chromosome inactivation equalizes X-linked

gene expression between XY male and XX female cells and is

controlled by a genomic region designated the X-inactivation

center. This region harbors two distinct chromatin segments,

each centered around noncoding genes transcribed in opposite

directions, Xss, ist and Tsix. A conserved CTCF binding element

positioned between these regions facilitates Xist induction and X

chromosome inactivation in female cells (Spencer et al., 2011).

Transcriptional Regulation Linked to CTCF Eviction

or Recruitment

Inducible noncoding RNA transcription has been reported to

evict CTCF from a site upstream of the chicken lysozyme

promoter (Lefevre et al., 2008). The RARb2 gene displays an

intriguing mechanism for regulated CTCF recruitment (Le May

et al., 2012). It starts with the introduction of DNA breaks by

the XPG endonuclease and is followed by DNA repair, which

replaces methylated with unmethylated DNA. This allows

CTCF to bind and to form chromatin loops that correlate with

locus transcription (Le May et al., 2012).

Regulation of RNA Polymerase Elongation

and Alternative Splicing

Fay et al. (2011) have shown that local cohesin binding can

impact the processivity of RNAPII. The rate of transcriptional

elongation is known to impact on alternative splicing (Ip et al.,

2011), and CTCF can promote the inclusion of weak exons by

mediating local RNAPII pausing at the alternatively spliced

CD45 locus as well as genome wide (Shukla et al., 2011). Both

CTCF binding and exon inclusion are sensitive to DNA methyla-

tion, linking the developmental regulation of splicing with epige-

netic marks.

The mechanisms described in this section are the result of

detailed locus-specific studies, and their general significance

remains to be tested on a genome-wide level.

PerspectiveCTCF binding is often associatedwith constitutive DNaseI hyper-

sensitive sites (Parelho et al., 2008). Within one species, some

CTCF sites can reflect cell-type-specific chromatin states

(Wang et al., 2012), but most CTCF sites are shared among dif-

ferent cell types (Kim et al., 2007). Most—but not all—CTCF sites

attract cohesin and, although the mechanisms of selective cohe-

sin recruitment by CTCF remain to be defined, it is clear that, in

isolation, CTCF-associated cohesin sites are relatively static

amongdiverse cell types and tissues.On the scale of evolutionary

time, the ancient and ongoing remodeling of the mammalian

genome by repeat elements that carry CTCF insures that even

these stable CTCF-cohesin anchorages diverge between spe-

cies. In contrast, cohesin binding at enhancers and promoters is

often cell-type specific and thus reflects the dynamic transcrip-

tional state of different cell types (Kagey et al., 2010; Schmidt

et al., 2010a; Kim et al., 2005; Rada-Iglesias et al., 2011; Heintz-

man et al., 2009; Visel et al., 2009; Shen et al., 2012). The interac-

tion of cohesin with both CTCF and active enhancers and

promoters can be thought of as a unifying mechanism that links

the rapidly evolvingbindingof tissue-specific transcription factors

with themoredevelopmentallyandevolutionarily stablebindingof

CTCF into networks of long-range interactions that reflect and

promote the transcriptional programs of specific cell types.

ACKNOWLEDGMENTS

We thank Vlad Seitan and other lab members for suggestions and Anthony

Lewis (MRC Clinical Sciences Centre) and Michelle Ward (University of

Cambridge) for help with graphics, and we apologize to our colleagues whose

work could not be cited due to space constraints. This work was supported by

the Medical Research Council, UK (M.M.), the Wellcome Trust (M.M.), Cancer

Research UK (D.T.O.), European Research Council (D.T.O.), and EMBOYoung

Investigators Programme (D.T.O.).

REFERENCES

Baniahmad, A., Steiner, C., Kohne, A.C., and Renkawitz, R. (1990). Modular

structure of a chicken lysozyme silencer: involvement of an unusual thyroid

hormone receptor binding site. Cell 61, 505–514.

Bannister, A.J., and Kouzarides, T. (2011). Regulation of chromatin by histone

modifications. Cell Res. 21, 381–395.

Bell, A.C., and Felsenfeld, G. (2000). Methylation of a CTCF-dependent

boundary controls imprinted expression of the Igf2 gene. Nature 405, 482–485.

Bell, A.C., West, A.G., and Felsenfeld, G. (1999). The protein CTCF is required

for the enhancer blocking activity of vertebrate insulators. Cell 98, 387–396.

Bell, O., Tiwari, V.K., Thoma, N.H., and Schubeler, D. (2011). Determinants and

dynamics of genome accessibility. Nat. Rev. Genet. 12, 554–564.

Bernstein, B.E., Kamal, M., Lindblad-Toh, K., Bekiranov, S., Bailey, D.K.,

Huebert, D.J., McMahon, S., Karlsson, E.K., Kulbokas, E.J., 3rd, Gingeras,

T.R., et al. (2005). Genomic maps and comparative analysis of histone modifi-

cations in human and mouse. Cell 120, 169–181.

Beygo, J., Citro, V., Sparago, A., De Crescenzo, A., Cerrato, F., Heitmann, M.,

Rademacher, K., Guala, A., Enklaar, T., Anichini, C., et al. (2013). The molec-

ular function and clinical phenotype of partial deletions of the IGF2/H19

imprinting control region depends on the spatial arrangement of the remaining

CTCF-binding sites. Hum. Mol. Genet. 22, 544–557.

Bird, A.P. (1980). DNA methylation and the frequency of CpG in animal DNA.

Nucleic Acids Res. 8, 1499–1504.

Borneman, A.R., Leigh-Bell, J.A., Yu, H., Bertone, P., Gerstein, M., and

Snyder, M. (2006). Target hub proteins serve as master regulators of develop-

ment in yeast. Genes Dev. 20, 435–448.

Bossen, C., Mansson, R., and Murre, C. (2012). Chromatin topology and the

regulation of antigen receptor assembly. Annu. Rev. Immunol. 30, 337–356.

Bourque, G., Leong, B., Vega, V.B., Chen, X., Lee, Y.L., Srinivasan, K.G.,

Chew, J.L., Ruan, Y., Wei, C.L., Ng, H.H., and Liu, E.T. (2008). Evolution of

the mammalian transcription factor binding repertoire via transposable

elements. Genome Res. 18, 1752–1762.

Boyle, S., Rodesch, M.J., Halvensleben, H.A., Jeddeloh, J.A., and Bickmore,

W.A. (2011). Fluorescence in situ hybridization with high-complexity repeat-

free oligonucleotide probes generated by massively parallel synthesis.

Chromosome Res. 19, 901–909.

Brockdorff, N. (2011). Chromosome silencing mechanisms in X-chromosome

inactivation: unknown unknowns. Development 138, 5057–5065.

Burcin, M., Arnold, R., Lutz, M., Kaiser, B., Runge, D., Lottspeich, F., Filippova,

G.N., Lobanenkov, V.V., and Renkawitz, R. (1997). Negative protein 1, which is

required for function of the chicken lysozyme gene silencer in conjunction with

hormone receptors, is identical to the multivalent zinc finger repressor CTCF.

Mol. Cell. Biol. 17, 1281–1288.

Campos, E.I., and Reinberg, D. (2009). Histones: annotating chromatin. Annu.

Rev. Genet. 43, 559–599.

Chen, H.S., Wikramasinghe, P., Showe, L., and Lieberman, P.M. (2012). Cohe-

sins repress Kaposi’s sarcoma-associated herpesvirus immediate early gene

transcription during latency. J. Virol. 86, 9454–9464.

Choufani, S., Shuman, C., and Weksberg, R. (2010). Beckwith-Wiedemann

syndrome. Am. J. Med. Genet. C. Semin. Med. Genet. 154C, 343–354.

Church, D.M., Goodstadt, L., Hillier, L.W., Zody, M.C., Goldstein, S., She, X.,

Bult, C.J., Agarwala, R., Cherry, J.L., DiCuccio, M., et al.; Mouse Genome

Sequencing Consortium. (2009). Lineage-specific biology revealed by

a finished genome assembly of the mouse. PLoS Biol. 7, e1000112.

Ciofani, M., Madar, A., Galan, C., Sellars, M., Mace, K., Pauli, F., Agarwal, A.,

Huang, W., Parkurst, C.N., Muratet, M., et al. (2012). A validated regulatory

network for Th17 cell specification. Cell 151, 289–303.

Cohen, N.M., Kenigsberg, E., and Tanay, A. (2011). Primate CpG islands are

maintained by heterogeneous evolutionary regimes involving minimal selec-

tion. Cell 145, 773–786.

Conaway, R.C., and Conaway, J.W. (2011). Function and regulation of the

Mediator complex. Curr. Opin. Genet. Dev. 21, 225–230.

Cuddapah, S., Jothi, R., Schones, D.E., Roh, T.Y., Cui, K., and Zhao, K. (2009).

Global analysis of the insulator binding protein CTCF in chromatin barrier

regions reveals demarcation of active and repressive domains. Genome

Res. 19, 24–32.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1293

Deardorff, M.A., Bando, M., Nakato, R., Watrin, E., Itoh, T., Minamino, M.,

Saitoh, K., Komata, M., Katou, Y., Clark, D., et al. (2012). HDAC8 mutations

in Cornelia de Lange syndrome affect the cohesin acetylation cycle. Nature

489, 313–317.

Degner, S.C., Verma-Gaur, J., Wong, T.P., Bossen, C., Iverson, G.M.,

Torkamani, A., Vettermann, C., Lin, Y.C., Ju, Z., Schulz, D., et al. (2011).

CCCTC-binding factor (CTCF) and cohesin influence the genomic architecture

of the Igh locus and antisense transcription in pro-B cells. Proc. Natl. Acad.

Sci. USA 108, 9566–9571.

Dekker, J. (2008). Gene regulation in the third dimension. Science 319, 1793–

1794.

Dixon, J.R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., Hu, M., Liu, J.S., and

Ren, B. (2012). Topological domains in mammalian genomes identified by

analysis of chromatin interactions. Nature 485, 376–380.

Dostie, J., and Bickmore, W.A. (2012). Chromosome organization in the

nucleus - charting new territory across the Hi-Cs. Curr. Opin. Genet. Dev.

22, 125–131.

Dostie, J., Richmond, T.A., Arnaout, R.A., Selzer, R.R., Lee, W.L., Honan, T.A.,

Rubio, E.D., Krumm, A., Lamb, J., Nusbaum, C., et al. (2006). Chromosome

Conformation Capture Carbon Copy (5C): a massively parallel solution for

mapping interactions between genomic elements. Genome Res. 16, 1299–

1309.

Dunham, I., Kundaje, A., Aldred, S.F., Collins, P.J., Davis, C.A., Doyle, F.,

Epstein, C.B., Frietze, S., Harrow, J., Kaul, R., et al.; ENCODE Project Consor-

tium. (2012). An integrated encyclopedia of DNA elements in the human

genome. Nature 489, 57–74.

Faure, A.J., Schmidt, D., Watt, S., Schwalie, P.C., Wilson, M.D., Xu, H.,

Ramsay, R.G., Odom, D.T., and Flicek, P. (2012). Cohesin regulates tissue-

specific expression by stabilising highly occupied cis-regulatory modules.

Genome Res. 22, 2163–2175.

Fay, A., Misulovin, Z., Li, J., Schaaf, C.A., Gause, M., Gilmour, D.S., and

Dorsett, D. (2011). Cohesin selectively binds and regulates genes with paused

RNA polymerase. Curr. Biol. 21, 1624–1634.

Featherstone, K., Wood, A.L., Bowen, A.J., and Corcoran, A.E. (2010). The

mouse immunoglobulin heavy chain V-D intergenic sequence contains

insulators that may regulate ordered V(D)J recombination. J. Biol. Chem.

285, 9327–9338.

Fedoriw, A.M., Stein, P., Svoboda, P., Schultz, R.M., and Bartolomei, M.S.

(2004). Transgenic RNAi reveals essential function for CTCF in H19 gene

imprinting. Science 303, 238–240.

Fedoroff, N.V. (2012). Presidential address. Transposable elements, epige-

netics, and genome evolution. Science 338, 758–767.

Filippova, G.N., Fagerlie, S., Klenova, E.M., Myers, C., Dehner, Y., Goodwin,

G., Neiman, P.E., Collins, S.J., and Lobanenkov, V.V. (1996). An exceptionally

conserved transcriptional repressor, CTCF, employs different combinations of

zinc fingers to bind diverged promoter sequences of avian and mammalian

c-myc oncogenes. Mol. Cell. Biol. 16, 2802–2813.

Galande, S., Purbey, P.K., Notani, D., and Kumar, P.P. (2007). The third dimen-

sion of gene regulation: organization of dynamic chromatin loopscape by

SATB1. Curr. Opin. Genet. Dev. 17, 408–414.

Giles, K.E., Gowher, H., Ghirlando, R., Jin, C., and Felsenfeld, G. (2010). Chro-

matin boundaries, insulators, and long-range interactions in the nucleus. Cold

Spring Harb. Symp. Quant. Biol. 75, 79–85.

Guo, C., Yoon, H.S., Franklin, A., Jain, S., Ebert, A., Cheng, H.L., Hansen, E.,

Despo, O., Bossen, C., Vettermann, C., et al. (2011). CTCF-binding elements

mediate control of V(D)J recombination. Nature 477, 424–430.

Guy, J., Gan, J., Selfridge, J., Cobb, S., and Bird, A. (2007). Reversal of neuro-

logical defects in a mouse model of Rett syndrome. Science 315, 1143–1147.

Hadjur, S., Williams, L.M., Ryan, N.K., Cobb, B.S., Sexton, T., Fraser, P.,

Fisher, A.G., and Merkenschlager, M. (2009). Cohesins form chromosomal

cis-interactions at the developmentally regulated IFNG locus. Nature 460,

410–413.

1294 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Handoko, L., Xu, H., Li, G., Ngan, C.Y., Chew, E., Schnapp, M., Lee, C.W., Ye,

C., Ping, J.L., Mulawadi, F., et al. (2011). CTCF-mediated functional chromatin

interactome in pluripotent cells. Nat. Genet. 43, 630–638.

Harbison, C.T., Gordon, D.B., Lee, T.I., Rinaldi, N.J., Macisaac, K.D., Danford,

T.W., Hannett, N.M., Tagne, J.B., Reynolds, D.B., Yoo, J., et al. (2004). Tran-

scriptional regulatory code of a eukaryotic genome. Nature 431, 99–104.

Hark, A.T., Schoenherr, C.J., Katz, D.J., Ingram, R.S., Levorse, J.M., and

Tilghman, S.M. (2000). CTCF mediates methylation-sensitive enhancer-

blocking activity at the H19/Igf2 locus. Nature 405, 486–489.

He, Q., Bardet, A.F., Patton, B., Purvis, J., Johnston, J., Paulson, A., Gogol, M.,

Stark, A., and Zeitlinger, J. (2011). High conservation of transcription factor

binding and evidence for combinatorial regulation across six Drosophila

species. Nat. Genet. 43, 414–420.

Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P., Stark, A., Harp,

L.F., Ye, Z., Lee, L.K., Stuart, R.K., Ching, C.W., et al. (2009). Histone modifi-

cations at human enhancers reflect global cell-type-specific gene expression.

Nature 459, 108–112.

Hesselberth, J.R., Chen, X., Zhang, Z., Sabo, P.J., Sandstrom, R., Reynolds,

A.P., Thurman, R.E., Neph, S., Kuehn, M.S., Noble, W.S., et al. (2009). Global

mapping of protein-DNA interactions in vivo by digital genomic footprinting.

Nat. Methods 6, 283–289.

Hirayama, T., Tarusawa, E., Yoshimura, Y., Galjart, N., and Yagi, T. (2012).

CTCF is required for neural development and stochastic expression of clus-

tered Pcdh genes in neurons. Cell Rep. 2, 345–357.

Holdorf, M.M., Cooper, S.B., Yamamoto, K.R., andMiranda, J.J. (2011). Occu-

pancy of chromatin organizers in the Epstein-Barr virus genome. Virology 415,

1–5.

Hughes, D.J., Marendy, E.M., Dickerson, C.A., Yetming, K.D., Sample, C.E.,

and Sample, J.T. (2012). Contributions of CTCF and DNA methyltransferases

DNMT1 and DNMT3B to Epstein-Barr virus restricted latency. J. Virol. 86,

1034–1045.

Ip, J., Schmidt, D., Pan, Q., Ramani, A., Fraser, A., Odom, D.T., and Blencowe,

B. (2011). Global impact of RNA polymerase II elongation inhibition on alterna-

tive splicing regulation. Genome Res. 21, 390–401.

Jang, M.K., Mochizuki, K., Zhou, M., Jeong, H.S., Brady, J.N., and Ozato, K.

(2005). The bromodomain protein Brd4 is a positive regulatory component of

P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol.

Cell 19, 523–534.

Johnson, D.S., Mortazavi, A., Myers, R.M., andWold, B. (2007). Genome-wide

mapping of in vivo protein-DNA interactions. Science 316, 1497–1502.

Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A.,

van Berkum, N.L., Ebmeier, C.C., Goossens, J., Rahl, P.B., Levine, S.S.,

et al. (2010). Mediator and cohesin connect gene expression and chromatin

architecture. Nature 467, 430–435.

Kanduri, C., Pant, V., Loukinov, D., Pugacheva, E., Qi, C.F., Wolffe, A.,

Ohlsson, R., and Lobanenkov, V.V. (2000). Functional association of CTCF

with the insulator upstream of the H19 gene is parent of origin-specific and

methylation-sensitive. Curr. Biol. 10, 853–856.

Kang, H., Wiedmer, A., Yuan, Y., Robertson, E., and Lieberman, P.M. (2011).

Coordination of KSHV latent and lytic gene control by CTCF-cohesin mediated

chromosome conformation. PLoS Pathog. 7, e1002140.

Kim, T.H., Barrera, L.O., Zheng, M., Qu, C., Singer, M.A., Richmond, T.A., Wu,

Y., Green, R.D., and Ren, B. (2005). A high-resolution map of active promoters

in the human genome. Nature 436, 876–880.

Kim, T.H., Abdullaev, Z.K., Smith, A.D., Ching, K.A., Loukinov, D.I., Green,

R.D., Zhang, M.Q., Lobanenkov, V.V., and Ren, B. (2007). Analysis of the verte-

brate insulator protein CTCF-binding sites in the human genome. Cell 128,

1231–1245.

Klenova, E.M., Nicolas, R.H., Paterson, H.F., Carne, A.F., Heath, C.M.,

Goodwin, G.H., Neiman, P.E., and Lobanenkov, V.V. (1993). CTCF, a

conserved nuclear factor required for optimal transcriptional activity of the

chicken c-myc gene, is an 11-Zn-finger protein differentially expressed in

multiple forms. Mol. Cell. Biol. 13, 7612–7624.

Kunarso, G., Chia, N.Y., Jeyakani, J., Hwang, C., Lu, X., Chan, Y.S., Ng, H.H.,

and Bourque, G. (2010). Transposable elements have rewired the core regula-

tory network of human embryonic stem cells. Nat. Genet. 42, 631–634.

Le May, N., Fradin, D., Iltis, I., Bougneres, P., and Egly, J.M. (2012). XPG and

XPF endonucleases trigger chromatin looping and DNA demethylation for

accurate expression of activated genes. Mol. Cell 47, 622–632.

Lee, T.I., Rinaldi, N.J., Robert, F., Odom, D.T., Bar-Joseph, Z., Gerber, G.K.,

Hannett, N.M., Harbison, C.T., Thompson, C.M., Simon, I., et al. (2002). Tran-

scriptional regulatory networks in Saccharomyces cerevisiae. Science 298,

799–804.

Lefevre, P., Witham, J., Lacroix, C.E., Cockerill, P.N., and Bonifer, C. (2008).

The LPS-induced transcriptional upregulation of the chicken lysozyme

locus involves CTCF eviction and noncoding RNA transcription. Mol. Cell 32,

129–139.

Li, G., Ruan, X., Auerbach, R.K., Sandhu, K.S., Zheng, M., Wang, P., Poh,

H.M., Goh, Y., Lim, J., Zhang, J., et al. (2012). Extensive promoter-centered

chromatin interactions provide a topological basis for transcription regulation.

Cell 148, 84–98.

Lieberman-Aiden, E., van Berkum, N.L., Williams, L., Imakaev, M., Ragoczy,

T., Telling, A., Amit, I., Lajoie, B.R., Sabo, P.J., Dorschner, M.O., et al.

(2009). Comprehensive mapping of long-range interactions reveals folding

principles of the human genome. Science 326, 289–293.

Lin, T., Chao, C., Saito, S., Mazur, S.J., Murphy, M.E., Appella, E., and Xu, Y.

(2005). p53 induces differentiation of mouse embryonic stem cells by sup-

pressing Nanog expression. Nat. Cell Biol. 7, 165–171.

Lindblad-Toh, K., Garber, M., Zuk, O., Lin, M.F., Parker, B.J., Washietl, S.,

Kheradpour, P., Ernst, J., Jordan, G., Mauceli, E., et al.; Broad Institute

Sequencing Platform and Whole Genome Assembly Team; Baylor College of

Medicine Human Genome Sequencing Center Sequencing Team; Genome

Institute at Washington University. (2011). A high-resolution map of human

evolutionary constraint using 29 mammals. Nature 478, 476–482.

Lobanenkov, V.V., Nicolas, R.H., Adler, V.V., Paterson, H., Klenova, E.M.,

Polotskaja, A.V., and Goodwin, G.H. (1990). A novel sequence-specific DNA

binding protein which interacts with three regularly spaced direct repeats of

the CCCTC-motif in the 50-flanking sequence of the chicken c-myc gene.

Oncogene 5, 1743–1753.

Lopez-Flores, I., and Garrido-Ramos, M.A. (2010). The repetitive DNA content

of eukaryotic genomes. Genome Dyn. 7, 1–28.

Loukinov, D.I., Pugacheva, E., Vatolin, S., Pack, S.D., Moon, H., Chernukhin, I.,

Mannan, P., Larsson, E., Kanduri, C., Vostrov, A.A., et al. (2002). BORIS,

a novel male germ-line-specific protein associated with epigenetic reprogram-

ming events, shares the same 11-zinc-finger domain with CTCF, the insulator

protein involved in reading imprinting marks in the soma. Proc. Natl. Acad. Sci.

USA 99, 6806–6811.

Lusk, R.W., and Eisen, M.B. (2010). Evolutionary mirages: selection on binding

site composition creates the illusion of conserved grammars in Drosophila

enhancers. PLoS Genet. 6, e1000829.

Lynch, M. (2007). The frailty of adaptive hypotheses for the origins of organ-

ismal complexity. Proc. Natl. Acad. Sci. USA 104(Suppl 1), 8597–8604.

MacArthur, S., Li, X.Y., Li, J., Brown, J.B., Chu, H.C., Zeng, L., Grondona, B.P.,

Hechmer, A., Simirenko, L., Keranen, S.V., et al. (2009). Developmental roles of

21 Drosophila transcription factors are determined by quantitative differences

in binding to an overlapping set of thousands of genomic regions. Genome

Biol. 10, R80.

Magdinier, F., Yusufzai, T.M., and Felsenfeld, G. (2004). Both CTCF-depen-

dent and -independent insulators are found between themouse T cell receptor

alpha and Dad1 genes. J. Biol. Chem. 279, 25381–25389.

Matthews, A.G., Kuo, A.J., Ramon-Maiques, S., Han, S., Champagne, K.S.,

Ivanov, D., Gallardo, M., Carney, D., Cheung, P., Ciccone, D.N., et al. (2007).

RAG2 PHD finger couples histone H3 lysine 4 trimethylation with V(D)J

recombination. Nature 450, 1106–1110.

Morey, L., and Helin, K. (2010). Polycomb group protein-mediated repression

of transcription. Trends Biochem. Sci. 35, 323–332.

Mortazavi, A., Leeper Thompson, E.C., Garcia, S.T., Myers, R.M., andWold, B.

(2006). Comparative genomics modeling of the NRSF/REST repressor

network: from single conserved sites to genome-wide repertoire. Genome

Res. 16, 1208–1221.

Murrell, A., Heeson, S., and Reik, W. (2004). Interaction between differentially

methylated regions partitions the imprinted genes Igf2 and H19 into parent-

specific chromatin loops. Nat. Genet. 36, 889–893.

Nasmyth, K., and Haering, C.H. (2009). Cohesin: its roles and mechanisms.

Annu. Rev. Genet. 43, 525–558.

Nativio, R., Wendt, K.S., Ito, Y., Huddleston, J.E., Uribe-Lewis, S., Woodfine,

K., Krueger, C., Reik, W., Peters, J.M., and Murrell, A. (2009). Cohesin is

required for higher-order chromatin conformation at the imprinted IGF2-H19

locus. PLoS Genet. 5, e1000739.

Neph, S., Vierstra, J., Stergachis, A.B., Reynolds, A.P., Haugen, E., Vernot, B.,

Thurman, R.E., John, S., Sandstrom, R., Johnson, A.K., et al. (2012). An expan-

sive human regulatory lexicon encoded in transcription factor footprints.

Nature 489, 83–90.

Nguyen, P., Cui, H., Bisht, K.S., Sun, L., Patel, K., Lee, R.S., Kugoh, H.,

Oshimura, M., Feinberg, A.P., and Gius, D. (2008). CTCFL/BORIS is a methyl-

ation-independent DNA-binding protein that preferentially binds to the

paternal H19 differentially methylated region. Cancer Res. 68, 5546–5551.

Noonan, J.P., and McCallion, A.S. (2010). Genomics of long-range regulatory

elements. Annu. Rev. Genomics Hum. Genet. 11, 1–23.

Noordermeer, D., Leleu, M., Splinter, E., Rougemont, J., De Laat, W., and

Duboule, D. (2011). The dynamic architecture of Hox gene clusters. Science

334, 222–225.

Nora, E.P., Lajoie, B.R., Schulz, E.G., Giorgetti, L., Okamoto, I., Servant, N.,

Piolot, T., van Berkum, N.L., Meisig, J., Sedat, J., et al. (2012). Spatial partition-

ing of the regulatory landscape of the X-inactivation centre. Nature 485,

381–385.

Odom, D.T., Zizlsperger, N., Gordon, D.B., Bell, G.W., Rinaldi, N.J., Murray,

H.L., Volkert, T.L., Schreiber, J., Rolfe, P.A., Gifford, D.K., et al. (2004). Control

of pancreas and liver gene expression by HNF transcription factors. Science

303, 1378–1381.

Odom, D.T., Dowell, R.D., Jacobsen, E.S., Nekludova, L., Rolfe, P.A., Danford,

T.W., Gifford, D.K., Fraenkel, E., Bell, G.I., and Young, R.A. (2006). Core tran-

scriptional regulatory circuitry in human hepatocytes. Mol. Syst. Biol. 2,

2006.0017.

Parelho, V., Hadjur, S., Spivakov, M., Leleu, M., Sauer, S., Gregson, H.C.,

Jarmuz, A., Canzonetta, C., Webster, Z., Nesterova, T., et al. (2008). Cohesins

functionally associate with CTCF on mammalian chromosome arms. Cell 132,

422–433.

Pauli, A., Althoff, F., Oliveira, R.A., Heidmann, S., Schuldiner, O., Lehner, C.F.,

Dickson, B.J., and Nasmyth, K. (2008). Cell-type-specific TEV protease

cleavage reveals cohesin functions in Drosophila neurons. Dev. Cell 14,

239–251.

Pauli, A., van Bemmel, J.G., Oliveira, R.A., Itoh, T., Shirahige, K., van Steensel,

B., and Nasmyth, K. (2010). A direct role for cohesin in gene regulation and

ecdysone response in Drosophila salivary glands. Curr. Biol. 20, 1787–1798.

Phillips, J.E., and Corces, V.G. (2009). CTCF: master weaver of the genome.

Cell 137, 1194–1211.

Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S.A., Flynn, R.A., and

Wysocka, J. (2011). A unique chromatin signature uncovers early develop-

mental enhancers in humans. Nature 470, 279–283.

Remeseiro, S., Cuadrado, A., Gomez-Lopez, G., Pisano, D.G., and Losada, A.

(2012). A unique role of cohesin-SA1 in gene regulation and development.

EMBO J. 31, 2090–2102.

Rhee, H.S., and Pugh, B.F. (2011). Comprehensive genome-wide protein-DNA

interactions detected at single-nucleotide resolution. Cell 147, 1408–1419.

Ribeiro de Almeida, C., Stadhouders, R., de Bruijn, M.J., Bergen, I.M., Thong-

juea, S., Lenhard, B., van Ijcken, W., Grosveld, F., Galjart, N., Soler, E., and

Hendriks, R.W. (2011). The DNA-binding protein CTCF limits proximal Vk

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1295

recombination and restricts k enhancer interactions to the immunoglobulin k

light chain locus. Immunity 35, 501–513.

Rollins, R.A., Morcillo, P., and Dorsett, D. (1999). Nipped-B, a Drosophila

homologue of chromosomal adherins, participates in activation by remote

enhancers in the cut and Ultrabithorax genes. Genetics 152, 577–593.

Rubio, E.D., Reiss, D.J., Welcsh, P.L., Disteche, C.M., Filippova, G.N., Baliga,

N.S., Aebersold, R., Ranish, J.A., and Krumm, A. (2008). CTCF physically links

cohesin to chromatin. Proc. Natl. Acad. Sci. USA 105, 8309–8314.

Samstein, R.M., Arvey, A., Josefowicz, S.Z., Peng, X., Reynolds, A., Sand-

strom, R., Neph, S., Sabo, P., Kim, J.M., Liao, W., Li, M.O., Leslie, C., Stama-

toyannopoulos, J.A., and Rudensky, A.Y. (2012). Foxp3 exploits a pre-existent

enhancer landscape for regulatory T cell lineage specification. Cell 151,

153–166.

Sanyal, A., Lajoie, B.R., Jain, G., and Dekker, J. (2012). The long-range inter-

action landscape of gene promoters. Nature 489, 109–113.

Schatz, D.G. (2004). Antigen receptor genes and the evolution of a recombi-

nase. Semin. Immunol. 16, 245–256.

Schmidt, D., Schwalie, P.C., Ross-Innes, C.S., Hurtado, A., Brown, G.D.,

Carroll, J.S., Flicek, P., and Odom, D.T. (2010a). A CTCF-independent role

for cohesin in tissue-specific transcription. Genome Res. 20, 578–588.

Schmidt, D., Wilson, M.D., Ballester, B., Schwalie, P.C., Brown, G.D.,

Marshall, A., Kutter, C., Watt, S., Martinez-Jimenez, C.P., Mackay, S., et al.

(2010b). Five-vertebrate ChIP-seq reveals the evolutionary dynamics of tran-

scription factor binding. Science 328, 1036–1040.

Schmidt, D., Schwalie, P.C., Wilson, M.D., Ballester, B., Goncalves, A., Kutter,

C., Brown, G.D., Marshall, A., Flicek, P., and Odom, D.T. (2012a). Waves of

retrotransposon expansion remodel genome organization and CTCF binding

in multiple mammalian lineages. Cell 148, 335–348.

Schmidt, E.F., Warner-Schmidt, J.L., Otopalik, B.G., Pickett, S.B., Greengard,

P., and Heintz, N. (2012b). Identification of the cortical neurons that mediate

antidepressant responses. Cell 149, 1152–1163.

Schoenborn, J.R., Dorschner, M.O., Sekimata,M., Santer, D.M., Shnyreva,M.,

Fitzpatrick, D.R., Stamatoyannopoulos, J.A., and Wilson, C.B. (2007).

Comprehensive epigenetic profiling identifies multiple distal regulatory

elements directing transcription of the gene encoding interferon-gamma.

Nat. Immunol. 8, 732–742.

Schuldiner, O., Berdnik, D., Levy, J.M., Wu, J.S., Luginbuhl, D., Gontang, A.C.,

and Luo, L. (2008). piggyBac-based mosaic screen identifies a postmitotic

function for cohesin in regulating developmental axon pruning. Dev. Cell 14,

227–238.

Segal, E., andWidom, J. (2009). FromDNA sequence to transcriptional behav-

iour: a quantitative approach. Nat. Rev. Genet. 10, 443–456.

Seitan, V.C., Hao, B., Tachibana-Konwalski, K., Lavagnolli, T., Mira-Bonten-

bal, H., Brown, K.E., Teng, G., Carroll, T., Terry, A., Horan, K., et al. (2011). A

role for cohesin in T cell receptor rearrangement and thymocyte differentiation.

Nature 476, 467–471.

Seitan, V.C., Krangel, M.S., and Merkenschlager, M. (2012). Cohesin, CTCF

and lymphocyte antigen receptor locus rearrangement. Trends Immunol. 33,

153–159.

Sekimata, M., Perez-Melgosa, M., Miller, S.A., Weinmann, A.S., Sabo, P.J.,

Sandstrom, R., Dorschner, M.O., Stamatoyannopoulos, J.A., and Wilson,

C.B. (2009). CCCTC-binding factor and the transcription factor T-bet orches-

trate T helper 1 cell-specific structure and function at the interferon-gamma

locus. Immunity 31, 551–564.

She, X., Cheng, Z., Zollner, S., Church, D.M., and Eichler, E.E. (2008). Mouse

segmental duplication and copy number variation. Nat. Genet. 40, 909–914.

Shen, Y., Yue, F., McCleary, D.F., Ye, Z., Edsall, L., Kuan, S., Wagner, U.,

Dixon, J., Lee, L., Lobanenkov, V.V., and Ren, B. (2012). Amap of the cis-regu-

latory sequences in the mouse genome. Nature 488, 116–120.

Shrimali, S., Srivastava, S., Varma, G., Grinberg, A., Pfeifer, K., and Srivastava,

M. (2012). An ectopic CTCF-dependent transcriptional insulator influences the

choice of Vb gene segments for VDJ recombination at TCRb locus. Nucleic

Acids Res. 40, 7753–7765.

1296 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Shukla, S., Kavak, E., Gregory, M., Imashimizu,M., Shutinoski, B., Kashlev,M.,

Oberdoerffer, P., Sandberg, R., and Oberdoerffer, S. (2011). CTCF-promoted

RNA polymerase II pausing links DNA methylation to splicing. Nature 479,

74–79.

Simonis, N., Rual, J.F., Carvunis, A.R., Tasan, M., Lemmens, I., Hirozane-Kish-

ikawa, T., Hao, T., Sahalie, J.M., Venkatesan, K., Gebreab, F., et al. (2009).

Empirically controlled mapping of the Caenorhabditis elegans protein-protein

interactome network. Nat. Methods 6, 47–54.

Siomi, H., and Siomi, M.C. (2011). Stress signaling etches heritable marks on

chromatin. Cell 145, 1005–1007.

Skibbens, R.V. (2009). Establishment of sister chromatid cohesion. Curr. Biol.

19, R1126–R1132.

Sopher, B.L., Ladd, P.D., Pineda, V.V., Libby, R.T., Sunkin, S.M., Hurley, J.B.,

Thienes, C.P., Gaasterland, T., Filippova, G.N., and La Spada, A.R. (2011).

CTCF regulates ataxin-7 expression through promotion of a convergently

transcribed, antisense noncoding RNA. Neuron 70, 1071–1084.

Soufi, A., Donahue, G., and Zaret, K.S. (2012). Facilitators and impediments of

the pluripotency reprogramming factors’ initial engagement with the genome.

Cell 151, 994–1004.

Soutourina, J., Wydau, S., Ambroise, Y., Boschiero, C., andWerner, M. (2011).

Direct interaction of RNA polymerase II and mediator required for transcription

in vivo. Science 331, 1451–1454.

Spencer, R.J., del Rosario, B.C., Pinter, S.F., Lessing, D., Sadreyev, R.I., and

Lee, J.T. (2011). A boundary element between Tsix and Xist binds the chro-

matin insulator Ctcf and contributes to initiation of X-chromosome inactivation.

Genetics 189, 441–454.

Splinter, E., Heath, H., Kooren, J., Palstra, R.J., Klous, P., Grosveld, F., Galjart,

N., and de Laat, W. (2006). CTCF mediates long-range chromatin looping and

local histonemodification in the beta-globin locus. Genes Dev. 20, 2349–2354.

Stadler, M.B., Murr, R., Burger, L., Ivanek, R., Lienert, F., Scholer, A.,

van Nimwegen, E., Wirbelauer, C., Oakeley, E.J., Gaidatzis, D., et al. (2011).

DNA-binding factors shape the mouse methylome at distal regulatory regions.

Nature 480, 490–495.

Stanhope-Baker, P., Hudson, K.M., Shaffer, A.L., Constantinescu, A., and

Schlissel, M.S. (1996). Cell type-specific chromatin structure determines the

targeting of V(D)J recombinase activity in vitro. Cell 85, 887–897.

Stedman, W., Kang, H., Lin, S., Kissil, J.L., Bartolomei, M.S., and Lieberman,

P.M. (2008). Cohesins localize with CTCF at the KSHV latency control region

and at cellular c-myc and H19/Igf2 insulators. EMBO J. 27, 654–666.

Stevens, H.C., Cham, K.S., Hughes, D.J., Sun, R., Sample, J.T., Bubb, V.J.,

Stewart, J.P., and Quinn, J.P. (2012). CTCF and Sp1 interact with the Murine

gammaherpesvirus 68 internal repeat elements. Virus Genes 45, 265–273.

Strachan, T. (2005). Cornelia de Lange Syndrome and the link between chro-

mosomal function, DNA repair and developmental gene regulation. Curr. Opin.

Genet. Dev. 15, 258–264.

Strom, L., Lindroos, H.B., Shirahige, K., and Sjogren, C. (2004). Postreplicative

recruitment of cohesin to double-strand breaks is required for DNA repair. Mol.

Cell 16, 1003–1015.

Suganuma, T., and Workman, J.L. (2011). Signals and combinatorial functions

of histone modifications. Annu. Rev. Biochem. 80, 473–499.

Tempera, I., Wiedmer, A., Dheekollu, J., and Lieberman, P.M. (2010). CTCF

prevents the epigenetic drift of EBV latency promoter Qp. PLoS Pathog. 6,

e1001048.

Thomas, M.C., and Chiang, C.M. (2006). The general transcription machinery

and general cofactors. Crit. Rev. Biochem. Mol. Biol. 41, 105–178.

Thomson, J.P., Skene, P.J., Selfridge, J., Clouaire, T., Guy, J., Webb, S., Kerr,

A.R., Deaton, A., Andrews, R., James, K.D., et al. (2010). CpG islands influence

chromatin structure via the CpG-binding protein Cfp1. Nature 464, 1082–1086.

Tittel-Elmer, M., Lengronne, A., Davidson, M.B., Bacal, J., Francois, P., Hohl,

M., Petrini, J.H., Pasero, P., and Cobb, J.A. (2012). Cohesin association to

replication sites depends on rad50 and promotes fork restart. Mol. Cell 48,

98–108.

Unal, E., Arbel-Eden, A., Sattler, U., Shroff, R., Lichten, M., Haber, J.E., and

Koshland, D. (2004). DNA damage response pathway uses histone modifica-

tion to assemble a double-strand break-specific cohesin domain. Mol. Cell

16, 991–1002.

Vahedi, G., Takahashi, H., Nakayamada, S., Sun, H., Sartorelli, V., Kanno, Y.,

and O’Shea, J.J. (2012). STATs shape the active enhancer landscape of T cell

populations. Cell 151, 981–993.

van de Nobelen, S., Rosa-Garrido, M., Leers, J., Heath, H., Soochit, W.,

Joosen, L., Jonkers, I., Demmers, J., van der Reijden, M., Torrano, V., et al.

(2010). CTCF regulates the local epigenetic state of ribosomal DNA repeats.

Epigenetics Chromatin 3, 19.

van Steensel, B. (2011). Chromatin: constructing the big picture. EMBO J. 30,

1885–1895.

Vega, H., Waisfisz, Q., Gordillo, M., Sakai, N., Yanagihara, I., Yamada, M., van

Gosliga, D., Kayserili, H., Xu, C., Ozono, K., et al. (2005). Roberts syndrome is

caused by mutations in ESCO2, a human homolog of yeast ECO1 that is

essential for the establishment of sister chromatid cohesion. Nat. Genet. 37,

468–470.

Vermeulen, M., Mulder, K.W., Denissov, S., Pijnappel, W.W., van Schaik, F.M.,

Varier, R.A., Baltissen, M.P., Stunnenberg, H.G., Mann, M., and Timmers, H.T.

(2007). Selective anchoring of TFIID to nucleosomes by trimethylation of

histone H3 lysine 4. Cell 131, 58–69.

Visel, A., Blow, M.J., Li, Z., Zhang, T., Akiyama, J.A., Holt, A., Plajzer-Frick, I.,

Shoukry, M., Wright, C., Chen, F., et al. (2009). ChIP-seq accurately predicts

tissue-specific activity of enhancers. Nature 457, 854–858.

Wallace, J.A., and Felsenfeld, G. (2007). We gather together: insulators and

genome organization. Curr. Opin. Genet. Dev. 17, 400–407.

Walsh, C.P., and Bestor, T.H. (1999). Cytosine methylation and mammalian

development. Genes Dev. 13, 26–34.

Wang, H., Maurano, M.T., Qu, H., Varley, K.E., Gertz, J., Pauli, F., Lee, K.,

Canfield, T., Weaver, M., Sandstrom, R., et al. (2012). Widespread

plasticity in CTCF occupancy linked to DNA methylation. Genome Res. 22,

1680–1688.

Wendt, K.S., Yoshida, K., Itoh, T., Bando, M., Koch, B., Schirghuber, E.,

Tsutsumi, S., Nagae, G., Ishihara, K., Mishiro, T., et al. (2008). Cohesin

mediates transcriptional insulation by CCCTC-binding factor. Nature 451,

796–801.

Yang, Z., Yik, J.H., Chen, R., He, N., Jang, M.K., Ozato, K., and Zhou, Q.

(2005). Recruitment of P-TEFb for stimulation of transcriptional elongation

by the bromodomain protein Brd4. Mol. Cell 19, 535–545.

Zhang, B., Chang, J., Fu, M., Huang, J., Kashyap, R., Salavaggione, E., Jain,

S., Kulkarni, S., Deardorff, M.A., Uzielli, M.L., et al. (2009). Dosage effects of

cohesin regulatory factor PDS5 on mammalian development: implications

for cohesinopathies. PLoS ONE 4, e5232.

Zhong, X.P., and Krangel, M.S. (1999). Enhancer-blocking activity within the

DNase I hypersensitive site 2 to 6 region between the TCR alpha and Dad1

genes. J. Immunol. 163, 295–300.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1297

Leading Edge

Review

Long Noncoding RNAs: Cellular AddressCodes in Development and Disease

Pedro J. Batista1 and Howard Y. Chang1,*1Howard Hughes Medical Institute and Program in Epithelial Biology, Stanford University School of Medicine, Stanford, CA 94305, USA

*Correspondence: howchang@stanford.eduhttp://dx.doi.org/10.1016/j.cell.2013.02.012

In biology as in real estate, location is a cardinal organizational principle that dictates the accessi-bility and flow of informational traffic. An essential question in nuclear organization is the nature ofthe address code—how objects are placed and later searched for and retrieved. Long noncodingRNAs (lncRNAs) have emerged as key components of the address code, allowing proteincomplexes, genes, and chromosomes to be trafficked to appropriate locations and subject toproper activation and deactivation. lncRNA-based mechanisms control cell fates during develop-ment, and their dysregulation underlies some human disorders caused by chromosomal deletionsand translocations.

IntroductionFrom a single cell to an entire organism, spatial positioning is

a key problem in biology. It is well appreciated that robust

systems sort and distribute macromolecules, a property essen-

tial for the function of cells and tissues (Shevtsov and Dundr,

2011; Wolpert, 2011). A historical example illustrates the general

utility of spatial organization. As the Roman Empire expanded

and the Romans were faced with the need to construct cities

in new lands, they developed a city prototype that included

a group of answers to the many practical problems related to

the creation and maintenance of a city (Figure 1A). This was

a universal plan of simple execution. City walls protected the citi-

zens from attack and delimited the city. At the center stood the

forum, where the business and political activities of the city

were concentrated. Fountains were placed throughout the city

to supply water, and other spaces, such as amphitheaters,

temples, and baths, were dedicated to organize daily activities.

Thus, a group of structures analogous in function was always

present in an organization that follows the original prototype

(Grimal and Woloch, 1983).

Just like the Roman city, the nucleus of the eukaryotic cell is

a highly organized space (Figure 1B). Evolution gave rise to

a ‘‘nuclear’’ prototype that provides answers to the many chal-

lenges the cell has to respond to maintain homeostasis and

growth, though subject to developmental specialization (Solovei

et al., 2009). Chromosomes are not randomly organized in the

nucleus, and during interphase, each chromosome occupies

a discrete territory (reviewed in Cremer and Cremer, 2010).

Furthermore, whereas the densely compacted heterochromatin

is localized at the nuclear envelope, euchromatin localizes to the

interior regions of the nucleus. Gene expression is also localized

and occurs mostly at nuclear center. In addition, active genes

that are coregulated are often found forming clusters. During

development, individual loci such as immunoglobulin or Hox

genes are known to change position within the nucleus accord-

ing to their transcriptional status (reviewed in Misteli, 2007).

1298 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Large portions of the genome are partitioned into topological

domains of chromatin interaction ranging from hundreds of

kilobases to megabases (the resolution of current methods),

within which the genes tend to be more coregulated (Dixon

et al., 2012; Nora et al., 2012). The complex task of gene expres-

sion—ensuring the proper timing, space, and rate of expres-

sion—involves noncoding regions of the genome, chromatin

modifications, and the arrangement of chromosomes and

nuclear domains. Here, we review the evidence that lncRNAs

are a rich source of molecular addresses in the eukaryotic

nucleus.

Biogenesis and CharacteristicsEfforts over the last decade revealed that a large fraction of the

noncoding genome is transcribed. Extensive annotation of

lncRNA has been performed in multiple model organisms

(reviewed in Rinn and Chang, 2012), and there is now evidence

that, whereas 2% of the genome encodes for proteins (IHGSC,

2004), primary transcripts cover 75% of the human genome,

with processed transcripts covering 62.1% of the genome

(Djebali et al., 2012). In this Review, we focus on a particular

class of noncoding transcripts known as long noncoding RNAs

(lncRNAs) and the roles that they play in nuclear organization.

lncRNAs are currently defined as transcripts of greater than

200 nucleotides without evident protein coding function (Rinn

and Chang, 2012). It is important to note that lncRNA is a broad

definition that encompasses different classes of RNA tran-

scripts, including enhancer RNAs, small nucleolar RNA (snoRNA)

hosts, intergenic transcripts, and transcripts overlapping other

transcripts in either sense or antisense orientation. lncRNAs

predominantly localize to the nucleus and have, on average,

a lower level of expression than protein coding genes, although

details vary for different classes (Djebali et al., 2012; Ravasi et al.,

2006). Multiple studies have shown that lncRNA expression is

more cell type specific than protein-coding genes (Cabili et al.,

2011; Djebali et al., 2012; Ravasi et al., 2006). At the DNA and

Figure 1. Comparison between a Roman City and the Cell Nucleus

Reveals the Importance of Spatial Organization(A) Depiction of the basic features of a Roman city. City walls delimit the city,with gates at the two main roads that intersect at the center of the city. TheForum was the business and political center of the city, and many buildingsprovided specific functions that were essential for city life.(B) Schematic representation of the typical nuclear organization during inter-phase. Each chromosome occupies a discrete territory. Euchromatin localizesto the interior regions of the nucleus, and the densely compacted hetero-chromatin localizes near the nuclear envelope. Many specialized functions areexecuted in distinct regions in the nucleus, known as nuclear bodies. Oneexample is the nucleolus, where ribosomes are assembled. Adapted fromSolovei et al., 2009.

chromatin level, lncRNA loci are similar to mRNA loci, but

lncRNAs show a bias for having just one intron and a trend for

less-efficient cotranscriptional splicing (Derrien et al., 2012;

Tilgner et al., 2012). Although lncRNAs are under lower selective

pressure than protein-coding genes, sequence analysis shows

that lncRNAs are under higher selective pressures than ancestral

repeat sequences, which are considered to be under neutral

selection. Interestingly, the promoters of lncRNAs are the region

of the lncRNA gene under higher selective pressure, displaying

levels of selection comparable to the promoters of protein-

coding genes (Derrien et al., 2012; Guttman et al., 2009;Marques

and Ponting, 2009; Ørom et al., 2010; Ponjavic et al., 2007). This

analysis has also revealed a high number of correlated positions

between lncRNA in sequence alignments, an observation that

fits the hypothesis that lncRNAs are under selective pressure

to maintain a functional RNA structure (Derrien et al., 2012).

Comparison between mammalian and zebrafish lncRNAs

revealed that short stretches of conserved sequence are func-

tionally important and that location and structure of lncRNAs

can be conserved, even in the absence of strong sequence

conservation. The ability to induce a loss-of-function phenotype

by blocking the short conserved motif in addition to the ability to

rescue loss of function of two lncRNAs with the addition of

human and mouse lncRNAs (Ulitsky et al., 2011) demonstrates

that these ‘‘in silico’’ observations are of biological significance.

Sequence analysis of lncRNAs, focusing on presence and

size of open reading frames as well as codon conservation

frequency, has been used to exclude protein coding potential.

Ribosome profiling, a method that enumerates transcripts asso-

ciated with ribosomes, had detected many lncRNAs, but it was

unclear whether these lncRNAs are just being scanned similarly

to 50 untranslated regions or actually are productively engaged in

translation (Ingolia et al., 2011). Comparison of RNA sequencing

(RNA-seq) data to tandem mass spectrometry data for two cell

lines suggests that�92% of the annotated lncRNAs do not yield

detectable peptides in these cell lines (Banfai et al., 2012; Derrien

et al., 2012). Although the differences between these two studies

may stem from measuring two different endpoints, they suggest

that lncRNAs have low translational potential even when ribo-

somes attempt to decode them. Current annotations suggest

that the actual number of lncRNAs exceeds that of protein

coding genes (Derrien et al., 2012).

The repertoire of roles performed by lncRNAs is growing, as

there is now evidence that lncRNAs participate in multiple

networks regulating gene expression and function. Several char-

acteristics of lncRNAsmake them the ideal system to provide the

nucleus with a system of molecular addresses. lncRNAs, unlike

proteins, can function both in cis, at the site of transcription, or

in trans. An RNA-based address code may be deployed more

rapidly and economically than a system that relies only on

proteins. lncRNAs do not need to be translated and do not

require transport between the cytoplasm and the nucleus.

lncRNAs can also interact with multiple proteins, enabling scaf-

folding functions and combinatorial control (Wang and Chang,

2011). As such, the act of transcription can rapidly create an

anchor that will lead to the formation, or remodeling, of nuclear

domains through the recruitment or sequestration of proteins

already present in the nuclear compartment. Using lncRNAs

allows cells to create addresses that are regional-, locus- or

even allele-specific (Lee, 2009). At the regional level, lncRNAs

can influence the formation of nuclear domains and the tran-

scriptional status of an entire chromosome, and they can partic-

ipate in the interaction of two different chromosomal regions. At

a more fine-grained level, lncRNAs can control the chromatin

state and activity of a chromosomal locus or specific gene. We

explore each of these concepts below with recently published

examples.

Locus Control of Gene RegulationCells can use noncoding RNAs to modulate gene expression by

changing the accessibility of gene promoters. These mecha-

nisms can be used to fine-tune gene expression in response to

environmental conditions or to silence a gene as part of a devel-

opmental program.

First, the act of noncoding RNA (ncRNA) transcription itself

can be purposed for regulatory function. For example, transcrip-

tion through a regulatory sequence, such as a promoter, can

block its function, a mechanism termed transcriptional interfer-

ence (Figure 2A) first identified in yeast (Martens et al., 2004).

In such instances, the lncRNA promoter is finely tuned to receive

appropriate inputs to exert regulatory function; the lncRNA

product is typically a faithful biomarker of transcriptional interfer-

ence in action but is not required for its success. In conditions

that limit vegetative growth, diploid S. cerevisiae cells enter spor-

ulation, a differentiation program that results in the formation of

haploid daughter cells. Entry into meiosis has catastrophic

consequences in haploid cells and is therefore inhibited via

a transcriptional interference mechanism. A transcription factor

in haploid cells activates the expression of IRT1(SUT643),

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1299

Figure 2. Functional Modules of lncRNAs in the Nucleus(A) The act of transcription at noncoding regions can modulate gene expres-sion through the recruitment of chromatin modifiers to the site of transcription.These complexes can create a local chromatin environment that facilitates orblocks the binding of other regulators.(B) lncRNAs can function in cis, recruiting protein complexes to their site oftranscription and thus creating a locus-specific address. Cells can use thismechanism to repress or activate gene expression.

1300 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

a noncoding RNA that overlaps the promoter of IME1, themaster

regulator of sporulation. Transcription of IRT1 establishes

a repressive chromatin state at the IME1 promoter through the

recruitment of histone methyltransferase Set2 and the histone

deacetylase Set3 (van Werven et al., 2012). The use of noncod-

ing transcription to control chromatin modification is a wide-

spread strategy. The Set3 histone deacetylase has also been

implicated in the modulation of gene induction kinetics during

changes of carbon source. Transcription of ncRNAs that overlap

the regulated genes leads to the establishment of H3K4me2,

which recruits Set3 and leads to the deacetylation of the gene

promoter. Deacetylation of the promoter results in delayed or

reduced induction of the regulated genes. This mechanism is

also involved in the inhibition of cryptic promoters (Kim et al.,

2012). Expression of GAL10-ncRNA, driven by Reb1, leads to

deacetylation across the GAL1-10 promoter, facilitating glucose

repression of GAL1-10 (Houseley et al., 2008).

In mammalian imprinting, the noncoding RNA Air (also known

as Airn) is expressed from the paternal chromosome and is

involved in silencing the paternal alleles of multiple genes. The

promoter of one of these genes, Igf2r, overlaps with the Air tran-

scriptional unit and is silenced by transcriptional interference

(Latos et al., 2012).

Transcriptional interference can also be used to activate gene

expression by inhibiting the action of repressor elements, func-

tioning as an antisilencing mechanism. In Drosophila embryo-

genesis, transcription through Polycomb response elements

(PRE) alters the function of these elements, blocking the estab-

lishment of repressive chromatin (Schmitt et al., 2005).

Second, lncRNAs can silence or activate gene expression in

cis, acting on neighboring genes of the lncRNA locus. Some of

the first studied examples of lncRNA function involve dosage

compensation and genomic imprinting, whereby lncRNAs

provide allele-specific gene regulation to differentially control

two copies of the same gene within one cell (see the Review

by Lee and Bartolomei on page 1308 of this issue; Lee and Bar-

tolomei, 2013) (Figure 2B). Several such lncRNAs are now recog-

nized to interact with and recruit histone modification

complexes, including Xist (recruits PRC2 for H3K27me3 and

RYBP-PRC1 for H2A ubiquitylation) and Kcnq1ot1 (recruits

G9a for H3K9me3 and PRC2) (Pandey et al., 2008; Tavares

et al., 2012; Zhao et al., 2010). The Air lncRNA (the transcription

of which inhibits Igfr2) targets G9a andH3K9me3 to silencemore

distantly located genes on the paternal chromosome (Nagano

et al., 2008); hence, one lncRNA gene can employ multiple

mechanisms to regulate nearby and distantly located genes. In

genome-wide studies, numerous lncRNAs have now been found

to interact with chromatin modification complexes (Guil et al.,

2012; Guttman et al., 2011; Khalil et al., 2009; Zhao et al.,

2010). In the plant A. thaliana, two cold-inducible lncRNAs,

COOLAIR and COLDAIR, are embedded antisense or intronic

to the flowering control locus gene FLC, and they help to recruit

PRC2 to stably silence FLC in a cold-dependent manner, a key

(C) lncRNAs can function in trans and recruit protein complexes to chromatinloci away from their site of transcription.(D) lncRNAs can bind and sequester transcription factors away from theirtarget chromosomal regions.

mechanism to ensure the proper flowering time after winter

termed ‘‘vernalization’’ (reviewed in Ietswaart et al., 2012). In

an analogous fashion, DNA damage induces a lncRNA from

the promoter of cyclin D1 gene (CCND1); this lncRNA binds to

TLS protein to allosterically inhibit histone acetyltransferase in

cis, which suppresses CCND1 transcription (Wang et al., 2008).

DNA methylation can occur as a long-term silencing mecha-

nism downstream of repressive histone modifications, and

lncRNAs may also guide DNA methylation in addition to histone

modification. The ribosomal DNA (rDNA) loci are tandemly

repeated in the genome, with some copies being transcription-

ally active, whereas others are silenced by DNA methylation

and histone modifications. Each ribosomal DNA transcribes

rRNA separated by intergenic spacers (IGSs) as a polycistronic

unit, and IGSs can be processed to 150–250 nt fragments

termed ‘‘promoter RNAs (pRNAs)’’ (reviewed in Bierhoff et al.,

2010). pRNA serves as a platform to recruit the de novo cytosine

methylase DNMT3 and the NoRC complex containing poly-ADP

ribose polymerase-1 (PARP-1) to promote silencing of rDNA

(Guetg et al., 2012; Mayer et al., 2006). Notably, a stretch of 20

nt in pRNA binds the rDNA promoter, forming a RNA:DNA:DNA

triplex (Schmitz et al., 2010). This triplex structure is proposed

to recruit DNMT3 and also serves as the specific recognition

mechanism between lncRNA and genomic DNA—a model that

likely applies to other lncRNA-DNA interactions (Martianov

et al., 2007).

A distinct family of lncRNAs serves to activate gene expres-

sion. Many active enhancer elements transcribe lncRNAs,

termed ‘‘eRNAs’’ (De Santa et al., 2010; Kim et al., 2010), and

several lncRNAs are required to activate gene expression, which

are termed ‘‘enhancer-like RNAs’’ (Ørom et al., 2010). Evf is a cis-

acting lncRNA that is required for the activation of Dlx5/6 genes

and generation of GABAergic interneurons in vivo (Bond et al.,

2009). A key mechanism of lncRNA specificity in cis is the

higher-order chromosomal configuration (Wang et al., 2011).

The noncoding RNA HOTTIP is expressed from the 50 end tip

of the HoxA locus and drives histone H3 lysine 4 trimethylation

and gene transcription of HoxA distal genes through the recruit-

ment of the WDR5/MLL complex (Wang et al., 2011). Endoge-

nous HOTTIP is brought to its target genes by chromosomal

looping, and ectopic HOTTIP only activates transcription when

it is artificially tethered to the reporter gene (Wang et al., 2011).

The MLL complex is also recruited to the Hox locus by the non-

coding RNA Mistral, located between Hoxa6 and Hoxa7. Mistral

directly interacts with MLL1, leading to changes at the chromatin

level that activateHoxa6 andHoxa7 (Bertani et al., 2011). Hence,

lncRNA interaction with MLL/Trx complexes and likely additional

proteins will define their function in enforcing active chromatin

states and gene activation.

Third, lncRNAs can control chromatin states at distantly

located genes (i.e., in trans) for both gene silencing and activa-

tion (Figure 2C). These lncRNAs bind to some of the same

effector chromatin modification complexes but target them to

genomic loci genome-wide. For instance, human HOTAIR

lncRNA binds to PRC2 and LSD1 complexes and couples

H3K27methylation andH3K4 demethylation activity to hundreds

of sites genome-wide (Chu et al., 2011; Tsai et al., 2010). HOTAIR

is located in the HOXC locus and is regulated in an anatomic

position-specific fashion. Linc-p21 is induced by p53 during

DNA damage and recruits hnRNPK via physical interaction to

mediate p53-mediated gene repression (Huarte et al., 2010).

Linc-p21 also has a recently recognized role in translational

control (Yoon et al., 2012). In contrast, PANDA, another lncRNA

induced by p53, acts as a decoy by binding to the transcription

factor NF-YA and preventing NF-YA from activating genes en-

coding cell death proteins (Hung et al., 2011) (Figure 2D).

lncRNA-mediated activation can also occur in trans. Jpx, an X-

linked lncRNA that activates Xist expression, is important for X

chromosome inactivation in female cells, and Jpx deletion can

be rescued by Jpx supplied in trans (Tian et al., 2010).

Nuclear DomainsThe concept of lncRNA recruitment of factors to genes may be

more properly considered a two-way street, with genes being

moved into specific cytotopic locations by lncRNAs. One type

of molecular address can be found in the formation of nuclear

domains. These are regions of the nucleus where specific func-

tions are performed. Unlike cellular organelles, these domains

are not membrane delimited. They are instead characterized

by the components that form them. These domains are believed

to form through molecular interactions between its components.

Once a stable interaction is found, the components remain asso-

ciated. These domains are often formed around the sites of tran-

scription of RNA components, which function as molecular

anchors (reviewed in Dundr and Misteli, 2010). The noncoding

RNA NEAT1, an essential component of the Paraspeckle, is

a well-characterized example of how noncoding RNAs can func-

tion as structural components of nuclear bodies. Upon transcrip-

tion of NEAT1, diffusible components of this domain nucleate at

the site of NEAT1 accumulation, leading to the formation of the

Paraspeckle (Figure 3A) (Chen and Carmichael, 2009; Clemson

et al., 2009; Mao et al., 2011; Sasaki et al., 2009; Shevtsov and

Dundr, 2011; Sunwoo et al., 2009).

Nuclear domains can be dynamically regulated in an RNA-

dependent fashion. In response to serum stimulation, the deme-

thylase KDM4C is recruited to the promoters of genes controlled

by the cell-cycle-specific transcription factor E2F, where it de-

methylates Polycomb protein Pc2. Whereas methylated Pc2

interacts with the noncoding RNA TUG1, a component of Poly-

comb bodies, unmethylated Pc2 interacts with the noncoding

RNA MALAT1/NEAT2, a component of interchromatin granules.

Therefore, changes in the methylation status of Pc2 lead to the

relocation of growth control genes from an environment that

inhibits gene expression, the Polycomb body, to a domain that

is permissive of gene expression, the interchromatin granule

(Figure 3B). Interestingly, the reading ability of Pc2 is modulated

by the noncoding RNA that it is interacting with. When bound to

TUG1, Pc2 reads H4R3me2s and H3K27me2, whereas it reads

H2AK5ac and H2AK13ac when interacting with MALAT1/

NEAT2 (Yang et al., 2011). These interplays control the growth-

factor-dependent expression of cell-cycle genes in vitro, but it

came as a surprise that mouse knockouts of either NEAT1 or

MALAT1/NEAT2 had no little overt phenotype (Eissmann et al.,

2012; Nakagawa et al., 2012; Nakagawa et al., 2011; Zhang

et al., 2012). Clearly, the question of redundancy or compensa-

tion in vivo needs to be addressed in the future.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1301

Figure 3. Schematic Representation of the Cell Nucleus, Showing

the Nucleolus and Chromosomal Territories(A) Protein components of the Paraspeckle diffused throughout the nucleo-plasm aggregate upon the transcription of NEAT1, forming the Paraspecklenuclear domain.(B) Pc2 differentially binds MALAT1/NEAT2 or TUG1 depending on methyla-tion status. Methylated Pc2 interacts with TUG1, bringing associated growthcontrol genes to a repressive environment, the polycomb body (PcG).Unmethylated Pc2 interacts with MALAT1/NEAT2 at the interchromatingranule (ICG), where gene expression is permitted.(C) Expression of lncRNAs with snoRNA ends from the Prader-Willi syndromelocus functions as a sink for the FOX2 protein, leading to redistribution of thissplicing factor in this nuclear region.(D) In response to cellular stress, transcription of specific IGSRNAs leads to theretention of targeted proteins at the nucleolus. Different types of stress lead tothe retention of different proteins through the expression of specific noncodingRNAs.

Unusual processing mechanisms may explain the localization

activity of certain lncRNAs. An imprinted region in chromosome

15 (15q11-q13) that had been implicated in Prader-Willi

syndrome (PWS) hosts multiple intron-derived lncRNAs with

small nucleolar RNAs at their ends—so called ‘‘sno-lncRNAs.’’

It is probable that the presence of structured snoRNAs at the

ends of lncRNAs stabilizes these molecules, which have no 50

cap or polyA tail. These RNAs are retained in the nucleus and

localize to, or remain near, their sites of transcription. Knock-

down of sno-lncRNAs has little effect on the expression of

nearby genes, suggesting that it does not affect gene expression

in cis. Instead, these sno-lncRNAs seem to create a ‘‘domain’’

where the splicing factor Fox2 is enriched. These sno-lncRNAs

contain multiple binding sites for Fox2, and altering the level of

sno-lncRNAs led to a redistribution of Fox2 in the nucleus and

changes in mRNA splicing patterns. Hence, the sno-lncRNAs

appear to function as Fox2 sinks, participating in the regulation

of splicing in specific subnuclear domains (Yin et al., 2012)

(Figure 3C). Similarly, formation of a blunt-ended triplex RNA

1302 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

structure at the 30 end of MALAT1/NEAT2 lncRNA, which lacks

a polyA tail, stabilizes the lncRNA and presumably limits its

export to the cytoplasm (Brown et al., 2012; Wilusz et al.,

2012). Viral nuclear lncRNAs have also adapted this strategy

and hide their 30 polyA tails in a triplex RNA structure to prevent

decay (Mitton-Fry et al., 2010; Tycowski et al., 2012).

Gene Control through SequestrationIn contrast to themodel of nuclear domains that concentrate and

thereby facilitate molecular interactions, spatial control can also

separate reactants until the moment is right. For example,

certain environmental stresses trigger the retention of select

proteins in the nucleolus away from their normal site of action.

The retention at the nucleolus requires a signal sequence

and the expression of specific noncoding RNAs expressed

from the large intergenic spacer (IGS) of the rDNA repeats. IGS

ncRNAs turn out to gate the responses to cellular stress. Unique

IGS ncRNAs are transcriptionally induced by specific stressors,

functioning as baits for proteins with specific signal sequences.

Interfering with a specific IGSRNA does not affect the function of

other IGSRNAs (Audas et al., 2012) (Figure 3D).

In S. pombe, both mRNAs and lncRNAs function together to

form heterochromatin and sequester genes in the control of

meiosis. During vegetative growth, the expression of meiotic

genes is repressed through selective elimination of meiotic

mRNAs. Meiotic genes contain within their transcripts a region

known as determinant of selective removal (DSR) that deter-

mines their degradation. This sequence is recognized by

Mmi1, which promotes both mRNA degradation (Harigaya

et al., 2006) as well as formation of facultative heterochromatic

islands (Zofall et al., 2012). Hence, aberrant nascent mRNAs

can function in an lncRNA-like fashion to tether the formation

for heterochromatin. Furthermore, during vegetative growth,

Mei2p, an RNA-binding protein that is crucial for entry inmeiosis,

is kept in an inactive form. When cells commit to the meiosis

expression program, Mei2p accumulates in its active form and

sequesters Mmi1 to a structure known as Mei2 dot, where

Mmi1 function is inhibited. The Mei2 dot forms at the sme2 locus

at the site of transcription of two noncoding RNAs, meiRNA-S

and meiRNA-L, which are necessary for the formation of the

Mei2 dot structure and, therefore, entry in meiosis (Yamamoto,

2010).

Higher-Order Chromosomal InteractionsAn intriguing possibility is that lncRNAs can regulate the three-

dimensional structure of the chromosomes by facilitating the

interaction of specific chromosomal loci. The act of transcription

itself can influence gene expression and genome organization by

promoting chromatin modifications, by recruiting gene active

regions to common transcription factories, or by exposing the

DNA strands to enzymatic activity. Hence, the presence of

multiple lncRNA genes in a region may help chromosomal loci

adopt distinct conformation with transcriptional activation. For

example, in the Hox loci, collinear expression of Hox mRNA

genes and Hox lncRNAs along the chromosome is associated

with the progressive recruitment of those chromosomal

segments into a tightly interacting domain that is distinct from

the transcriptionally silent portion of the loci (Noordermeer

Figure 4. lncRNAs Regulate Gene Expression in the Cytoplasm(A) The lncRNA TINCR interacts with STAU1 and target mRNAs containing theTINCR box motif, promoting their stability.(B) lncRNAs of the 1/2-sbsRNAs class hybridize with 30-UTR-containing Aluelements and promote the degradation of these target mRNAs.(C) Under stress conditions, the lncRNA antisense to Uchl1 moves from thenucleus to the cytoplasm and binds the 50 end of the Uchl1mRNA to promoteits translation under stress conditions.(D) lincRNA-p21 interacts with and targets RcK to mRNAs, resulting in trans-lation inhibition.

et al., 2011; Wang et al., 2011). A similar phenomenon was first

appreciated in the b-globin locus, and intergenic transcripts

from its locus control regions (Ashe et al., 1997). Transcription-

coupled looping is likely to be related to the fact that theMediator

complex that links transcription factors to basal transcription

machinery promotes long-range enhancer-promoter interac-

tions (Kagey et al., 2010). A similar transcription-directed mech-

anism has also been proposed to guide DNA recombination of

lymphocyte receptor genes over megabases (Verma-Gaur

et al., 2012). The lncRNA transcripts are useful readouts of the

chromosomal configuration but are not necessarily required for

the chromosomal interactions.

lncRNAs can also regulate chromosome structure through

direct mechanisms. High-throughput chromosomal conforma-

tion assays revealed that the active and inactive X chromosomes

adopt quite distinct conformations. The inactive X (Xi) is coated

by the Xist lncRNA, which is required for choosing the inactive

X chromosome. Importantly, conditional knockout of Xist has

demonstrated that the folding of inactive X requires the Xist

RNA. After Xist deletion, the Xi chromosome adopts a conforma-

tion that is more similar to that of the active X chromosome (Xa)

without reactivation of Xi gene expression. Hence, Xist appears

to regulate X chromosome structure through mechanisms other

than the relocation of active genes to transcriptional factories

(Splinter et al., 2011). One intriguing clue is that conditional Xist

deletion also led to loss of PRC2 and H3K27me3 marks. The

conformations of the two X chromosomes appear to be regu-

lated by distinct mechanisms because PRC2 is dispensable for

the topological domains of Xa (Nora et al., 2012). Whether one

or several Xa-expressed lncRNA controls Xa conformation

remains to be seen.

lncRNAs can also regulate the interaction between chromo-

somes, a concept that is exemplified by S. pombe meiosis. In

order for chromosomes to properly segregate in meiosis and

prevent aneuploidy, homologous chromosomes must interact

and generate stable associations. The sme2 locus plays a key

role in the mutual identification of homologous chromosomes

during meiosis, in addition to its role in the mitosis/meiosis

switch discussed above. The meiRNA-L transcript accumulates

at the sme2 locus and is necessary for the robust chromosomal

pairing (Ding et al., 2012). These studies suggest that noncoding

RNAs can be components of a cis-acting pairing factor that

allows homologous chromosomes to identify each other.

Cytoplasmic FunctionsThe ultimate function of mRNAs is to be translated, and like other

steps of gene expression, multiple layers of posttranscriptional

regulation exist in the cytoplasm (Figure 4). lncRNAs can also

‘‘identify’’ mRNAs in the cytoplasm and modulate their life

cycle. Recent works demonstrated that lncRNAs impact both

themRNA half-life and translation of mRNAs. The lncRNA TINCR

(terminal differentiation-induced ncRNA) is induced during

epidermal differentiation and is required for normal induction of

key mediators of epidermal differentiation. TINCR localizes to

the cytoplasm, where it interacts with Staufen 1 protein

(STAU1) to promote the stability of mRNAs containing the TINCR

box motif (Kretz et al., 2013) (Figure 4A). Hence, the TINCR

mechanism is the diametric opposite of posttranscriptional

silencing by small regulatory RNAs like siRNA or miRNAs.

STAU1 can also be programmed by other lncRNAs to facilitate

mRNA degradation. The half-STAU1-binding site RNAs (1/2-

sbsRNAs) contain Alu elements that bind to Alu elements in the

30UTR of actively transcribed target genes, generating

a STAU1-binding site. These mRNAs are therefore identified as

STAU1-mediated messenger RNA decay (SMD) targets (Gong

and Maquat, 2011) (Figure 4B). In addition, a recently identified

class of lncRNA impacts gene expression by promoting transla-

tion of targets mRNAs. Expression of antisenseUchl1RNA leads

to an increase in Uchl1 protein level without any change at the

mRNA level. Antisense Uchl1 lncRNA is composed by a region

that overlaps with the first 73 nucleotides of Uchl1 and two

embedded repetitive sequences, one of which (SINEB2) is

required for the ability of the lncRNA to induce protein transla-

tion. Under stress conditions in which cap-dependent translation

is inhibited, antisense Uchl1 lncRNA, previously enriched in the

nucleus, moves into the cytoplasm and hybridizes with Uchl1

mRNA to enable cap-independent translation of Uchl1. In other

words, the lncRNA acts like a mobile internal ribosomal entry

element to promote selective translation. Other SINEB2-contain-

ing antisense lncRNAs may function in a similar way (Carrieri

et al., 2012) (Figure 4C). Conversely, lincRNA-p21 can inhibit

the translation of target mRNAs. In the absence of HuR,

lincRNA-p21 is stable and interacts with the mRNAs CTNNB1

and JUNB and translational repressor Rck, repressing the trans-

lation of the targeted mRNAs (Yoon et al., 2012) (Figure 4D).

These emerging examples illustrate that lncRNAs can provide

a rich palette of regulatory capacities in the cytoplasm.

Human DiseasesConsidering the wide range of roles that lncRNAs play in cellular

networks, it is not surprising that noncoding RNAs have been

implicated in disease. Genome-wide association studies have

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1303

revealed that only 7% of disease or trait-associated single-

nucleotide polymorphisms (SNPs) reside in protein-coding

exons, whereas 43% of trait-/disease-associated SNP are found

outside of protein-coding genes (Hindorff et al., 2009). In addition

to the example of sno-lncRNAs in Prader-Willi syndrome dis-

cussed above, several recent discoveries of lncRNAs in

Mendelian disorders illustrate the emerging recognition of

lncRNAs in human diseases.

Facioscapulohumeral muscular dystrophy (FSHD) is the third

most common myopathy and is predominantly caused by

a contraction in copy number of the D4Z4 repeats mapping to

4q35. The D4Z4 repeat is the target of several chromatin modifi-

cations, including H3K9me3 and H3K27me3, which are reduced

in FSHD patients. Cabianca et al. found that a long array of

D4Z4 repeats recruit Polycomb complexes to promote the

formation of a repressive chromatin state that inhibits the

expression of genes at 4q35. Loss of D4Z4 repeats results in

derepression of DBE-T, a novel lncRNA that functions in cis

and localizes to the FSHD locus. DBE-T recruits ASH1L (a

component of MLL/TrX complex), leading to improper establish-

ment of active chromatin and expression of genes from 4q35

(Cabianca et al., 2012). Hence, DBE-T is a lncRNA that functions

as a locus control element by promoting active chromatin

domain, and FSHD results from lncRNA ‘‘promoter mutations’’

that perturb DBE-T regulation.

HELLP syndrome (hemolysis, elevated liver enzymes, low

platelets) is a recessively inherited life-threatening pregnancy

complication. Linkage analysis narrowed the HELLP locus to

a gene desert between C12orf48 and IGF1 on 12q23.2, where

a single 205 kb capped and polyadenylated lncRNA is tran-

scribed (van Dijk et al., 2012). Knockdown of this lncRNA

revealed a role in the transition from G2 to mitosis and tropho-

blast cell invasion, although the precise mechanism is still

unclear. Notably, morpholino oligonucleotides complementary

to the mutation site in HELLP lncRNA boosted lncRNA level

and reversed the gene expression and cell invasion defects.

Similarly, deletions in a coding-gene desert at 16q24.1 lead to

alveolar capillary dysplasia with misalignment of pulmonary

veins (ACD/MPV) (Szafranski et al., 2013). This region contains

a distant enhancer of FOXF1, a key regulator of lung develop-

ment. This enhancer element interacts with FOXF1 in human

pulmonary microvascular endothelial cells, but not in lympho-

blasts, suggesting that FOXF1 expression in the lung endothe-

lium is regulated at the chromatin structure levels. In addition

to transcription-factor-binding sites, the focal deletion includes

two lncRNA expressed specifically in the lung. An intriguing

possibility is that the expression of these lncRNAs, which

happens specifically in the lung, contributes to the establishment

of a chromatin loop that brings the enhancer in close proximity to

FOXF1.

Chromosomal translocations lead to inheritable structural and

genetic changes and, as such, are relevant causes of genetic

disease. One way that chromosomal translocations can lead

to disease is through disruption of the higher-order chromatin

organization and the cis-regulatory landscape. Recently, two

different translocations have been identified in brachydactyly

type E (BDE) that implicate lncRNA dysregulation (Maass et al.,

2012). These translocations affect a regulatory region that inter-

1304 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

acts in cis with PTHLH and in trans with SOX9. Interestingly, this

region is home to a lncRNA whose expression is important for

the proper expression of PTHLH and SOX9. Depletion of this

lncRNA (DA125942) resulted in downregulation of PTHLH and

SOX9. The lncRNA interacts with both loci, and the occupancy

is reduced in chromatin originated from BDE patients. This study

demonstrates how lncRNAs and chromatin higher-order organi-

zation collaborate in the regulation of gene expression.

Recognition of the roles of lncRNAs in human disease has

unveiled new diagnostic and therapeutic opportunities. lncRNAs

are expressed in a more tissue-specific fashion than mRNA

genes, a pattern that has been found to hold true in pathologic

states such as cancer (Brunner et al., 2012). lncRNA measure-

ments could hence trace cancer metastases or circulating

cancer cells to their origins. In addition, a strong connection

between lncRNAs and cancer has been clearly established, as

many lncRNAs are dysregulated in human cancers. The lncRNA

HOTAIR in overexpressed in breast, colon, pancreas, and liver

cancers, and overexpression of HOTAIR has been shown to

drive breast cancer metastasis in vivo (Gupta et al., 2010;

Gutschner and Diederichs, 2012). lncRNAs appear to be more

structured and stable than mRNA transcripts, which facilitate

their detection as free nucleic acids in body fluid such as urine

and blood—knowledge already put to good use in clinically

approved tests for prostate cancer (Fradet et al., 2004; Shappell,

2008; Tinzl et al., 2004). Aberrant lncRNAs can be knocked down

in vivo using oligonucleotide ‘‘drugs’’ (Modarresi et al., 2012;

Wheeler et al., 2012), which should spur advance in lncRNA

genetics and therapeutics.

ConclusionslncRNAs are well poised to be molecular address codes,

particularly in the nucleus. On the one hand, transcription of

lncRNAs is often exquisitely regulated, reflecting the particular

developmental stage and external environment that the cell has

experienced. On the other, the capacity of lncRNAs to function

as guides, scaffolds, and decoys endows them with enormous

regulatory potential in gene expression and for spatial control

within the cell. These outstanding properties of long RNAs

have already been leveraged to make designer RNA scaffolds

for synthetic cell circuits (Delebecque et al., 2011). Many ques-

tions remain to be addressed in this rapidly expanding field.

First, the in vivo function of most lncRNAs has not been deter-

mined. An extensive catalog of lncRNAs has recently been

described available for several model organisms (Nam and

Bartel, 2012; Pauli et al., 2012; Ulitsky et al., 2011), opening

the door of a wide array of powerful techniques to be used in

the in vivo study of lncRNAs that will complement the study

of human lncRNAs. In addition, detailed knowledge of struc-

ture-function relationship in lncRNAs is still lacking, which

prohibits the de novo prediction of lncRNA domains and func-

tions that we take for granted in protein-coding transcripts.

New technologies to deconvolute RNA structure and function

(Martin et al., 2012; Wan et al., 2012), probe RNA-chromatin

interactions (Chu et al., 2011; Simon et al., 2011), and track

RNA movement in real time (Paige et al., 2011) will be crucial

for understanding lncRNAs and realizing their therapeutic

potential.

ACKNOWLEDGMENTS

We thank members of the Chang lab for discussion and apologize to

colleagues whose works are not discussed due to space limitation. We

acknowledge support from NIH and California Institute for Regenerative Medi-

cine (H.Y.C.). P.J.B. is the Kenneth G. and Elaine A. Langone Fellow of the

Damon Runyon Cancer Research Foundation. H.Y.C. is an Early Career

Scientist of the Howard Hughes Medical Institute. H.Y.C. is on the Scientific

Advisory Board of RaNA Therapeutics, which works on long noncoding RNAs.

REFERENCES

Ashe, H.L., Monks, J., Wijgerde, M., Fraser, P., and Proudfoot, N.J. (1997).

Intergenic transcription and transinduction of the human beta-globin locus.

Genes Dev. 11, 2494–2509.

Audas, T.E., Jacob, M.D., and Lee, S. (2012). Immobilization of proteins in

the nucleolus by ribosomal intergenic spacer noncoding RNA. Mol. Cell 45,

147–157.

Banfai, B., Jia, H., Khatun, J., Wood, E., Risk, B., Gundling, W.E., Jr., Kundaje,

A., Gunawardena, H.P., Yu, Y., Xie, L., et al. (2012). Long noncoding RNAs are

rarely translated in two human cell lines. Genome Res. 22, 1646–1657.

Bertani, S., Sauer, S., Bolotin, E., and Sauer, F. (2011). The noncoding RNA

Mistral activates Hoxa6 and Hoxa7 expression and stem cell differentiation

by recruiting MLL1 to chromatin. Mol. Cell 43, 1040–1046.

Bierhoff, H., Schmitz, K., Maass, F., Ye, J., and Grummt, I. (2010). Noncoding

transcripts in sense and antisense orientation regulate the epigenetic state of

ribosomal RNA genes. Cold Spring Harb. Symp. Quant. Biol. 75, 357–364.

Bond, A.M., Vangompel, M.J., Sametsky, E.A., Clark, M.F., Savage, J.C.,

Disterhoft, J.F., and Kohtz, J.D. (2009). Balanced gene regulation by an

embryonic brain ncRNA is critical for adult hippocampal GABA circuitry. Nat.

Neurosci. 12, 1020–1027.

Brown, J.A., Valenstein, M.L., Yario, T.A., Tycowski, K.T., and Steitz, J.A.

(2012). Formation of triple-helical structures by the 30-end sequences of

MALAT1 and MENb noncoding RNAs. Proc. Natl. Acad. Sci. USA 109,

19202–19207.

Brunner, A.L., Beck, A.H., Edris, B., Sweeney, R.T., Zhu, S.X., Li, R., Montgom-

ery, K., Varma, S., Gilks, T., Guo, X., et al. (2012). Transcriptional profiling of

lncRNAs and novel transcribed regions across a diverse panel of archived

human cancers. Genome Biol. 13, R75.

Cabianca, D.S., Casa, V., Bodega, B., Xynos, A., Ginelli, E., Tanaka, Y., and

Gabellini, D. (2012). A long ncRNA links copy number variation to a poly-

comb/trithorax epigenetic switch in FSHD muscular dystrophy. Cell 149,

819–831.

Cabili, M.N., Trapnell, C., Goff, L., Koziol, M., Tazon-Vega, B., Regev, A., and

Rinn, J.L. (2011). Integrative annotation of human large intergenic noncoding

RNAs reveals global properties and specific subclasses. Genes Dev. 25,

1915–1927.

Carrieri, C., Cimatti, L., Biagioli, M., Beugnet, A., Zucchelli, S., Fedele, S.,

Pesce, E., Ferrer, I., Collavin, L., Santoro, C., et al. (2012). Long non-coding

antisense RNA controls Uchl1 translation through an embedded SINEB2

repeat. Nature 491, 454–457.

Chen, L.L., and Carmichael, G.G. (2009). Altered nuclear retention of mRNAs

containing inverted repeats in human embryonic stem cells: functional role

of a nuclear noncoding RNA. Mol. Cell 35, 467–478.

Chu, C., Qu, K., Zhong, F.L., Artandi, S.E., and Chang, H.Y. (2011). Genomic

maps of long noncoding RNA occupancy reveal principles of RNA-chromatin

interactions. Mol. Cell 44, 667–678.

Clemson, C.M., Hutchinson, J.N., Sara, S.A., Ensminger, A.W., Fox, A.H.,

Chess, A., and Lawrence, J.B. (2009). An architectural role for a nuclear non-

coding RNA: NEAT1 RNA is essential for the structure of paraspeckles. Mol.

Cell 33, 717–726.

Cremer, T., and Cremer, M. (2010). Chromosome territories. Cold Spring Harb.

Perspect. Biol. 2, a003889.

De Santa, F., Barozzi, I., Mietton, F., Ghisletti, S., Polletti, S., Tusi, B.K., Muller,

H., Ragoussis, J., Wei, C.L., and Natoli, G. (2010). A large fraction of extragenic

RNA pol II transcription sites overlap enhancers. PLoS Biol. 8, e1000384.

Delebecque, C.J., Lindner, A.B., Silver, P.A., and Aldaye, F.A. (2011). Organi-

zation of intracellular reactions with rationally designed RNA assemblies.

Science 333, 470–474.

Derrien, T., Johnson, R., Bussotti, G., Tanzer, A., Djebali, S., Tilgner, H.,

Guernec, G., Martin, D., Merkel, A., Knowles, D.G., et al. (2012). The

GENCODE v7 catalog of human long noncoding RNAs: analysis of their

gene structure, evolution, and expression. Genome Res. 22, 1775–1789.

Ding, D.Q., Okamasa, K., Yamane, M., Tsutsumi, C., Haraguchi, T., Yama-

moto, M., and Hiraoka, Y. (2012). Meiosis-specific noncoding RNA mediates

robust pairing of homologous chromosomes in meiosis. Science 336,

732–736.

Dixon, J.R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., Hu, M., Liu, J.S., and

Ren, B. (2012). Topological domains in mammalian genomes identified by

analysis of chromatin interactions. Nature 485, 376–380.

Djebali, S., Davis, C.A., Merkel, A., Dobin, A., Lassmann, T., Mortazavi, A.,

Tanzer, A., Lagarde, J., Lin, W., Schlesinger, F., et al. (2012). Landscape of

transcription in human cells. Nature 489, 101–108.

Dundr, M., and Misteli, T. (2010). Biogenesis of nuclear bodies. Cold Spring

Harb. Perspect. Biol. 2, a000711.

Eissmann, M., Gutschner, T., Hammerle, M., Gunther, S., Caudron-Herger,M.,

Groß, M., Schirmacher, P., Rippe, K., Braun, T., Zornig, M., and Diederichs, S.

(2012). Loss of the abundant nuclear non-coding RNA MALAT1 is compatible

with life and development. RNA Biol. 9, 1076–1087.

Fradet, Y., Saad, F., Aprikian, A., Dessureault, J., Elhilali, M., Trudel, C., Masse,

B., Piche, L., and Chypre, C. (2004). uPM3, a new molecular urine test for the

detection of prostate cancer. Urology 64, 311–315, discussion 315–316.

Gong, C., and Maquat, L.E. (2011). lncRNAs transactivate STAU1-mediated

mRNA decay by duplexing with 30 UTRs via Alu elements. Nature 470,

284–288.

Grimal, P., and Woloch, M. (1983). Roman cities = Les villes romaines

(Madison, Wis.: University of Wisconsin Press).

Guetg, C., Scheifele, F., Rosenthal, F., Hottiger, M.O., and Santoro, R. (2012).

Inheritance of silent rDNA chromatin is mediated by PARP1 via noncoding

RNA. Mol. Cell 45, 790–800.

Guil, S., Soler, M., Portela, A., Carrere, J., Fonalleras, E., Gomez, A., Villa-

nueva, A., and Esteller, M. (2012). Intronic RNAs mediate EZH2 regulation of

epigenetic targets. Nat. Struct. Mol. Biol. 19, 664–670.

Gupta, R.A., Shah, N., Wang, K.C., Kim, J., Horlings, H.M., Wong, D.J., Tsai,

M.C., Hung, T., Argani, P., Rinn, J.L., et al. (2010). Long non-coding RNA

HOTAIR reprograms chromatin state to promote cancer metastasis. Nature

464, 1071–1076.

Gutschner, T., and Diederichs, S. (2012). The hallmarks of cancer: a long non-

coding RNA point of view. RNA Biol. 9, 703–719.

Guttman,M., Amit, I., Garber, M., French, C., Lin, M.F., Feldser, D., Huarte, M.,

Zuk, O., Carey, B.W., Cassady, J.P., et al. (2009). Chromatin signature reveals

over a thousand highly conserved large non-coding RNAs inmammals. Nature

458, 223–227.

Guttman, M., Donaghey, J., Carey, B.W., Garber, M., Grenier, J.K., Munson,

G., Young, G., Lucas, A.B., Ach, R., Bruhn, L., et al. (2011). lincRNAs act in

the circuitry controlling pluripotency and differentiation. Nature 477, 295–300.

Harigaya, Y., Tanaka, H., Yamanaka, S., Tanaka, K., Watanabe, Y., Tsutsumi,

C., Chikashige, Y., Hiraoka, Y., Yamashita, A., and Yamamoto, M. (2006).

Selective elimination of messenger RNA prevents an incidence of untimely

meiosis. Nature 442, 45–50.

Hindorff, L.A., Sethupathy, P., Junkins, H.A., Ramos, E.M., Mehta, J.P.,

Collins, F.S., and Manolio, T.A. (2009). Potential etiologic and functional impli-

cations of genome-wide association loci for human diseases and traits. Proc.

Natl. Acad. Sci. USA 106, 9362–9367.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1305

Houseley, J., Rubbi, L., Grunstein, M., Tollervey, D., and Vogelauer, M. (2008).

A ncRNA modulates histone modification and mRNA induction in the yeast

GAL gene cluster. Mol. Cell 32, 685–695.

Huarte, M., Guttman, M., Feldser, D., Garber, M., Koziol, M.J., Kenzelmann-

Broz, D., Khalil, A.M., Zuk, O., Amit, I., Rabani, M., et al. (2010). A large inter-

genic noncoding RNA induced by p53 mediates global gene repression in the

p53 response. Cell 142, 409–419.

Hung, T., Wang, Y., Lin, M.F., Koegel, A.K., Kotake, Y., Grant, G.D., Horlings,

H.M., Shah, N., Umbricht, C., Wang, P., et al. (2011). Extensive and coordi-

nated transcription of noncoding RNAs within cell-cycle promoters. Nat.

Genet. 43, 621–629.

Ietswaart, R., Wu, Z., and Dean, C. (2012). Flowering time control: another

window to the connection between antisense RNA and chromatin. Trends

Genet. 28, 445–453.

Ingolia, N.T., Lareau, L.F., and Weissman, J.S. (2011). Ribosome profiling of

mouse embryonic stem cells reveals the complexity and dynamics of mamma-

lian proteomes. Cell 147, 789–802.

IHGSC (International HumanGenomeSequencingConsortium). (2004). Finish-

ing the euchromatic sequence of the human genome. Nature 431, 931–945.

Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A.,

van Berkum, N.L., Ebmeier, C.C., Goossens, J., Rahl, P.B., Levine, S.S.,

et al. (2010). Mediator and cohesin connect gene expression and chromatin

architecture. Nature 467, 430–435.

Khalil, A.M., Guttman, M., Huarte, M., Garber, M., Raj, A., Rivea Morales, D.,

Thomas, K., Presser, A., Bernstein, B.E., van Oudenaarden, A., et al. (2009).

Many human large intergenic noncoding RNAs associate with chromatin-

modifying complexes and affect gene expression. Proc. Natl. Acad. Sci.

USA 106, 11667–11672.

Kim, T.K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin,

D.A., Laptewicz, M., Barbara-Haley, K., Kuersten, S., et al. (2010). Widespread

transcription at neuronal activity-regulated enhancers. Nature 465, 182–187.

Kim, T., Xu, Z., Clauder-Munster, S., Steinmetz, L.M., and Buratowski, S.

(2012). Set3 HDAC mediates effects of overlapping noncoding transcription

on gene induction kinetics. Cell 150, 1158–1169.

Kretz, M., Siprashvili, Z., Chu, C.,Webster, D.E., Zehnder, A., Qu, K., Lee, C.S.,

Flockhart, R.J., Groff, A.F., Chow, J., et al. (2013). Control of somatic tissue

differentiation by the long non-coding RNA TINCR. Nature 493, 231–235.

Published online December 2, 2012. http://dx.doi.org/10.1038/nature11661.

Latos, P.A., Pauler, F.M., Koerner, M.V., Sxenergin, H.B., Hudson, Q.J., Stoc-

sits, R.R., Allhoff, W., Stricker, S.H., Klement, R.M., Warczok, K.E., et al.

(2012). Airn transcriptional overlap, but not its lncRNA products, induces

imprinted Igf2r silencing. Science 338, 1469–1472.

Lee, J.T. (2009). Lessons from X-chromosome inactivation: long ncRNA as

guides and tethers to the epigenome. Genes Dev. 23, 1831–1842.

Lee, J.T., and Bartolomei, M.S. (2013). X-inactivation, imprinting, and long

noncoding RNAs in health and disease. Cell 152, this issue, 1308–1323.

Maass, P.G., Rump, A., Schulz, H., Stricker, S., Schulze, L., Platzer, K., Aydin,

A., Tinschert, S., Goldring,M.B., Luft, F.C., andBahring, S. (2012). Amisplaced

lncRNA causes brachydactyly in humans. J. Clin. Invest. 122, 3990–4002.

Mao, Y.S., Sunwoo, H., Zhang, B., and Spector, D.L. (2011). Direct visualiza-

tion of the co-transcriptional assembly of a nuclear body by noncoding

RNAs. Nat. Cell Biol. 13, 95–101.

Marques, A.C., and Ponting, C.P. (2009). Catalogues of mammalian long non-

coding RNAs: modest conservation and incompleteness. Genome Biol. 10,

R124.

Martens, J.A., Laprade, L., and Winston, F. (2004). Intergenic transcription is

required to repress the Saccharomyces cerevisiae SER3 gene. Nature 429,

571–574.

Martianov, I., Ramadass, A., Serra Barros, A., Chow, N., and Akoulitchev, A.

(2007). Repression of the human dihydrofolate reductase gene by a non-

coding interfering transcript. Nature 445, 666–670.

1306 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Martin, L., Meier, M., Lyons, S.M., Sit, R.V., Marzluff, W.F., Quake, S.R., and

Chang, H.Y. (2012). Systematic reconstruction of RNA functional motifs with

high-throughput microfluidics. Nat. Methods 9, 1192–1194.

Mayer, C., Schmitz, K.M., Li, J., Grummt, I., and Santoro, R. (2006). Intergenic

transcripts regulate the epigenetic state of rRNA genes. Mol. Cell 22, 351–361.

Misteli, T. (2007). Beyond the sequence: cellular organization of genome func-

tion. Cell 128, 787–800.

Mitton-Fry, R.M., DeGregorio, S.J., Wang, J., Steitz, T.A., and Steitz, J.A.

(2010). Poly(A) tail recognition by a viral RNA element through assembly of

a triple helix. Science 330, 1244–1247.

Modarresi, F., Faghihi, M.A., Lopez-Toledano, M.A., Fatemi, R.P., Magistri, M.,

Brothers, S.P., van der Brug, M.P., and Wahlestedt, C. (2012). Inhibition of

natural antisense transcripts in vivo results in gene-specific transcriptional

upregulation. Nat. Biotechnol. 30, 453–459.

Nagano, T., Mitchell, J.A., Sanz, L.A., Pauler, F.M., Ferguson-Smith, A.C., Feil,

R., and Fraser, P. (2008). The Air noncoding RNA epigenetically silences

transcription by targeting G9a to chromatin. Science 322, 1717–1720.

Nakagawa, S., Naganuma, T., Shioi, G., and Hirose, T. (2011). Paraspeckles

are subpopulation-specific nuclear bodies that are not essential in mice. J.

Cell Biol. 193, 31–39.

Nakagawa, S., Ip, J.Y., Shioi, G., Tripathi, V., Zong, X., Hirose, T., and

Prasanth, K.V. (2012). Malat1 is not an essential component of nuclear

speckles in mice. RNA 18, 1487–1499.

Nam, J.W., and Bartel, D.P. (2012). Long noncoding RNAs in C. elegans.

Genome Res. 22, 2529–2540.

Noordermeer, D., Leleu, M., Splinter, E., Rougemont, J., De Laat, W., and

Duboule, D. (2011). The dynamic architecture of Hox gene clusters. Science

334, 222–225.

Nora, E.P., Lajoie, B.R., Schulz, E.G., Giorgetti, L., Okamoto, I., Servant, N.,

Piolot, T., van Berkum, N.L., Meisig, J., Sedat, J., et al. (2012). Spatial partition-

ing of the regulatory landscape of the X-inactivation centre. Nature 485,

381–385.

Ørom, U.A., Derrien, T., Beringer, M., Gumireddy, K., Gardini, A., Bussotti, G.,

Lai, F., Zytnicki, M., Notredame, C., Huang, Q., et al. (2010). Long noncoding

RNAs with enhancer-like function in human cells. Cell 143, 46–58.

Paige, J.S., Wu, K.Y., and Jaffrey, S.R. (2011). RNA mimics of green fluores-

cent protein. Science 333, 642–646.

Pandey, R.R.,Mondal, T., Mohammad, F., Enroth, S., Redrup, L., Komorowski,

J., Nagano, T., Mancini-Dinardo, D., and Kanduri, C. (2008). Kcnq1ot1 anti-

sense noncoding RNA mediates lineage-specific transcriptional silencing

through chromatin-level regulation. Mol. Cell 32, 232–246.

Pauli, A., Valen, E., Lin,M.F., Garber,M., Vastenhouw, N.L., Levin, J.Z., Fan, L.,

Sandelin, A., Rinn, J.L., Regev, A., and Schier, A.F. (2012). Systematic identi-

fication of long noncoding RNAs expressed during zebrafish embryogenesis.

Genome Res. 22, 577–591.

Ponjavic, J., Ponting, C.P., and Lunter, G. (2007). Functionality or transcrip-

tional noise? Evidence for selection within long noncoding RNAs. Genome

Res. 17, 556–565.

Ravasi, T., Suzuki, H., Pang, K.C., Katayama, S., Furuno, M., Okunishi, R.,

Fukuda, S., Ru, K., Frith, M.C., Gongora, M.M., et al. (2006). Experimental vali-

dation of the regulated expression of large numbers of non-coding RNAs from

the mouse genome. Genome Res. 16, 11–19.

Rinn, J.L., and Chang, H.Y. (2012). Genome regulation by long noncoding

RNAs. Annu. Rev. Biochem. 81, 145–166.

Sasaki, Y.T., Ideue, T., Sano, M., Mituyama, T., and Hirose, T. (2009).

MENepsilon/beta noncoding RNAs are essential for structural integrity of

nuclear paraspeckles. Proc. Natl. Acad. Sci. USA 106, 2525–2530.

Schmitt, S., Prestel, M., and Paro, R. (2005). Intergenic transcription through

a polycomb group response element counteracts silencing. Genes Dev. 19,

697–708.

Schmitz, K.M., Mayer, C., Postepska, A., and Grummt, I. (2010). Interaction of

noncoding RNAwith the rDNA promoter mediates recruitment of DNMT3b and

silencing of rRNA genes. Genes Dev. 24, 2264–2269.

Shappell, S.B. (2008). Clinical utility of prostate carcinoma molecular diag-

nostic tests. Rev. Urol. 10, 44–69.

Shevtsov, S.P., and Dundr, M. (2011). Nucleation of nuclear bodies by RNA.

Nat. Cell Biol. 13, 167–173.

Simon, M.D., Wang, C.I., Kharchenko, P.V., West, J.A., Chapman, B.A., Alek-

seyenko, A.A., Borowsky, M.L., Kuroda, M.I., and Kingston, R.E. (2011). The

genomic binding sites of a noncoding RNA. Proc. Natl. Acad. Sci. USA 108,

20497–20502.

Solovei, I., Kreysing, M., Lanctot, C., Kosem, S., Peichl, L., Cremer, T., Guck,

J., and Joffe, B. (2009). Nuclear architecture of rod photoreceptor cells adapts

to vision in mammalian evolution. Cell 137, 356–368.

Splinter, E., de Wit, E., Nora, E.P., Klous, P., van de Werken, H.J., Zhu, Y.,

Kaaij, L.J., van Ijcken, W., Gribnau, J., Heard, E., and de Laat, W. (2011).

The inactive X chromosome adopts a unique three-dimensional conformation

that is dependent on Xist RNA. Genes Dev. 25, 1371–1383.

Sunwoo, H., Dinger, M.E., Wilusz, J.E., Amaral, P.P., Mattick, J.S., and

Spector, D.L. (2009). MEN epsilon/beta nuclear-retained non-coding RNAs

are up-regulated upon muscle differentiation and are essential components

of paraspeckles. Genome Res. 19, 347–359.

Szafranski, P., Dharmadhikari, A.V., Brosens, E., Gurha, P., Kolodziejska, K.E.,

Ou, Z., Dittwald, P., Majewski, T., Mohan, K.N., Chen, B., et al. (2013). Small

non-coding differentially-methylated copy-number variants, including lncRNA

genes, cause a lethal lung developmental disorder. Genome Res. 23, 23–33.

Published online October 3, 2012. http://dx.doi.org/10.1101/gr.141887.112.

Tavares, L., Dimitrova, E., Oxley, D., Webster, J., Poot, R., Demmers, J.,

Bezstarosti, K., Taylor, S., Ura, H., Koide, H., et al. (2012). RYBP-PRC1

complexes mediate H2A ubiquitylation at polycomb target sites independently

of PRC2 and H3K27me3. Cell 148, 664–678.

Tian, D., Sun, S., and Lee, J.T. (2010). The long noncoding RNA, Jpx, is

a molecular switch for X chromosome inactivation. Cell 143, 390–403.

Tilgner, H., Knowles, D.G., Johnson, R., Davis, C.A., Chakrabortty, S., Djebali,

S., Curado, J., Snyder, M., Gingeras, T.R., and Guigo, R. (2012). Deep

sequencing of subcellular RNA fractions shows splicing to be predominantly

co-transcriptional in the human genome but inefficient for lncRNAs. Genome

Res. 22, 1616–1625.

Tinzl, M., Marberger, M., Horvath, S., and Chypre, C. (2004). DD3PCA3 RNA

analysis in urine—a new perspective for detecting prostate cancer. Eur.

Urol. 46, 182–186, discussion 187.

Tsai, M.C., Manor, O., Wan, Y., Mosammaparast, N., Wang, J.K., Lan, F., Shi,

Y., Segal, E., and Chang, H.Y. (2010). Long noncoding RNA as modular

scaffold of histone modification complexes. Science 329, 689–693.

Tycowski, K.T., Shu, M.D., Borah, S., Shi, M., and Steitz, J.A. (2012). Conser-

vation of a triple-helix-forming RNA stability element in noncoding and

genomic RNAs of diverse viruses. Cell Rep. 2, 26–32.

Ulitsky, I., Shkumatava, A., Jan, C.H., Sive, H., and Bartel, D.P. (2011).

Conserved function of lincRNAs in vertebrate embryonic development despite

rapid sequence evolution. Cell 147, 1537–1550.

van Dijk, M., Thulluru, H.K., Mulders, J., Michel, O.J., Poutsma, A., Windhorst,

S., Kleiverda, G., Sie, D., Lachmeijer, A.M., and Oudejans, C.B. (2012). HELLP

babies link a novel lincRNA to the trophoblast cell cycle. J. Clin. Invest. 122,

4003–4011.

van Werven, F.J., Neuert, G., Hendrick, N., Lardenois, A., Buratowski, S., van

Oudenaarden, A., Primig, M., and Amon, A. (2012). Transcription of two long

noncoding RNAs mediates mating-type control of gametogenesis in budding

yeast. Cell 150, 1170–1181.

Verma-Gaur, J., Torkamani, A., Schaffer, L., Head, S.R., Schork, N.J., and

Feeney, A.J. (2012). Noncoding transcription within the Igh distal V(H) region

at PAIR elements affects the 3D structure of the Igh locus in pro-B cells.

Proc. Natl. Acad. Sci. USA 109, 17004–17009.

Wan, Y., Qu, K., Ouyang, Z., Kertesz, M., Li, J., Tibshirani, R., Makino, D.L.,

Nutter, R.C., Segal, E., and Chang, H.Y. (2012). Genome-wide measurement

of RNA folding energies. Mol. Cell 48, 169–181.

Wang, K.C., and Chang, H.Y. (2011). Molecular mechanisms of long noncod-

ing RNAs. Mol. Cell 43, 904–914.

Wang, X., Arai, S., Song, X., Reichart, D., Du, K., Pascual, G., Tempst, P.,

Rosenfeld, M.G., Glass, C.K., and Kurokawa, R. (2008). Induced ncRNAs allo-

sterically modify RNA-binding proteins in cis to inhibit transcription. Nature

454, 126–130.

Wang, K.C., Yang, Y.W., Liu, B., Sanyal, A., Corces-Zimmerman, R., Chen, Y.,

Lajoie, B.R., Protacio, A., Flynn, R.A., Gupta, R.A., et al. (2011). A long noncod-

ing RNA maintains active chromatin to coordinate homeotic gene expression.

Nature 472, 120–124.

Wheeler, T.M., Leger, A.J., Pandey, S.K., MacLeod, A.R., Nakamori, M.,

Cheng, S.H., Wentworth, B.M., Bennett, C.F., and Thornton, C.A. (2012).

Targeting nuclear RNA for in vivo correction of myotonic dystrophy. Nature

488, 111–115.

Wilusz, J.E., Jnbaptiste, C.K., Lu, L.Y., Kuhn, C.D., Joshua-Tor, L., and Sharp,

P.A. (2012). A triple helix stabilizes the 30 ends of long noncoding RNAs that

lack poly(A) tails. Genes Dev. 26, 2392–2407.

Wolpert, L. (2011). Positional information and patterning revisited. J. Theor.

Biol. 269, 359–365.

Yamamoto, M. (2010). The selective elimination of messenger RNA underlies

the mitosis-meiosis switch in fission yeast. Proc. Jpn. Acad., Ser. B, Phys.

Biol. Sci. 86, 788–797.

Yang, L., Lin, C., Liu,W., Zhang, J., Ohgi, K.A., Grinstein, J.D., Dorrestein, P.C.,

and Rosenfeld, M.G. (2011). ncRNA- and Pc2 methylation-dependent gene

relocation between nuclear structures mediates gene activation programs.

Cell 147, 773–788.

Yin, Q.F., Yang, L., Zhang, Y., Xiang, J.F., Wu, Y.W., Carmichael, G.G., and

Chen, L.L. (2012). Long noncoding RNAs with snoRNA ends. Mol. Cell 48,

219–230.

Yoon, J.H., Abdelmohsen, K., Srikantan, S., Yang, X., Martindale, J.L., De, S.,

Huarte, M., Zhan, M., Becker, K.G., and Gorospe, M. (2012). LincRNA-p21

suppresses target mRNA translation. Mol. Cell 47, 648–655.

Zhang, B., Arun, G., Mao, Y.S., Lazar, Z., Hung, G., Bhattacharjee, G., Xiao, X.,

Booth, C.J., Wu, J., Zhang, C., and Spector, D.L. (2012). The lncRNAMalat1 is

dispensable for mouse development but its transcription plays a cis-regulatory

role in the adult. Cell Rep. 2, 111–123.

Zhao, J., Ohsumi, T.K., Kung, J.T., Ogawa, Y., Grau, D.J., Sarma, K., Song,

J.J., Kingston, R.E., Borowsky, M., and Lee, J.T. (2010). Genome-wide identi-

fication of polycomb-associated RNAs by RIP-seq. Mol. Cell 40, 939–953.

Zofall, M., Yamanaka, S., Reyes-Turcu, F.E., Zhang, K., Rubin, C., and Grewal,

S.I. (2012). RNA elimination machinery targeting meiotic mRNAs promotes

facultative heterochromatin formation. Science 335, 96–100.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1307

Leading Edge

Review

X-Inactivation, Imprinting, and LongNoncoding RNAs in Health and Disease

Jeannie T. Lee1,2,* and Marisa S. Bartolomei3,*1Howard Hughes Medical Institute, Department of Molecular Biology, Massachusetts General Hospital, Boston, MA 02114, USA2Department of Genetics, Harvard Medical School, Boston, MA 02114, USA3Department of Cell and Developmental Biology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA 19104, USA

*Correspondence: lee@molbio.mgh.harvard.edu (J.T.L.), bartolom@mail.med.upenn.edu (M.S.B.)

http://dx.doi.org/10.1016/j.cell.2013.02.016

X chromosome inactivation and genomic imprinting are classic epigenetic processes that causedisease when not appropriately regulated in mammals. Whereas X chromosome inactivationevolved to solve the problem of gene dosage, the purpose of genomic imprinting remains contro-versial. Nevertheless, the two phenomena are united by allelic control of large gene clusters,such that only one copy of a gene is expressed in every cell. Allelic regulation poses significantchallenges because it requires coordinated long-range control in cis and stable propagation overtime. Long noncoding RNAs have emerged as a common theme, and their contributions to diseasesof imprinting and the X chromosome have become apparent. Here, we review recent advancesin basic biology, the connections to disease, and preview potential therapeutic strategies for futuredevelopment.

IntroductionEvery organism faces the challenge of regulating gene dosage.

In diploids, genes are generally assumed to be expressed from

both alleles but, in mammals, several classes of genes are ex-

pressed from only one allele per cell. Two of the most prominent

examples of allelic phenomena are X chromosome inactivation

(XCI) and genomic imprinting. Because of XCI, only one copy

of each X-linked gene is active in female cells (XX). Because

male cells carry only one X chromosome (XY), they are inherently

hemizygous for all X-linked genes. In genomic imprinting,

genes within a discrete domain are coordinately regulated and

expressed according to parent of origin. Research over the

past 50 years has identified many similarities between XCI and

genomic imprinting. Apart from monoallelic expression, genes

subject to the two processes tend to cluster, are influenced at

long-range by a master control region, and are associated with

multiple long noncoding RNAs (lncRNA). Some of the most

fascinating stories to emerge in recent years have been related

to lncRNAs as master regulators. Some of the first epigenetic

lncRNAs in mammals were, in fact, identified from genomic

imprinting and XCI studies. Such lncRNAs have been proposed

to serve as recruiting tools for chromatin-modifying complexes

that would in turn silence or activate genes residing within the

allelically regulated clusters.

Together, XCI and imprinting affect expression of �5%–10%

of genes in the mammalian genome. From a functional stand-

point, mutations within these regions could be easily unmasked,

as they are often unbuffered by contributions from the silenced

wild-type copy and could thereby cause severe developmental

defects. This explains why X-linked and imprinted diseases

are among the most common congenital human disorders,

accounting for easily recognizable childhood syndromes such

1308 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

as Rett, fragile X, Prader-Willi/Angelman, and Beckwith-Wiede-

mann syndromes, as well as conditions such as hemophilia,

testicular feminization, and red-green color blindness. More

recently, imprinting and X-linked anomalies have also been of

interest for stem cell maturation and reprogramming, cancer,

assisted reproductive technology (ART), and cognition. This

article will review the state-of-the-art in genomic imprinting and

XCI, focusing on recent advances in studying mechanism,

the emerging roles of lncRNAs, and their relevance for under-

standing and treating various human conditions.

X Chromosome Inactivation and Genomic ImprintingGenomic Imprinting

Mammals require both maternal and paternal genomic contribu-

tions to develop into healthy, viable organisms (Solter, 1988).

This is, in large part, due to the inheritance of autosomally im-

printed genes, which are expressed only from a single allele in

accordance with its parent of origin (Bartolomei, 2009). That is,

imprinted genes are expressed either from the maternally or

paternally inherited allele, so that, when summed across the

whole genome, contributions from both parents are necessary

for expression of the full complement of imprinted genes and

for proper development. The elegant nuclear transplantation

experiments by Solter and Surani in the 1980s were the first

to suggest that the mammalian genome harbored imprinted

genes (McGrath and Solter, 1984; Surani et al., 1984). They

showed that diploid androgenetic embryos derived from two

male pronuclei or diploid gynogenetic embryos derived from

two female pronuclei failed to develop and reasoned that this

must be due to genes that are expressed exclusively from one

of the parental genomes. Later genetic experiments extended

these findings by demonstrating that the proposed imprinted

Figure 1. Mechanisms of Imprinting(A)The insulator model is exemplified by the H19/Igf2 domain. Here, the in-tergenic ICR is paternally methylated. On the unmethylated maternal allele,CTCF binding prevents enhancers from interacting with the Igf2 promoter.Instead, the enhancers activate H19 expression. On the paternal allele,methylation of the ICR spreads to the H19 promoter, silencing its expression,and prevents CTCF from binding the ICR, allowing the enhancers to activateIgf2 expression.(B–D) The ncRNA model is illustrated by the Kcnq1 (B), Igf2r (C), and Snprn (D)domains.(B) For Kcnq1, the ICR contains the promoter of the Kcnq1ot1 lncRNA. On thepaternal allele, the ICR is unmethylated, allowing Kcnq1ot1 expression.Kcnq1ot1 expression silences the paternal allele of the linked genes in cis. Onthe maternal allele, Kcnq1ot1 is not expressed due to methylation of the ICR,and the adjacent imprinted genes are expressed.(C) For the Igf2r domain, transcription of the Airn lncRNA is governed bya promoter within the ICR and is expressed from the unmethylated paternalallele. In somatic cells, transcription of Airn over the Igf2r promoter precludesIgf2r expression, in part by kicking RNA polymerase II (POL-II) off of thepromoter. In extraembryonic lineages, Airn lncRNA is postulated to recruitenzymes that confer repressive histone modifications to silence genes in cis.(D) The Snrpn locus uses the ncRNA model. Ube3a is expressed from thematernal allele exclusively in brain (in other tissues, it is biallelically expressed).The lncRNA on the paternal allele occurs in multiple, variably processed

genes mapped to specific mouse chromosomes (Searle and

Beechey, 1978; Cattanach, 1986).

The current number of imprinted genes in the mouse is

approximately 150 (http://www.mousebook.org/catalog.php?

catalog=imprinting), with a smaller number identified in humans,

in part because many genes have not been tested in humans

(Weksberg, 2010). The imprinted genes are typically located

in clusters of 3–12 (or more) genes that are spread over 20–

3,700 kb of DNA (Barlow, 2011), but, interestingly, genes within

one cluster are not necessarily expressed from the same

parental chromosome (Figure 1). Most imprinted clusters contain

protein coding genes and noncoding RNAs (ncRNAs). The

ncRNAs are of different varieties (microRNAs, snoRNAs, and

lncRNAs), some of which are essential to the mechanism that

imprints these genes in cis. Each well-studied cluster has a

discrete imprinting control region (ICR) that exhibits parent-of-

origin-specific epigenetic modifications (DNA methylation and

posttranslational histone modifications) and governs the

imprinting of the locus. Although the mechanism(s) that confer

the allele-specific epigenetic modifications is poorly understood,

DNA methylation has been shown to be imposed at a precise

time in germ cells by a mechanism that is postulated to involve

transcription (Chotalia et al., 2009; Henckel et al., 2012) and is

maintained after fertilization despite extensive reprogramming

of the genome (Bartolomei and Ferguson-Smith, 2011) (Figure 1).

Moreover, germline deletion of the ICR results in the loss of

imprinting of multiple genes in the cluster, thus demonstrating

that the clustering of imprinted genes is required for their appro-

priate expression.

Many imprinted genes undergo tissue-specific imprinting. Of

the approximately 150 imprinted genes identified in mouse,

a few are imprinted exclusively in the placenta, an extraembry-

onic organ that plays a crucial role in regulating fetal growth by

controlling the supply of nutrients (Frost and Moore, 2010).

Imprinting is hypothesized to be a mechanism to balance

growth, with many imprinted genes having a demonstrated

role in growth control. Thus, the placenta is particularly signifi-

cant in the physiology and study of imprinting. Interestingly, as

described below, mechanisms that control imprinting in the

placenta—a short-lived organ—may differ from those mecha-

nisms that regulate imprinting in the much longer-lived somatic

lineages. This holds true not only for autosomal but also X-linked

imprinting.

X Chromosome Inactivation

In 1949, Murray Barr showed that the sex of cat cells could be

deduced by a subnuclear structure now called the ‘‘Barr body’’

in honor of his seminal work (Barr and Bertram, 1949). Susumu

forms, some of which are brain-specific variants that contain upstreampromoters/exons and sequences overlapping withUbe3a. Expression of theselncRNAs occurs when the ICR is unmethylated, with the result that expressionofUbe3a is repressed. On the maternal allele, transcription of the upstream (U)exons is proposed to direct the maternal methylation imprint at the ICR.Topoisomerase I inhibitors identified by amouse screen activate Ube3a on thepaternal allele. As a result, Snprn and Ube3a-ATS were no longer expressedand the ICR exhibited increased methylation, relative to the wild-type paternalallele.All imprinted domains, which are not drawn to scale, are depicted forthe mouse, although the human regions are largely conserved. T refers to thetelomeric end of the cluster and C the proximal end of the chromosome.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1309

Figure 2. The X-Inactivation and X-Reactivation Cycle during Mouse DevelopmentMammalian dosage compensation occurs within a continual cycle of XCI and X reactivation. The XCI cycle begins in the male germline. During the first meioticprophase of spermatogenesis, the X and Y undergo MSCI. After meiosis, 85% of X-linked genes remain suppressed through spermiogenesis, formingpostmeiotic sex chromatin (PMSC). This germline-inactivated X has been proposed to be passed onto the next generation in a partially silent state, accounting forthe preferential XP inactivation of the early female mouse embryo. In the two-cell mouse embryo, repetitive elements on XP are already suppressed. XP-linkedcoding genes are initially active but become progressively inactivated during preimplantation development. The maternal germline also plays a crucial rolein imprinted XCI by marking the future XM during the oocyte growth phase, ensuring that XM is protected in both XX and XY embryos. Beyond the blastocyst,these marks persist only in the placenta of the mouse. Whereas extraembryonic tissues, including the primitive endoderm (PE) and the trophectoderm (TE),maintain imprinted XP inactivation, the epiblast lineage undergoes XCR and initiates zygotically driven random XCI. XCR also occurs in primordial germ cells(PGCs) in preparation for equal segregation during meiosis. Xp, paternal X; Xm, maternal X; Xa, active X; MSCI, meiotic sex chromosome inactivation. Adaptedfrom Payer and Lee, 2008.

Ohno later demonstrated that the Barr body is a condensed

X chromosome (Ohno et al., 1959), and Mary Lyon followed

with the understanding that the condensed X is the result of

whole-chromosome silencing (Lyon, 1961). We now know that

XCI compensates for dosage differences between males and

females by rendering all cells functionally monosomic for the

X chromosome (reviewed in Payer and Lee, 2008; Starmer and

1310 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Magnuson, 2009; Wutz, 2011). XCI is coordinated by an X-inac-

tivation center (Xic), which controls most, if not all, of the steps of

XCI, including X chromosome counting, random X chromosome

choice, and the initiation of silencing along �1,000 genes of the

X (Brown et al., 1991b) (Figure 2). These steps are completed in

the peri-implantation embryo within the 10–20 cell epiblast

lineage (which gives rise to all somatic cells) (Puck et al., 1992).

Once established, the pattern of XCI is stably propagated in the

soma, with the same X chromosome maintained as Xi in subse-

quent mitotic divisions. The mammalian female is therefore

a mosaic (Figure 2).

Whereas the choice of XCI in somatic cells of eutherian

mammals occurs randomly, the choice in marsupial mammals

is fixed. In marsupials (Sharman, 1971), and also in the extraem-

bryonic tissues of some eutherian mammals (Takagi and Sasaki,

1975), the paternal X (XP) is imprinted to undergo silencing,

providing a first example of mammalian imprinting. The phenom-

enon is conceptually similar to autosomal imprinting in that

monoallelic expression is determined by parent-of-origin and

has mechanistic underpinnings in the parental germline. Im-

printed XCI in the placenta adds to the number of imprinted

mammalian genes and further supports the idea that imprinting

balances fetal growth by controlling the nutrient supply (Frost

and Moore, 2010).

Mammalian dosage compensation occurs within a continual

cycle of XCI and reactivation (XCR) (Figure 2). Although random

XCI is female-specific, X silencing also occurs in the male germ-

line (Lifschytz and Lindsley, 1972). For somemammals, the male

germline is where the XCI cycle begins. During the first meiotic

prophase of spermatogenesis, the X and Y undergo ‘‘meiotic

sex chromosome inactivation’’ (MSCI) and form the ‘‘XY

body’’. The X and Y do not wholly reactivate after completion

of meiosis; in mice, 85% of genes on the X chromosome remain

transcriptionally suppressed in postmeiotic spermiogenesis

(Namekawa et al., 2006). This ‘‘postmeiotic sex chromatin’’

(PMSC) is decorated by distinct heterochromatic signatures

(Greaves et al., 2006; Namekawa et al., 2006; Turner et al.,

2006) and is consistent with the hypothesis that the germline-

inactivated X may be passed onto the next generation at least

in a partially preinactivated state, accounting for the preferential

XP inactivation that occurs in the early female embryo (Cooper,

1971; Lyon, 1999; Huynh and Lee, 2003).

At zygotic gene activation in the two-cell mouse embryo,

transcription of repetitive elements on XP is already suppressed,

reflecting their suppression in the male germline (Namekawa

et al., 2006; Namekawa et al., 2010) (Figure 2). Although X-linked

coding genes on the XP are initially active, they are progressively

inactivated during preimplantation development (Okamoto and

Heard, 2006; Kalantry et al., 2009; Namekawa et al., 2010).

The X-linked repetitive elements may facilitate formation of the

silent compartment for inactivation of XP genes (Namekawa

et al., 2010). Thus, imprinted XCI may be a process that begins

in the male germline, continues into the zygote as repeat

silencing, and progresses through the blastocyst stage with

genic silencing. However, the maternal germline also plays

a crucial role in imprinted XCI by marking the future XM to resist

silencing (Takagi and Abe, 1990; Goto and Takagi, 2000). This

occurs during the oocyte growth phase (Tada et al., 2000),

ensuring that XM (passed onto both XX and XY embryos) is pro-

tected (Figure 2). Thus, it is likely that both XP and XM are paren-

tally marked, with XP subject to imprinted XCI and XM protected

from it.

Beyond the blastocyst, these marks persist only in the

placenta of the mouse (Figure 2). The blastocyst consists of

the trophectodermal lineage, which gives rise to placental tissue,

and the inner cell mass, which gives rise to the epiblast lineage

that develops into the embryo proper. During peri-implantation

development, their epigenetic fates diverge with respect to

XCI. Whereas extraembryonic tissues, including the primitive

endoderm (PE) and the trophectoderm (TE), maintain imprinted

XP inactivation, the epiblast lineage undergoes XCR and initiates

a new round of inactivation—this time randomly without

a parent-of-origin bias (Mak et al., 2004).

MechanismsCis-Acting Control Regions

Both XCI and genomic imprinting are regulated by cis-acting

master control regions. For XCI, a single Xic has been mapped

to a 100–500 kb region (Brown et al., 1991b; Lee et al., 1999b;

Chureau et al., 2002) (Figure 3A). Genetic analyses based on

knockouts, gain-of-function mutations, and transgenic

overexpression have shown that the Xic is necessary and

sufficient to regulate XCI. Deleting the noncoding locus Xist

results in loss of silencing capability in cis (Penny et al., 1996;

Marahrens et al., 1997), and placing theXic at an ectopic location

results in counting, choosing, and silencing of the autosome (Lee

et al., 1996; Migeon et al., 1999; Wutz et al., 2002). The Xic there-

fore drives XCI without a requirement for additional X-specific

elements, such as those that might be responsible for the spread

of silencing.

Similarly, genomic imprinting is regulated by cis-acting ICRs

that influence allelic expression across long distances. Whereas

the Xic controls 150 Mb of a chromosome, ICRs control gene

clusters of 0.1–3.0 Mb. Within the clusters, the direction of

transcription and distribution of maternally versus paternally

imprinted genes can vary (Figure 1). However, nearly all im-

printed clusters studied to date contain at least one each of

maternally expressed and paternally expressed genes. ICRs

are usually just a few kilobases in length with allele-specific

DNA methylation and chromatin modifications, but their ICR’s

location relative to the genes can also vary. Most ICRs are meth-

ylated in the female germline during oocyte growth (Bartolomei

and Ferguson-Smith, 2011). A few, including the ICRs for the

H19/Igf2 and Gtl2/Dlk1 clusters, are methylated on the paternal

allele prior to birth in the male germline (Bartolomei and Fergu-

son-Smith, 2011) (Figure 1A). Maternally methylated ICRs

typically harbor the promoter for lncRNAs, examples of which

include the ICRs for Kcnq1ot1, Snprn, and Airn (Figures 1B–

1D). In contrast, paternally methylated ICRs are intergenic

(Barlow, 2011). In the case of the H19/Igf2 locus, the ICR serves

as a methylation-sensitive insulator (Figure 1A). In all cases, ICR

deletions result in the loss of imprinting of multiple genes within

the cluster.

Long Noncoding RNAs

The X-Inactivation Center. Noted early in the study of both

phenomena, a prominent feature of the Xic and ICRs is their

association with lncRNAs, the prototypes of which were dis-

covered within these regions (Brannan et al., 1990; Borsani

et al., 1991; Brown et al., 1991a; Lee et al., 1999a; Koerner

et al., 2009). With respect to function and mechanism, the Xic

harbors some of the best-characterized lncRNAs. The ‘‘X-inac-

tive-specific transcript’’ (XIST/Xist) (Brockdorff et al., 1992;

Brown et al., 1992) produces a 17–20 kb RNA that decorates

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1311

Figure 3. The Control Center and Steps of Initiation during X Chromosome Inactivation(A) The X-inactivation center consists of multiple genes encoding lncRNA, including Xist, RepA, Tsix, Xite, Jpx/Enox, Ftx, and Tsx. Regions involved in varioussteps of XCI (counting, choice, pairing, and silencing) are delineated.(B) Converging pathways of RNA-protein interactions during XCI. Yellow ovals represent chromatin complexes (PRC2, YY1, DNMT3a, RNF12, and REX1) thatinteract with indicated lncRNA or associated loci. Positive regulation shown by pointed arrows; negative regulation shown by blunted arrows. Various steps ofXCI are shown in blue lettering.(C) Initiation of XCI by lncRNA. (1) Biallelic Tsix prevents loading of RepA-PRC2 and initiation of XCI; (2) Two events enable Xist expression during cell differ-entiation: induction of the Jpx activation andmonoallelic loss of Tsix on Xi, which allows RepA-PRC2 to load; (3) Xist cotranscriptionally recruits PRC2. YY1 bindsXi nucleation center, but is blocked from binding Xa; (4) Xist-PRC2 complex cotranscriptionally loads onto the YY1-based nucleation center; (5) From thenucleation center, Xist-PRC2 spreads in a cis-limited fashion to �150 strong Polycomb stations, which in turn spread H3K27me3 via 3,000–4,000 moderatePolycomb sites.

the X chromosome during the initiation of XCI (Clemson et al.,

1996). Xist is expressed only from the Xi and is required for

whole-chromosome silencing (Penny et al., 1996; Marahrens

et al., 1997). Xist RNA directs chromatin and transcriptional

change by binding Polycomb repressive complex 2 (PRC2),

the epigenetic complex responsible for trimethylation of histone

H3 at lysine 27 (H3K27me3), and targeting PRC2 to the Xi (Zhao

et al., 2008) (Figure 3B). This discovery suggests RNA as a crucial

guiding factor in Polycomb targeting. However, PRC2 targeting

and binding to the chromatin are biologically separable, as

indeed chromatin loading is precluded when Xist’s antisense

partner, Tsix (Lee et al., 1999a), is expressed in cis (Zhao et al.,

2008). Only when Tsix expression is downregulated during

development does the Xist-PRC2 complex load onto the Xi

‘‘nucleation center’’ within Xist’s exon 1 (Jeon and Lee, 2011).

The nucleation center consists of three binding sites for the

transcription factor, YY1, a protein bound only to the Xi allele.

By cotranscriptionally tethering Xist RNA to the Xic, YY1 bridges

PRC2, Xist RNA, and Xi chromatin (Figure 3B).

From the nucleation center, PRC2 spreads initially to �150

strong binding sites along the future Xi, concentrating predomi-

1312 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

nantly within bivalent domains coinciding with CpG islands

(Pinter et al., 2012) (Figure 3C). As XCI proceeds, the coating

of the future Xi by Xist RNA correlates with recruitment of

3,000–4,000 moderate Polycomb sites, most of which are inter-

genic, nonbivalent, and lack CpG islands. The moderate sites

cluster around strong sites and facilitate the spreading of

H3K27me3 in a graded concentration relative to strong sites.

Interestingly, Polycomb stations are not enriched for the LINE1

repeats previously hypothesized to influence spreading (Lyon,

2003; Chow et al., 2010). Thus, the spreading of XCI is also

controlled by Xist RNA and is governed by a hierarchy of defined

Polycomb stations along the Xi.

Xist is itself controlled by other lncRNAs, some acting

negatively (e.g., Tsix), others positively (e.g., Jpx). Tsix RNA,

the antisense partner of Xist RNA (Lee et al., 1999a), represses

Xist in several ways. First, Tsix coordinates X chromosome

pairing to generate the epigenetic asymmetry required for

selection of future Xa and Xi (Bacher et al., 2006; Xu et al.,

2006; Xu et al., 2007). Second, Tsix also recruits DNA methyl-

transferase (Dnmt3a) to silence Xist (Sado et al., 2005;

Sun et al., 2006). Third, it blocks recruitment of PRC2 to Xist

(and RepA, see below), potentially by binding PRC2 and titrating

it from Xist/RepA RNAs (Zhao et al., 2008). Tsix also duplexes

with Xist/RepA RNA (Ogawa et al., 2008) and possibly serves

as decoy for PRC2 recruitment (by titrating Xist-RepA RNA or

PRC2). In these ways, Tsix determines allelic choice by repres-

sing Xist transcription on one allele (Figure 3B).

Xist is regulated positively by Jpx RNA (Tian et al., 2010)

(Figure 3C). Deleting Jpx abolishes Xist activation, indicating

that Jpx is a positive regulator. Because knocking down the

RNA recapitulates the deletion, Jpx must function as an RNA

and not merely through its act of being transcribed. Moreover,

because Jpx expression from an autosomal transgene can

rescue the X-linked deletion, Jpx RNA is trans acting, unlike

other elements of the Xic. The 1.6 kb RepA RNA (intragenic to

Xist) has also been implicated as a potential activator of Xist

expression, as its expression appears to be necessary for Xist

upregulation (Zhao et al., 2008) and deleting the Repeat A motif

(Hoki et al., 2009) abolishes Xist induction. The linked noncoding

Ftx locus has also been suggested to regulate Xist, as deleting

Ftx in male cells has mild effects on the chromatin profile of

Xist (Chureau et al., 2011), but its effects in female cells are

currently unknown. These Xist regulators work in parallel with

the E3 ubiquitin ligase, RNF12, encoded by an X-linked gene

near the Xic (Figures 3A and 3B): Its overexpression causes

partial derepression of Xist (Jonkers et al., 2009), and knockouts

of Rnf12 block imprinted XCI and delay random XCI (Shin et al.,

2010; Barakat et al., 2011). The pluripotency factor, REX1, has

been identified as a target of RNF12 (Gontan et al., 2012). It is

thought that elimination of REX1 binding to the Xist promoter

facilitates activation of Xist. These studies collectively point to

central functions for lncRNA-protein interactions, with the

lncRNAs targeting epigenetic complexes, serving as antisense

inhibitors, and activating sense transcription.

Imprinting Clusters. Every imprinted cluster harbors lncRNAs

(Figure 1), many of which originate within or near ICRs. These

lncRNAs are themselves imprinted. The most common mecha-

nism used for imprinting relies on expression of a lncRNA in cis

and exploits much of what has been identified for silencing of

the X chromosome during XCI (Figure 1). There are currently

six well-characterized clusters of imprinted genes (along with

at least nine additional less well-studied clusters), including

Igf2r/Airn, Kcnq1, Snprn/Ube3A, Gnas, Igf2/H19, Dlk1/Gtl2

(Barlow, 2011). All of these clusters contain lncRNAs. Three

imprinted lncRNAs are long mature RNAs: Airn is 108 kb (Lyle

et al., 2000), Kcnq1ot1 is approximately 100 kb (Pauler et al.,

2012), and, Ube3a-ATS may be in excess of 1,000 kb (Pauler

et al., 2012). The Gtl2 lncRNA contains multiple alternatively

spliced transcripts, however, downstream intergenic transcrip-

tion has also been noted, suggesting longer RNAs are likely

(Tierling et al., 2006). Nespas lncRNA exceeds 27 kb (Robson

et al., 2012).

Experiments that directly test the role of the lncRNA itself

have now been performed for a number of loci (Airn, Nespas,

Kcnq1ot1, and Ube3aats). Thus far, all have been analyzed by

genetic manipulation of the endogenous locus to truncate the

lncRNA by insertion of a polyadenylation signal. The 108 kb

Airn lncRNA has been examined in the most detail. Initially,

Barlow and colleagues reported that truncation of Airn to 3 kb

in a mouse model suggested that the lncRNA itself is necessary

to silence all 3 mRNA genes in the Igf2r cluster, indicating a clear

regulatory role for this lncRNA (Sleutels and Barlow, 2002)

(Figure 1A). Similarly, truncation of the �100 kb Kcnq1ot1

lncRNA to 1.5 kb showed that this lncRNA was directly needed

to silence all 10 mRNA genes in the larger Kcnq1 cluster

(Mancini-Dinardo et al., 2006) (Figure 1B), and truncation of the

�27 kb Nespas lncRNA showed it was necessary to silence

the overlappedNesp gene in theGnas imprinted cluster (William-

son et al., 2011). Lastly, truncated Ube3a-ATS in an embryonic

stem (ES) cell model resulted in activation of paternal Ube3a

(Figure 1D), consistent with the role for the Ube3a-ATS lncRNA

in repressing paternal Ube3a in neurons (Meng et al., 2012)

(Figure 1).

At this point, it is not clear how the lncRNAs silence imprinted

genes in cis. One possibility is that they overlap adjacent im-

printed genes and the sense-antisense overlap causes a form

of transcriptional interference of a promoter or an enhancer,

which in turn affects transcription from the mRNA promoter

(Pauler et al., 2012). In this case, the first event could be silencing

of the overlapped promoter or enhancer followed by accumu-

lation of repressive chromatin that can spread and induce

transcriptional gene silencing throughout the cluster. Evidence

for this model was recently obtained by Latos and colleagues

by generating a series of recombinant endogenous chromo-

somes at the Igf2r/Airn locus in ES cells (Latos et al., 2012)

(Figure 1C). Analogous to XCI, the onset of allele-specific

expression at this locus in the embryo can be recapitulated

by ES cell differentiation, where Igf2r is initially biallelically ex-

pressed but the initiation of Airn expression results in Igf2r

imprinting (Latos et al., 2009). To test whether Airn transcription

or the lncRNA itself was required for Igf2r silencing, Airn was

shortened to different lengths and it was shown that silencing

only required Airn transcription overlap of the Igf2r promoter,

which interferes with RNA polymerase II recruitment (Latos

et al., 2012). This model suggests that Airn acts predominantly

through its transcription, rather than as a lncRNA like Xist.

It is, however, also possible that imprinted lncRNAs act by

coating the local chromosomal region and directly recruiting

repressive chromatin proteins to the imprinted cluster, in a

manner similar to Xist lncRNA. Many imprinted lncRNAs, such

as Gtl2 and Nespas, appear to form a complex with Polycomb

proteins (Pandey et al., 2008; Zhao et al., 2010). Evidence for

a function of the lncRNA in recruitment of histone posttrans-

lational modification machinery comes from experiments in

placental tissues. RNA fluorescence in situ hybridization ex-

periments showed that Airn and Kcnq1ot1 form RNA clouds at

their site of transcription and are associated with a repressive

histone compartment and Polycomb proteins (Nagano et al.,

2008; Pandey et al., 2008; Terranova et al., 2008; Redrup

et al., 2009). This nuclear compartment is also devoid of RNA

polymerase II and exists in a three-dimensionally contracted

state. Other studies on the Airn lncRNA go further in suggesting

that the lncRNAs actively recruit repressive histone modifica-

tions (Nagano et al., 2008) but only in the placenta. In this latter

case, Airn was shown to actively recruit the histone H3 lysine

9 methyltransferase, G9a, and paternal-specific silencing of

the Slc22a3 gene but not the Igf2r gene, was dependent on

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1313

Figure 4. LncRNAs Tether Epigenetic Complexes to Chromatin, Enabling Allelic, and Locus-Specific Regulation(A–C) LncRNA transcribed by RNA polymerase II (POL-II) (A) cotranscriptionally binds to an epigenetic complex (such as PRC2) (B), which loads onto chromatinthrough DNA-bound factors such as YY1 (for Xist RNA) (C).(D) Epigenetic modifications silence the linked gene. Rapid lncRNA turnover prevents diffusion and action at ectopic loci. Adapted from Lee, 2012.

G9a—in a mechanism that contrasts with the promoter-

transcription model hypothesized on the basis of transcript

truncation experiments in somatic lineages (Latos et al., 2012).

These experiments indicate that lncRNA mediated silencing of

imprinted genes may depend on different downstream mecha-

nisms.

A new class of lncRNAs was recently discovered, sno-

lncRNAs, that arise from the imprinted Prader-Willi Syndrome

(PWS) critical region of human chromosome 15 (Yin et al.,

2012). Intriguingly, these lncRNAs, which have a snoRNA

sequence at each end as well as intervening sequence, accumu-

late near the sites of synthesis and strongly associate with Fox

family splicing regulators and alter splicing. The investigators

hypothesize that the sno-lncRNAs in the PWS locus function

as molecular sinks for Fox2, acting locally to alter splicing

patterns in specific subnuclear neighborhoods. Thus, the mech-

anisms by which lncRNAs operate at imprinted loci are diverse.

Why Are lncRNAs Central to Imprinting and XCI?. It has been

argued that lncRNAs make ideal factors for allelic regulation

(Lee, 2012). Indeed, lncRNA’s tethering capabilities and potential

for fast turnover renders them excellent allelic markers. These

transcripts are tethered to the site of synthesis through the

RNA polymerase II transcription complex and can therefore

function as allele-specific tags (Figure 4). As shown by Xist and

RepA RNA, such long transcripts can cotranscriptionally capture

chromatin complexes while tethered to the site of transcription

(Zhao et al., 2008). Tethering could be aided by bridge proteins,

such as YY1 in the case of Xist RNA (Jeon and Lee, 2011). Rapid

RNA turnover after transcription would prevent diffusion to

ectopic sites. At the Xic, the decoying effect of Tsix for Polycomb

proteins would be limited to the Xic by Tsix’s very short half-life

(30–60 min, the time required to synthesize the 40 kb RNA;

Sun et al., 2006) so that effective concentrations would only be

reached at the site of synthesis. Whereas lncRNAs can mark

alleles, proteins cannot retain such allelic memory because

their transcriptional origin is lost when mRNA is shuttled to the

cytoplasm.

Another property of lncRNAs is their ability to specify a unique

address (Lee, 2012). Although transcription factors can also

recruit epigenetic complexes, lncRNAs offer the possibility of

targeting to a single location. Transcription factors typically

target complexes to multiple genomic locations because they

recognize short DNA motifs that occur hundreds to thousands

of times in the genome. In contrast, lncRNAs such as Tsix and

RepA/Xist occur only once in the genome. Because of this

1314 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

uniqueness, lncRNAs can deliver epigenetic complexes to

a single address, offering a regulatory specificity not possible

with proteins or small RNAs. These properties may explain why

the protein-coding region syntenic to the present-day Xic was

rapidly transformed into a noncoding landscape �150 million

years ago when random XCI first appeared in eutherian

mammals (Duret et al., 2006). Prior to this time, Xist was a

ubiquitin ligase, Lnx3, and Jpxwas a peptidase,UspL1. It is likely

that lncRNAs evolved within imprinted domains and other

locations in the mammalian genome for similar reasons. For

a discussion of genome-wide lncRNAswith epigenetic functions,

we refer readers to the accompanying Review by Batista and

Chang on page 1298 of this issue (Batista and Chang, 2013).

Insulators

Despite the common occurrence of lncRNAs at imprinted loci,

insulators may play an equally important role in imprinted

regions. The insulator model, which has been described at the

Igf2/H19 locus (Figure 1A), is an evolutionarily older mechanism,

components of which are conserved in marsupials (Smits et al.,

2008). Key to this mechanism is CTCF-binding sites in the ICR,

which exhibit insulator or enhancer blocking properties (Bell

and Felsenfeld, 2000; Hark et al., 2000). On the maternal allele,

CTCF binds to the ICR and blocks the access of Igf2 to

enhancers shared with H19, which are located downstream,

thereby allowing H19 exclusive enhancer access. On the

paternal allele the ICR acquires DNA methylation in the male

germline, preventing CTCF binding, allowing Igf2 interaction

with the enhancers and paternal-specific expression (Fig-

ure 1A). The presence of DNA methylation on the paternal ICR

leads to secondary methylation of the H19 promoter and

paternal-specific H19 silencing (Thorvaldsen et al., 1998). The

involvement of CTCF in the insulator model has prompted the

identification of CTCF-binding sites at other imprinted genes

such as Rasgrf1, Grb10, and Kcnq1ot1, indicating that the

insulator model may operate in other imprinted clusters. CTCF

sites have also been identified within the Xic in regions important

for imprinted XCI (Chao et al., 2002); however, it is currently

unknown if CTCF is central to imprinting the X. Insulator-based

and lncRNA-based models are not mutually exclusive.

The Enigma of Imprinted XCIImprinted XCI in the Mouse

A further consideration of imprinted XCI is worthwhile for its

mechanistic differences and implications for human develop-

ment. The mechanism of X-imprinting not only differs from

Figure 5. Imprinted XCI in the Mouse(A) Pictorial representation of genic localization into the preformed silent compartment during imprinted XCI. XP repeats form a silent compartment next to thenucleolus by the two-cell stage and, although Xist RNA localizes within it, formation of this silent compartment does not require Xist. The repeats could potentiallycontribute to imprinted XCI by setting up a silencing compartment next to the nucleolus. The silent compartment is present by the two-cell stage and enlarges asgenic loci are translocated into it and become silenced. Genic silencing depends on Xist. XP silencing is completed by the blastocyst.(B) Pictorial representation of XP and XM in the early mouse embryo. Repeat elements of XP create the silent perinucleolar compartment, whereas XM and activegenic loci of XP reside in repeat-expressing regions.(C) Hypothesis: developmental history of the X chromosome from gamete to embryo. Hypothesized events in imprinted XCI of the mouse are shown. In the malegermline, during the first meiotic prophase, the X and Y are inactivated by MSCI and remain suppressed through spermiogenesis as PMSC. This germline-inactivated X may be passed onto the next generation with its repeats preinactivated. In the two-cell mouse embryo, repetitive elements on XP are alreadysuppressed in an Xist-independent manner. XP genic silencing occurs progressively during preimplantation development, strictly depends on Xist, and iscompleted in the blastocyst stage. Thus, imprinted XCI in the mouse embryo is a two-step process, with repeat silencing (Xist-independent) occurring prior togenic silencing (Xist-dependent). Repeat silencing could account partly for the transgenerational information (the imprint) involved in XP silencing. The maternalgermline also plays a crucial role in imprinted XCI by marking the future XM.Adapted from Namekawa et al. (2010).

random XCI but also differs between the imprinted marsupial

and eutherian forms. In mouse imprinted XCI, XP-repeat

silencing precedes genic inactivation (Figure 5A) (Namekawa

et al., 2010). The repeats form a silent compartment next to the

nucleolus by the two-cell stage and, although Xist RNA localizes

within it, formation of this silent compartment does not require

Xist. Repeats could potentially contribute to imprinted XCI

by setting up a silencing compartment next to the nucleolus

(Figure 5B). If their silencing were indeed carried over from the

male germline, repeats could account partly for the transgene-

rational information (the imprint) for XP silencing.

XP genic silencing follows repeat silencing (Namekawa et al.,

2010) and occurs predominantly in the morula-blastocyst stages

(Okamoto and Heard, 2006; Namekawa et al., 2010). Although

one study suggests an Xist-independent process (Kalantry

et al., 2009), the general consensus is that genic silencing

depends on Xist (Marahrens et al., 1997; Namekawa et al.,

2010). Xist must be marked by a second (presently unknown)

imprint that would promote imprinted genic XCI (Figure 5C). In

the mouse, Xist and Tsix are opposing regulatory factors for

imprinted genic silencing, as they are for random XCI. Deleting

Xist from XP precludes placental XCI (Marahrens et al., 1997),

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1315

whereas deleting Tsix from XM compromises maternal-specific

protection from imprinted silencing in the placenta (Lee, 2000;

Sado et al., 2001). Thus, the Xic plays at least a partial role in

imprinted XCI in eutherian mammals.

Imprinted XCI in Marsupials

The eutherian Xic is not recognizable in the marsupial (Duret

et al., 2006). The idea of an Xist-independent mechanism based

on repeat silencing raises the possibility of a similar mechanism

in marsupials. Notably, the opposum male germline demon-

strates postmeiotic silencing of X-linked repeat elements (Name-

kawa et al., 2007), but whether silencing is carried over into the

embryo is unknown. The recent identification of RSX indicates

that a lncRNA like XIST may be present (Grant et al., 2012).

The 27 kb RSX transcript also ‘‘coats’’ the marsupial Xi and

is specifically expressed in female cells. Introduction of RSX

transgenes into mouse ES cells results in partial silencing of

three autosomal genes near the site of integration. These find-

ings suggest that RSX may be the XIST equivalent in opposum,

though an RSX knockout has not been performed and the two

lncRNAs do not possess obvious homology. Like in the mouse,

an XIC mechanism may occur alongside a repeat-silencing

process to implement imprinted XCI.

Imprinted XCI in Humans?

The question of whether imprinted XCI occurs in the human

placenta has not been resolved, but implications for human

development are evident. In several studies, examination of

single X-linked genes from a small number of placentae sug-

gested preferential maternal expression (e.g., Harrison and

Warburton, 1986). Using transdifferentiation of a female human

ES line into trophoblast cells, another study found that FMR1

was expressed only from one X, consistent with imprinting

(Dhara and Benvenisty, 2004). However, other studies have

detected expression from both XM and XP (Moreira de Mello

et al., 2010; Okamoto et al., 2011; Penaherrera et al., 2012);

and, in a nonhuman primate model, XIST was detected from

either XM or XP of the trophectoderm (Tachibana et al., 2012).

The fact that the X chromosome contributing to Turner (XO)

and Klinefelter (XXY) syndrome could be of either XM or XP origin

(Skuse et al., 1997; Skuse, 2000, 2005) further argues against

imprinting. Although the preponderance of evidence may be

against imprinted XCI in human placentae, there is the intriguing

possibility of X-imprinting in the brain as a basis for male-female

differences in behavior and prevalence of autism (Skuse, 2000)

(more below). The question of imprinting therefore bears sig-

nificance for human development and disease, particularly

where X-linked mutations may contribute to early fetal loss,

and congenital or cognitive defects.

Human Diseases and ConditionsCongenital Diseases of Imprinting

Because of parental-origin effects, human disease syndromes

can result from genetic or epigenetic abnormalities on only

a single parental allele. In fact, most well-defined imprinted

gene clusters are associated with human diseases (Thorvaldsen

and Bartolomei, 2007). Interestingly, aberrant expression of

ICR-associated lncRNAs may be implicated in various imprint-

ing disorders. Two of the best-studied imprinting syndromes,

PWS and Angelman (AS) syndromes, map to human chromo-

1316 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

some 15 (Buiting, 2010). PWS involves loss of function of

a number of genes on 15q11-13, including SNORD116. People

with PWS are obese and have reduced muscle tone and mental

ability. AS syndrome is a complex disorder of the nervous

system that arises from loss of function of the UBE3A gene

(Figure 1D). AS symptoms include delayed development, intel-

lectual disability, and severe speech impairment. Most PWS

and AS cases involve large deletions containing the imprinted

genes from the chromosome on which they are expressed. In

PWS, there is biallelic repression of the ICR-associated lncRNA;

in AS, the lncRNA is biallelically expressed. A smaller number

of cases arise from either deletion or aberrant allelic DNA

methylation of the ICR, leading to expression changes. With

the recent identification of the new class of lncRNAs, sno-

lncRNAs, it is likely that absence of the sno-lncRNA in the

PWS critical region impairs brain-specific splicing possibly due

to mislocalization of Fox splicing factors.

Beckwith-Wiedemann syndrome (BWS), an overgrowth

disorder, and Silver-Russell syndrome (SRS), an undergrowth

and asymmetry disorder, are two other well-studied imprinting

disorders that map to human chromosome 11p15.5, where

IGF2 and H19 reside (Figure 1A). Unlike PWS and AS, the

majority of individuals with BWS or SRS have epigenetic errors.

For example, over half of BWS cases exhibit loss of methylation

at the KCNQ1 ICR, which results in biallelic expression of the

KCNQ1OT1 lncRNA (Weksberg et al., 2005) (Figure 1B). Inappro-

priate expression of the lncRNA may lead to aberrant repression

of associated disease genes in cis—in this case, CDKN1C was

silenced. Additionally, some BWS patients exhibit overexpres-

sion of IGF2. Most of these cases have small deletions in the

ICR on the maternal allele, which disrupts the CTCF-dependent

insulator, leading to biallelic IGF2 and loss of H19 expression

(Riccio et al., 2009). Curiously, the remaining ICR sequences in

these individuals are hypermethylated. Many individuals with

SRS have an opposite epigenetic phenotype where the ICR is

unmethylated, resulting in biallelic H19 expression and loss of

IGF2 expression. In many of these cases, it is unclear what event

leads to DNA hypomethylation but in some cases, multiple

imprinted loci exhibit loss of ICR methylation (Azzi et al., 2010).

Significantly, some examples of multilocus loss of imprinting

involve mutations in ZFP57, a zinc finger protein involved in the

postfertilization maintenance of genomic imprints, which was

first reported in individuals presenting with transient neonatal

diabetes (Mackay et al., 2008). It is possible that yet-to-be

identified proteins are mutated in other cases involving loss of

methylation. Alternatively, early environment insults can affect

DNA methylation patterns (see Imprinting and Assisted Repro-

ductive Technology as an example).

X-Linked Influences on Disease, Cognition,

and Behavior

The X chromosome is home to nearly 1,000 genes, many of

which result in discernible human phenotypes when mutated.

X-linked diseases result from single-gene mutations, which can

be classified as dominant or recessive, with the former manifest

in both XX and XY individuals and the latter manifest primarily in

XY individuals because they lack a wild-type allele. X-linked

mutations can cause serious disease, such as hemophilia

A (FVIII), Duchenne muscular dystrophy (DMD), Rett syndrome

(MECP2), and fragile X syndrome (FMR1), or less serious con-

ditions, such as red-green color blindness and male-pattern

baldness. Because of differential inheritance of sex chromo-

somes and the hemizygous state of the X chromosome in the

male population, more diseases have been described for the

X chromosome than any other (Puck and Willard, 1998).

As X-linked genes have existed in the hemizygous state for

much of the history of sex chromosomes, the X chromosome

has been engaged in selection of sexually dimorphic traits for

more than 300 million years since the X and Y began to diverge

(Arnold et al., 2004; Skuse, 2005). Genes for sexual dimorphism,

reproduction, and cognition are enriched on the X chromosome,

with their genetic patency making them easy substrates for

evolutionary selection. In mice, deleting the Xic-encoded

lncRNA, Tsx, has been shown to reduce fear and enhance hippo-

campal short-term memory in male mice (Anguera et al., 2011).

The fact that many X-linked genes are expressed in the brain,

some in a sex-specific manner, may explain why mental retarda-

tion and autism are up to ten times more common in males,

though the underlying mutations are not known for many such

disorders (Skuse et al., 1997). Genetic patency of X-linked

haplotypes has been hypothesized to increase the likelihood of

manifesting extreme behavioral and cognitive phenotypes in

males, and the likelihood would also be increased in females

when the XCI pattern is skewed to favor XM expression. XCI

profiles and mosaicism vary extensively between human

females, perhaps accounting for greater phenotypic variation

among females (Carrel and Willard, 2005). Genes that variably

escape XCI also contribute to this effect (Berletch et al., 2011).

In the mouse, X-linked modifiers such as the Xce can skew XCI

ratios (Cattanach and Isaacson, 1967; Percec et al., 2002;

Thorvaldsen et al., 2012), providing a mechanism by which

nonrandom XCI patterns could be generated. Nonrandom XCI

is also not uncommon in human females (Puck and Willard,

1998).

In the area of cognitive and behavioral development, the study

of X chromosome monosomies (XO, Turner syndrome) has

played a major role in elucidating X-linked contributions. Turner

syndrome girls usually have normal verbal intelligence but are

less developed in spatial and mathematical skills. By comparing

Turner syndrome girls who inherited their X chromosome from

mother (XMO) versus father (XPO), one study concluded that

the XP was associated with enhanced social cognitive function

(Skuse et al., 1997). Despite their genotypic similarity, the epige-

netically different XPO and XMO girls demonstrated measurable

phenotypic differences in social adjustment. The fact that the

XP chromosome is normally only inherited by daughters has

led some to suggest that it accounts for better social skills in

girls on average. XPO and XMO girls also exhibit differences in

visual memory and brain structure (Bishop et al., 2000; Kesler

et al., 2004). Candidate genes include USP9X, MAOA, and

MAOB (monoamine oxidases) on the short arm of the human

X chromosome (Good et al., 2003; Oreland et al., 2004).

Genes on the X chromosome may be imprinted tissue specif-

ically, particularly in the brain where many X-linked genes are

expressed. A transcriptome analysis of the mouse brain sug-

gested that hundreds of alleles on XM may be preferentially

expressed in glutamatergic neurons of the female cortex (Gregg

et al., 2010). Although XP alleles are not silenced, they are ex-

pressed at lower levels. This type of partial imprinting could

contribute to cognitive and behavioral differences. Follow-up

analyses have argued that the allelic skewing called by whole-

transcriptome analyses may have been an aberration caused

by unappreciated statistical limitations of a novel technology

(DeVeale et al., 2012; reviewed in Kelsey and Bartolomei,

2012). Thus, the question of how many and in what tissues im-

printed X-linked genesmay occur in eutherianmammals remains

open. This clinically important area has been underexplored.

Xist, the X Chromosome, and Cancer

An association between the X chromosome and cancer has been

noted since the discovery of the Barr body (Moore and Barr,

1955; Liao et al., 2003; Pageau et al., 2007). Breast and ovarian

cancer cells, for example, frequently duplicate their Xa. The

correlation also holds for men, as XXY men have a 20- to 50-

fold increased risk of breast cancer in a BRCA1 background

(Fentiman et al., 2006), and testicular germ cell tumors often

acquire supernumerary Xs (Kawakami et al., 2003). One recent

study directly tested the connection of the X to cancer by condi-

tionally deleting Xist RNA in the blood lineages of mice (Yildirim

et al., 2013). This deletion resulted in overexpression of the

X chromosome and a fulminant hematologic cancer known as

‘‘mixed MPN/MDS’’ (myeloproliferative neoplasm, myelodys-

plastic syndrome), a cancer that includes chronic myelomono-

cytic leukemia, erythroleukemia, histiocytic sarcoma, and bone

marrow fibrosis. The cancer is female specific and 100% pene-

trant. Intriguingly, in humans, MDS is more common in women,

with noted XIST deletions and X chromosome duplications

occurring in MPN, MDS, and myeloid cancers (see references

within Yildirim et al., 2013). The association is not restricted

to women, as extra X chromosomes are seen in a range of

leukemias in both sexes. The mouse study showed that loss

of Xist perturbed maturation as well as longevity of hematopoi-

etic stem cells. Thus, Xist plays a role not only in dosage

compensation but also in suppressing cancer and preserving

function of adult stem cell populations. This study illustrates

the importance of studying lncRNA function not only in cells

ex vivo but also within the context of the organism in vivo.

Epigenetic Reprogramming in Human Stem Cells

Xist RNA also influences the pluripotent stem cell population, as

shown by recent studies of induced pluripotent stem cells (iPSC)

in regenerative medicine. In mice, XCI is tightly linked to cell

differentiation in the epiblast and the possession of two Xa is

a hallmark of pluripotent cells of both mouse ESC and iPSC

(reviewed in Minkovsky et al., 2012). The tight linkage is ex-

plained by the physical convergence of many pluripotency

factors, such as OCT4, SOX2, NANOG, and REX1, at the Xic,

specifically within control regions of Xite, Tsix, and Xist (Navarro

et al., 2008; Donohoe et al., 2009; Navarro et al., 2010)

(Figure 3B). Binding of pluripotency factors to these regions

blocks initiation of XCI, and the loss of binding during cell differ-

entiation creates a permissive state for the initiation of XCI.

XIST currently provides one of a few tangible readouts for

stem cell quality. In human ESC (hESC) and iPSC (hiPSC),

XIST expression and XCI do not necessarily occur in the

expected manner. Female hESC and hiPSC lines occur in

three different epigenetic groups based on XIST expression

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1317

(Silva et al., 2008) (Minkovsky et al., 2012). ‘‘Class I’’ cells are

most similar to mESC in that they have two Xa in the undifferen-

tiated state. When placed in differentiation conditions, Class I

cells express XIST and initiate XCI. In contrast, ‘‘Class II’’ cells

already express XIST and carry one Xi, even before growth under

differentiation conditions. Finally, ‘‘Class III’’ cells once ex-

pressed XIST but irreversibly lost its expression, with evidence

of partial X reactivation (Shen et al., 2008; Silva et al., 2008;

Anguera et al., 2012; Tomoda et al., 2012). Epigenetic fluidity is

evident through irreversible progression from Class I to II to III

states (Tchieu et al., 2010; Anguera et al., 2012; Mekhoubad

et al., 2012). Class I is transient, whereas Class III is dominant

and stable.

Although failure of XIST expression is lethal in vivo (Penny

et al., 1996; Marahrens et al., 1997), loss of XIST does not

have the same dire consequences ex vivo, though these cells

lack full developmental potential (Silva et al., 2008; Anguera

et al., 2012). Class III hiPSCs have limited differentiation

capability (Anguera et al., 2012; Mekhoubad et al., 2012), In a

xenograft model, Class III hiPSCs produce cystic teratomas

composed of simple cystic epithelia and undifferentiatedmesen-

chyme, whereas Class II cells produce well-differentiated struc-

tures of three germ layers. Given the tumorigenic phenotype of

the murine Xist deletion (Yildirim et al., 2013), most concerning

would be the potential of XIST-negative hIPSC lines to cause

cancer when introduced in vivo in the clinical setting. Indeed,

Class III hiPSC also showed partial X reactivation, faster

doubling times, and a distinct gene expression signature of

cancer cells (Anguera et al., 2012), urging further careful consid-

eration before using hiPSCs in regenerative medicine.

Genomic imprinting also contributes to quality of human and

mouse iPSC (Pick et al., 2009; Sun et al., 2012). The imprinted

state of the imprinted Dlk1-Dio3 locus—in particular the expres-

sion of Gtl2 (aka Meg3) lncRNA—has been at the center of

attention. One study found that mouse iPSC cloneswith aberrant

Dlk1-Dio3 imprinting and low Gtl2 expression contributed

poorly to chimeras (Stadtfeld et al., 2010), whereas another did

not observe a difference (Carey et al., 2011). There is, however,

general agreement that loss of imprinting at this locus resulted

in lower efficiency of generating entirely iPSC-derived mice

(Stadtfeld et al., 2010; Carey et al., 2011; Stadtfeld et al.,

2012). With further investigation, it is likely that other imprinted

loci will affect stem cell quality. Nonetheless, despite a number

of claims that imprinting is aberrant, iPSCs can be an important

tool for studying imprinting perturbations in inaccessible cell

types such as neurons in AS (Chamberlain et al., 2010).

Imprinting and Assisted Reproductive TechnologyRelated to the issue of epigenetic change within imprinted and

X-linked loci in stem cells ex vivo is the question of whether

in vitro culture of early human embryos during use of ART might

have similar effects on imprinting and XCI. The ex vivo manipu-

lations utilized during ART coincide with the developmental

stages in which genome-wide epigenetic reprogramming occurs

(i.e., oocyte growth and preimplantation development). The use

of ART procedures to help couples with fertility issues conceive

children of their own has doubled in the last decade. In 2009,

ART contributed to 1.4% of all U.S. births (Sunderam et al.,

1318 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

2012). Nevertheless, there is growing concern about the safety

of these procedures (Manipalviratn et al., 2009). Of particular

concern, children conceived by ART have an increased inci-

dence of rare epigenetic disorders, with most of these patients

exhibiting loss of DNA methylation at ICRs (Manipalviratn et al.,

2009). Specifically, cases of AS and BWS in children conceived

by ART are associated with loss of methylation of the SNPRN

and KCNQ1 ICRs, respectively, which result in biallelic expres-

sion of the lncRNAs and loss of expression of UBE3A and

CDKN1C, respectively (Figure 1). Consistently, animal studies

have demonstrated that embryo culture and embryo transfer

as well as hormonal treatments, which are integral components

of ART, disrupt normal epigenetic programming in embryonic

and extraembryonic lineages (Mann et al., 2004; Fortier et al.,

2008; Rivera et al., 2008), although the mechanism for this

disruption remains poorly understood. Thus, a greater under-

standing of in vitro effects on epigenetic regulation during ART

is a rising need from a public health perspective of industrialized

countries.

Conclusions and Therapeutic ProspectsXCI, genomic imprinting, and lncRNA clearly havemajor implica-

tions for public health. Yet, in the arena of preventive, diagnostic,

and therapeutic medicine, few strategies have targeted regula-

tory factors for imprinted genes and the Xic to control X-linked

disease and conditions. This holds true also for regenerative

medicine and stem cell biology, where ex vivo cellular manipula-

tions have not universally considered the impact of imprinting

and XCI, in spite of converging indications that these processes

impact production, maintenance, and overall quality of stem

cells.

On a hopeful note, proof-of-concept was reported in one

recent study. One of the most intriguing aspects of disorders

that involve monoallelically expressed genes is the prospect

for therapy that involves derepressing the silenced allele

in situations where the expressed allele of an imprinted gene is

deleted or contains a loss of function mutation. A recent success

was reported for AS, where a screen revealed that small mole-

cule topoisomerase inhibitors reactivated the silenced UBE3A

gene and repressed the ICR-associated antisense RNA (Huang

et al., 2012). Ironically, however, because of the clustering of

imprinted genes, the biallelic activation ofUBE3Awas accompa-

nied by the loss of expression of the paternally-expressed genes

in the locus (Figure 1D). Although the mechanism for reactivation

is unclear, these strategies offer hope and suggest that other loci

could be subject to similar screens.

The successful reactivation of the silent copy of UBE3A raises

hopes that a treatment for various X-linked diseases might be

similarly achieved. Of particular interest has been Rett

syndrome, a neurologic disorder caused by mutations in

MECP2. The syndrome affects girls and is manifested by

a reversal of developmental milestones after the first year of

life (the disease is fatal in newborn males). Because Rett

syndrome is not accompanied by neurodegeneration, efforts

have been devoted to restoring expression of MECP2 after birth

in hopes of reversing the symptoms. Intriguingly, mouse models

have shown that restoration of MECP2 expression after disease

first becomes symptomatic can reverse the neurologic defects

(Giacometti et al., 2007; Guy et al., 2007). Because the possibility

that restoration of MECP2 expression might similarly cure

Rett syndrome in humans, ongoing studies are now aimed at

reactivating the wild-type copy of MECP2 in affected girls

through small molecules that target chromatin modifiers and

other regulators of XCI and XCR.

Furthermore, with knowledge that loss of XIST expression

and/or overdosage of the X chromosome could result in blood

cancer (Yildirim et al., 2013), cancer therapeutics might similarly

be directed at genes on the X chromosome. In the future, in addi-

tion to trans-acting factors such as topoisomerases, therapeutic

strategies could be targeted at control regions (ICRs, Xic) or

lncRNAs that regulate crucial genes in cis. For example, it may

be productive to determine ways to control expression of XIST

RNA, GTL2, and other imprinted genes. We may be some

ways from realizing commercial products, but the technologies

to develop them are evolving rapidly and may soon enable us

to produce drugs to influence cellular reprogramming ex vivo

and to treat human diseases and conditions in vivo.

ACKNOWLEDGMENTS

We thank the Lee and Bartolomei labs for many inspirational discussions, and

the National Institutes of Health and the Howard Hughes Medical Institute for

supporting their research.

REFERENCES

Anguera, M.C., Ma, W., Clift, D., Namekawa, S.H., Kelleher, R.J., 3rd, and Lee,

J.T. (2011). Tsx produces a long noncoding RNA and has general functions in

the germline, stem cells, and brain. PLoS Genet. 7, e1002248.

Anguera, M.C., Sadreyev, R.I., Zhang, Z., Szanto, A., Payer, B., Sheridan, S.D.,

Kwok, S., Haggarty, S.J., Sur, M., Alvarez, J., et al. (2012). Molecular signa-

tures of human induced pluripotent stem cells highlight sex differences and

cancer genes. Cell Stem Cell 11, 75–90.

Arnold, A.P., Xu, J., Grisham, W., Chen, X., Kim, Y.H., and Itoh, Y. (2004). Mini-

review: Sex chromosomes and brain sexual differentiation. Endocrinology

145, 1057–1062.

Azzi, S., Rossignol, S., Le Bouc, Y., and Netchine, I. (2010). Lessons from

imprinted multilocus loss of methylation in human syndromes: A step toward

understanding the mechanisms underlying these complex diseases. Epige-

netics 5, 373–377.

Bacher, C.P., Guggiari, M., Brors, B., Augui, S., Clerc, P., Avner, P., Eils, R.,

and Heard, E. (2006). Transient colocalization of X-inactivation centres

accompanies the initiation of X inactivation. Nat. Cell Biol. 8, 293–299.

Barakat, T.S., Gunhanlar, N., Pardo, C.G., Achame, E.M., Ghazvini, M., Boers,

R., Kenter, A., Rentmeester, E., Grootegoed, J.A., and Gribnau, J. (2011).

RNF12 activates Xist and is essential for X chromosome inactivation. PLoS

Genet. 7, e1002001.

Barlow, D.P. (2011). Genomic imprinting: a mammalian epigenetic discovery

model. Annu. Rev. Genet. 45, 379–403.

Barr, M.L., and Bertram, E.G. (1949). A morphological distinction between

neurones of the male and female, and the behaviour of the nucleolar satellite

during accelerated nucleoprotein synthesis. Nature 163, 676.

Bartolomei, M.S. (2009). Genomic imprinting: employing and avoiding epige-

netic processes. Genes Dev. 23, 2124–2133.

Bartolomei, M.S., and Ferguson-Smith, A.C. (2011). Mammalian genomic

imprinting. Cold Spring Harb. Perspect. Biol. 3. http://dx.doi.org/10.1101/

cshperspect.a002592.

Batista, P.J., and Chang, H.Y. (2013). Long noncoding RNAs: Cellular address

codes in development and disease. Cell 152, this issue, 1298–1307.

Bell, A.C., and Felsenfeld, G. (2000). Methylation of a CTCF-dependent

boundary controls imprinted expression of the Igf2 gene. Nature 405, 482–485.

Berletch, J.B., Yang, F., Xu, J., Carrel, L., and Disteche, C.M. (2011). Genes

that escape from X inactivation. Hum. Genet. 130, 237–245.

Bishop, D.V., Canning, E., Elgar, K., Morris, E., Jacobs, P.A., and Skuse, D.H.

(2000). Distinctive patterns of memory function in subgroups of females with

Turner syndrome: evidence for imprinted loci on the X-chromosome affecting

neurodevelopment. Neuropsychologia 38, 712–721.

Borsani, G., Tonlorenzi, R., Simmler, M.C., Dandolo, L., Arnaud, D., Capra, V.,

Grompe, M., Pizzuti, A., Muzny, D., Lawrence, C., et al. (1991). Characteriza-

tion of a murine gene expressed from the inactive X chromosome. Nature

351, 325–329.

Brannan, C.I., Dees, E.C., Ingram, R.S., and Tilghman, S.M. (1990). The

product of the H19 gene may function as an RNA. Mol. Cell. Biol. 10, 28–36.

Brockdorff, N., Ashworth, A., Kay, G.F., McCabe, V.M., Norris, D.P., Cooper,

P.J., Swift, S., and Rastan, S. (1992). The product of the mouse Xist gene is

a 15 kb inactive X-specific transcript containing no conserved ORF and

located in the nucleus. Cell 71, 515–526.

Brown, C.J., Ballabio, A., Rupert, J.L., Lafreniere, R.G., Grompe, M.,

Tonlorenzi, R., and Willard, H.F. (1991a). A gene from the region of the

human X inactivation centre is expressed exclusively from the inactive

X chromosome. Nature 349, 38–44.

Brown, C.J., Lafreniere, R.G., Powers, V.E., Sebastio, G., Ballabio, A.,

Pettigrew, A.L., Ledbetter, D.H., Levy, E., Craig, I.W., and Willard, H.F.

(1991b). Localization of the X inactivation centre on the human X chromosome

in Xq13. Nature 349, 82–84.

Brown, C.J., Hendrich, B.D., Rupert, J.L., Lafreniere, R.G., Xing, Y., Lawrence,

J., and Willard, H.F. (1992). The human XIST gene: analysis of a 17 kb inactive

X-specific RNA that contains conserved repeats and is highly localized within

the nucleus. Cell 71, 527–542.

Buiting, K. (2010). Prader-Willi syndrome and Angelman syndrome. Am. J.

Med. Genet. C. Semin. Med. Genet. 154C, 365–376.

Carey, B.W., Markoulaki, S., Hanna, J.H., Faddah, D.A., Buganim, Y., Kim, J.,

Ganz, K., Steine, E.J., Cassady, J.P., Creyghton, M.P., et al. (2011). Reprog-

ramming factor stoichiometry influences the epigenetic state and biological

properties of induced pluripotent stem cells. Cell Stem Cell 9, 588–598.

Carrel, L., and Willard, H.F. (2005). X-inactivation profile reveals extensive

variability in X-linked gene expression in females. Nature 434, 400–404.

Cattanach, B.M. (1986). Parental origin effects in mice. J. Embryol. Exp.

Morphol. 97 (Suppl), 137–150.

Cattanach, B.M., and Isaacson, J.H. (1967). Controlling elements in the mouse

X chromosome. Genetics 57, 331–346.

Chamberlain, S.J., Chen, P.F., Ng, K.Y., Bourgois-Rocha, F., Lemtiri-Chlieh,

F., Levine, E.S., and Lalande, M. (2010). Induced pluripotent stem cell models

of the genomic imprinting disorders Angelman and Prader-Willi syndromes.

Proc. Natl. Acad. Sci. USA 107, 17668–17673.

Chao, W., Huynh, K.D., Spencer, R.J., Davidow, L.S., and Lee, J.T. (2002).

CTCF, a candidate trans-acting factor for X-inactivation choice. Science

295, 345–347.

Chotalia, M., Smallwood, S.A., Ruf, N., Dawson, C., Lucifero, D., Frontera, M.,

James, K., Dean,W., and Kelsey, G. (2009). Transcription is required for estab-

lishment of germline methylation marks at imprinted genes. Genes Dev. 23,

105–117.

Chow, J.C., Ciaudo, C., Fazzari, M.J., Mise, N., Servant, N., Glass, J.L.,

Attreed, M., Avner, P., Wutz, A., Barillot, E., et al. (2010). LINE-1 activity in

facultative heterochromatin formation during X chromosome inactivation.

Cell 141, 956–969.

Chureau, C., Prissette, M., Bourdet, A., Barbe, V., Cattolico, L., Jones, L.,

Eggen, A., Avner, P., and Duret, L. (2002). Comparative sequence analysis

of the X-inactivation center region in mouse, human, and bovine. Genome

Res. 12, 894–908.

Chureau, C., Chantalat, S., Romito, A., Galvani, A., Duret, L., Avner, P., and

Rougeulle, C. (2011). Ftx is a non-coding RNA which affects Xist expression

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1319

and chromatin structure within the X-inactivation center region. Hum. Mol.

Genet. 20, 705–718.

Clemson, C.M., McNeil, J.A., Willard, H.F., and Lawrence, J.B. (1996).

XIST RNA paints the inactive X chromosome at interphase: evidence for

a novel RNA involved in nuclear/chromosome structure. J. Cell Biol. 132,

259–275.

Cooper, D.W. (1971). Directed genetic change model for X chromosome inac-

tivation in eutherian mammals. Nature 230, 292–294.

DeVeale, B., van der Kooy, D., and Babak, T. (2012). Critical evaluation of

imprinted gene expression by RNA-Seq: a new perspective. PLoS Genet. 8,

e1002600.

Dhara, S.K., and Benvenisty, N. (2004). Gene trap as a tool for genome

annotation and analysis of X chromosome inactivation in human embryonic

stem cells. Nucleic Acids Res. 32, 3995–4002.

Donohoe, M.E., Silva, S.S., Pinter, S.F., Xu, N., and Lee, J.T. (2009). The

pluripotency factor Oct4 interacts with Ctcf and also controls X-chromosome

pairing and counting. Nature 460, 128–132.

Duret, L., Chureau, C., Samain, S., Weissenbach, J., and Avner, P. (2006). The

Xist RNA gene evolved in eutherians by pseudogenization of a protein-coding

gene. Science 312, 1653–1655.

Fentiman, I.S., Fourquet, A., and Hortobagyi, G.N. (2006). Male breast cancer.

Lancet 367, 595–604.

Fortier, A.L., Lopes, F.L., Darricarrere, N., Martel, J., and Trasler, J.M. (2008).

Superovulation alters the expression of imprinted genes in the midgestation

mouse placenta. Hum. Mol. Genet. 17, 1653–1665.

Frost, J.M., and Moore, G.E. (2010). The importance of imprinting in the

human placenta. PLoS Genet. 6, e1001015.

Giacometti, E., Luikenhuis, S., Beard, C., and Jaenisch, R. (2007). Partial

rescue of MeCP2 deficiency by postnatal activation of MeCP2. Proc. Natl.

Acad. Sci. USA 104, 1931–1936.

Gontan, C., Achame, E.M., Demmers, J., Barakat, T.S., Rentmeester, E., van

IJcken, W., Grootegoed, J.A., and Gribnau, J. (2012). RNF12 initiates

X-chromosome inactivation by targeting REX1 for degradation. Nature 485,

386–390.

Good, C.D., Lawrence, K., Thomas, N.S., Price, C.J., Ashburner, J., Friston,

K.J., Frackowiak, R.S., Oreland, L., and Skuse, D.H. (2003). Dosage-sensitive

X-linked locus influences the development of amygdala and orbitofrontal

cortex, and fear recognition in humans. Brain 126, 2431–2446.

Goto, Y., and Takagi, N. (2000). Maternally inherited X chromosome is not

inactivated in mouse blastocysts due to parental imprinting. Chromosome

Res. 8, 101–109.

Grant, J., Mahadevaiah, S.K., Khil, P., Sangrithi, M.N., Royo, H., Duckworth, J.,

McCarrey, J.R., VandeBerg, J.L., Renfree, M.B., Taylor, W., et al. (2012). Rsx is

a metatherian RNA with Xist-like properties in X-chromosome inactivation.

Nature 487, 254–258.

Greaves, I.K., Rangasamy, D., Devoy, M., Marshall Graves, J.A., and

Tremethick, D.J. (2006). The X and Y chromosomes assemble into H2A.Z-

containing facultative heterochromatin following meiosis. Mol. Cell. Biol. 26,

5394–5405.

Gregg, C., Zhang, J., Butler, J.E., Haig, D., and Dulac, C. (2010). Sex-specific

parent-of-origin allelic expression in the mouse brain. Science 329, 682–685.

Guy, J., Gan, J., Selfridge, J., Cobb, S., and Bird, A. (2007). Reversal of neuro-

logical defects in a mouse model of Rett syndrome. Science 315, 1143–1147.

Hark, A.T., Schoenherr, C.J., Katz, D.J., Ingram, R.S., Levorse, J.M., and

Tilghman, S.M. (2000). CTCF mediates methylation-sensitive enhancer-

blocking activity at the H19/Igf2 locus. Nature 405, 486–489.

Harrison, K.B., and Warburton, D. (1986). Preferential X-chromosome activity

in human female placental tissues. Cytogenet. Cell Genet. 41, 163–168.

Henckel, A., Chebli, K., Kota, S.K., Arnaud, P., and Feil, R. (2012). Transcription

and histone methylation changes correlate with imprint acquisition in male

germ cells. EMBO J. 31, 606–615.

1320 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Hoki, Y., Kimura, N., Kanbayashi, M., Amakawa, Y., Ohhata, T., Sasaki, H.,

and Sado, T. (2009). A proximal conserved repeat in the Xist gene is essential

as a genomic element for X-inactivation in mouse. Development 136,

139–146.

Huang, H.S., Allen, J.A., Mabb, A.M., King, I.F., Miriyala, J., Taylor-Blake, B.,

Sciaky, N., Dutton, J.W., Jr., Lee, H.M., Chen, X., et al. (2012). Topoisomerase

inhibitors unsilence the dormant allele of Ube3a in neurons. Nature 481,

185–189.

Huynh, K.D., and Lee, J.T. (2003). Inheritance of a pre-inactivated paternal X

chromosome in early mouse embryos. Nature 426, 857–862.

Jeon, Y., and Lee, J.T. (2011). YY1 tethers Xist RNA to the inactive X nucleation

center. Cell 146, 119–133.

Jonkers, I., Barakat, T.S., Achame, E.M., Monkhorst, K., Kenter, A., Rent-

meester, E., Grosveld, F., Grootegoed, J.A., and Gribnau, J. (2009). RNF12

is an X-Encoded dose-dependent activator of X chromosome inactivation.

Cell 139, 999–1011.

Kalantry, S., Purushothaman, S., Bowen, R.B., Starmer, J., and Magnuson, T.

(2009). Evidence of Xist RNA-independent initiation of mouse imprinted

X-chromosome inactivation. Nature 460, 647–651.

Kawakami, T., Okamoto, K., Sugihara, H., Hattori, T., Reeve, A.E., Ogawa, O.,

and Okada, Y. (2003). The roles of supernumerical X chromosomes and XIST

expression in testicular germ cell tumors. J. Urol. 169, 1546–1552.

Kelsey, G., and Bartolomei, M.S. (2012). Imprinted genes . and the number

is? PLoS Genet. 8, e1002601.

Kesler, S.R., Garrett, A., Bender, B., Yankowitz, J., Zeng, S.M., and Reiss, A.L.

(2004). Amygdala and hippocampal volumes in Turner syndrome: a high-reso-

lution MRI study of X-monosomy. Neuropsychologia 42, 1971–1978.

Koerner, M.V., Pauler, F.M., Huang, R., and Barlow, D.P. (2009). The function

of non-coding RNAs in genomic imprinting. Development 136, 1771–1783.

Latos, P.A., Stricker, S.H., Steenpass, L., Pauler, F.M., Huang, R., Senergin,

B.H., Regha, K., Koerner, M.V., Warczok, K.E., Unger, C., and Barlow, D.P.

(2009). An in vitro ES cell imprinting model shows that imprinted expression

of the Igf2r gene arises from an allele-specific expression bias. Development

136, 437–448.

Latos, P.A., Pauler, F.M., Koerner, M.V., Sxenergin, H.B., Hudson, Q.J.,

Stocsits, R.R., Allhoff, W., Stricker, S.H., Klement, R.M., Warczok, K.E.,

et al. (2012). Airn transcriptional overlap, but not its lncRNA products, induces

imprinted Igf2r silencing. Science 338, 1469–1472.

Lee, J.T. (2000). Disruption of imprinted X inactivation by parent-of-origin

effects at Tsix. Cell 103, 17–27.

Lee, J.T. (2012). Epigenetic regulation by long noncoding RNAs. Science 338,

1435–1439.

Lee, J.T., Strauss, W.M., Dausman, J.A., and Jaenisch, R. (1996). A 450 kb

transgene displays properties of the mammalian X-inactivation center. Cell

86, 83–94.

Lee, J.T., Davidow, L.S., and Warshawsky, D. (1999a). Tsix, a gene antisense

to Xist at the X-inactivation centre. Nat. Genet. 21, 400–404.

Lee, J.T., Lu, N., and Han, Y. (1999b). Genetic analysis of the mouse

X inactivation center defines an 80-kb multifunction domain. Proc. Natl.

Acad. Sci. USA 96, 3836–3841.

Liao, D.J., Du, Q.Q., Yu, B.W., Grignon, D., and Sarkar, F.H. (2003). Novel

perspective: focusing on the X chromosome in reproductive cancers. Cancer

Invest. 21, 641–658.

Lifschytz, E., and Lindsley, D.L. (1972). The role of X-chromosome inacti-

vation during spermatogenesis (Drosophila-allocycly-chromosome evolu-

tion-male sterility-dosage compensation). Proc. Natl. Acad. Sci. USA 69,

182–186.

Lyle, R., Watanabe, D., te Vruchte, D., Lerchner, W., Smrzka, O.W., Wutz, A.,

Schageman, J., Hahner, L., Davies, C., and Barlow, D.P. (2000). The imprinted

antisense RNA at the Igf2r locus overlaps but does not imprint Mas1. Nat.

Genet. 25, 19–21.

Lyon, M.F. (1961). Gene action in the X-chromosome of the mouse

(Mus musculus L.). Nature 190, 372–373.

Lyon, M.F. (1999). Imprinting and X chromosome inactivation. In Results and

problems in cell differentiation, R. Ohlsson, ed. (Heidelberg: Springer-Verlag),

pp. 73–90.

Lyon,M.F. (2003). The Lyon and the LINE hypothesis. Semin. Cell Dev. Biol. 14,

313–318.

Mackay, D.J., Callaway, J.L., Marks, S.M., White, H.E., Acerini, C.L., Boonen,

S.E., Dayanikli, P., Firth, H.V., Goodship, J.A., Haemers, A.P., et al. (2008).

Hypomethylation of multiple imprinted loci in individuals with transient

neonatal diabetes is associated with mutations in ZFP57. Nat. Genet. 40,

949–951.

Mak, W., Nesterova, T.B., de Napoles, M., Appanah, R., Yamanaka, S., Otte,

A.P., and Brockdorff, N. (2004). Reactivation of the paternal X chromosome in

early mouse embryos. Science 303, 666–669.

Mancini-Dinardo, D., Steele, S.J., Levorse, J.M., Ingram, R.S., and Tilghman,

S.M. (2006). Elongation of the Kcnq1ot1 transcript is required for genomic

imprinting of neighboring genes. Genes Dev. 20, 1268–1282.

Manipalviratn, S., DeCherney, A., and Segars, J. (2009). Imprinting disorders

and assisted reproductive technology. Fertil. Steril. 91, 305–315.

Mann, M.R., Lee, S.S., Doherty, A.S., Verona, R.I., Nolen, L.D., Schultz, R.M.,

and Bartolomei, M.S. (2004). Selective loss of imprinting in the placenta

following preimplantation development in culture. Development 131, 3727–

3735.

Marahrens, Y., Panning, B., Dausman, J., Strauss, W., and Jaenisch, R. (1997).

Xist-deficient mice are defective in dosage compensation but not spermato-

genesis. Genes Dev. 11, 156–166.

McGrath, J., and Solter, D. (1984). Completion of mouse embryogenesis

requires both the maternal and paternal genomes. Cell 37, 179–183.

Mekhoubad, S., Bock, C., de Boer, A.S., Kiskinis, E., Meissner, A., and Eggan,

K. (2012). Erosion of dosage compensation impacts human iPSC disease

modeling. Cell Stem Cell 10, 595–609.

Meng, L., Person, R.E., and Beaudet, A.L. (2012). Ube3a-ATS is an atypical

RNA polymerase II transcript that represses the paternal expression of

Ube3a. Hum. Mol. Genet. 21, 3001–3012.

Migeon, B.R., Kazi, E., Haisley-Royster, C., Hu, J., Reeves, R., Call, L., Lawler,

A., Moore, C.S., Morrison, H., and Jeppesen, P. (1999). Human X inactivation

center induces random X chromosome inactivation in male transgenic mice.

Genomics 59, 113–121.

Minkovsky, A., Patel, S., and Plath, K. (2012). Concise review: Pluripotency

and the transcriptional inactivation of the female Mammalian X chromosome.

Stem Cells 30, 48–54.

Moore, K.L., and Barr, M.L. (1955). The sex chromatin in benign tumours and

related conditions in man. Br. J. Cancer 9, 246–252.

Moreira de Mello, J.C., de Araujo, E.S.S., Stabellini, R., Fraga, A.M., de Souza,

J.E.S., Sumita, D.R., Camargo, A.A., and Pereira, L.V. (2010). Random

X inactivation and extensive mosaicism in human placenta revealed by

analysis of allele-specific gene expression along the X chromosome. PLoS

ONE 5, e10947.

Nagano, T., Mitchell, J.A., Sanz, L.A., Pauler, F.M., Ferguson-Smith, A.C., Feil,

R., and Fraser, P. (2008). The Air noncoding RNA epigenetically silences

transcription by targeting G9a to chromatin. Science 322, 1717–1720.

Namekawa, S.H., Park, P.J., Zhang, L.F., Shima, J.E., McCarrey, J.R.,

Griswold, M.D., and Lee, J.T. (2006). Postmeiotic sex chromatin in the male

germline of mice. Curr. Biol. 16, 660–667.

Namekawa, S.H., VandeBerg, J.L., McCarrey, J.R., and Lee, J.T. (2007). Sex

chromosome silencing in the marsupial male germ line. Proc. Natl. Acad.

Sci. USA 104, 9730–9735.

Namekawa, S.H., Payer, B., Huynh, K.D., Jaenisch, R., and Lee, J.T. (2010).

Two-step imprinted X inactivation: repeat versus genic silencing in the mouse.

Mol. Cell. Biol. 30, 3187–3205.

Navarro, P., Chambers, I., Karwacki-Neisius, V., Chureau, C., Morey, C.,

Rougeulle, C., and Avner, P. (2008). Molecular coupling of Xist regulation

and pluripotency. Science 321, 1693–1695.

Navarro, P., Oldfield, A., Legoupi, J., Festuccia, N., Dubois, A., Attia, M.,

Schoorlemmer, J., Rougeulle, C., Chambers, I., and Avner, P. (2010).

Molecular coupling of Tsix regulation and pluripotency. Nature 468, 457–460.

Ogawa, Y., Sun, B.K., and Lee, J.T. (2008). Intersection of the RNA interference

and X-inactivation pathways. Science 320, 1336–1341.

Ohno, S., Kaplan, W.D., and Kinosita, R. (1959). Formation of the sex chro-

matin by a single X-chromosome in liver cells of Rattus norvegicus. Exp. Cell

Res. 18, 415–418.

Okamoto, I., and Heard, E. (2006). The dynamics of imprinted X inactivation

during preimplantation development in mice. Cytogenet. Genome Res. 113,

318–324.

Okamoto, I., Patrat, C., Thepot, D., Peynot, N., Fauque, P., Daniel, N., Diaban-

gouaya, P., Wolf, J.P., Renard, J.P., Duranthon, V., and Heard, E. (2011).

Eutherian mammals use diverse strategies to initiate X-chromosome inactiva-

tion during development. Nature 472, 370–374.

Oreland, L., Hallman, J., and Damberg, M. (2004). Platelet MAO and person-

ality—function and dysfunction. Curr. Med. Chem. 11, 2007–2016.

Pageau, G.J., Hall, L.L., Ganesan, S., Livingston, D.M., and Lawrence, J.B.

(2007). The disappearing Barr body in breast and ovarian cancers. Nat. Rev.

Cancer 7, 628–633.

Pandey, R.R.,Mondal, T., Mohammad, F., Enroth, S., Redrup, L., Komorowski,

J., Nagano, T., Mancini-Dinardo, D., and Kanduri, C. (2008). Kcnq1ot1 anti-

sense noncoding RNA mediates lineage-specific transcriptional silencing

through chromatin-level regulation. Mol. Cell 32, 232–246.

Pauler, F.M., Barlow, D.P., and Hudson, Q.J. (2012). Mechanisms of long

range silencing by imprinted macro non-coding RNAs. Curr. Opin. Genet.

Dev. 22, 283–289.

Payer, B., and Lee, J.T. (2008). X chromosome dosage compensation: how

mammals keep the balance. Annu. Rev. Genet. 42, 733–772.

Penaherrera, M.S., Jiang, R., Avila, L., Yuen, R.K.C., Brown, C.J., and

Robinson, W.P. (2012). Patterns of placental development evaluated by

X chromosome inactivation profiling provide a basis to evaluate the origin of

epigenetic variation. Hum. Reprod. 27, 1745–1753.

Penny, G.D., Kay, G.F., Sheardown, S.A., Rastan, S., and Brockdorff, N.

(1996). Requirement for Xist in X chromosome inactivation. Nature 379,

131–137.

Percec, I., Plenge, R.M., Nadeau, J.H., Bartolomei, M.S., and Willard, H.F.

(2002). Autosomal dominant mutations affecting X inactivation choice in the

mouse. Science 296, 1136–1139.

Pick, M., Stelzer, Y., Bar-Nur, O., Mayshar, Y., Eden, A., and Benvenisty, N.

(2009). Clone- and gene-specific aberrations of parental imprinting in human

induced pluripotent stem cells. Stem Cells 27, 2686–2690.

Pinter, S.F., Sadreyev, R.I., Yildirim, E., Jeon, Y., Ohsumi, T.K., Borowsky, M.,

and Lee, J.T. (2012). Spreading of X chromosome inactivation via a hierarchy

of defined Polycomb stations. Genome Res. 22, 1864–1876.

Puck, J.M., and Willard, H.F. (1998). X inactivation in females with X-linked

disease. N. Engl. J. Med. 338, 325–328.

Puck, J.M., Stewart, C.C., and Nussbaum, R.L. (1992). Maximum-likelihood

analysis of human T-cell X chromosome inactivation patterns: normal women

versus carriers of X-linked severe combined immunodeficiency. Am. J. Hum.

Genet. 50, 742–748.

Redrup, L., Branco, M.R., Perdeaux, E.R., Krueger, C., Lewis, A., Santos, F.,

Nagano, T., Cobb, B.S., Fraser, P., and Reik, W. (2009). The long noncoding

RNA Kcnq1ot1 organises a lineage-specific nuclear domain for epigenetic

gene silencing. Development 136, 525–530.

Riccio, A., Sparago, A., Verde, G., De Crescenzo, A., Citro, V., Cubellis, M.V.,

Ferrero, G.B., Silengo, M.C., Russo, S., Larizza, L., and Cerrato, F. (2009).

Inherited and Sporadic Epimutations at the IGF2-H19 locus in Beckwith-

Wiedemann syndrome and Wilms’ tumor. Endocr. Dev. 14, 1–9.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1321

Rivera, R.M., Stein, P., Weaver, J.R., Mager, J., Schultz, R.M., and Bartolomei,

M.S. (2008). Manipulations of mouse embryos prior to implantation result in

aberrant expression of imprinted genes on day 9.5 of development. Hum.

Mol. Genet. 17, 1–14.

Robson, J.E., Eaton, S.A., Underhill, P., Williams, D., and Peters, J. (2012).

MicroRNAs 296 and 298 are imprinted and part of the GNAS/Gnas cluster

and miR-296 targets IKBKE and Tmed9. RNA 18, 135–144.

Sado, T., Wang, Z., Sasaki, H., and Li, E. (2001). Regulation of imprinted

X-chromosome inactivation in mice by Tsix. Development 128, 1275–1286.

Sado, T., Hoki, Y., and Sasaki, H. (2005). Tsix silences Xist through modifica-

tion of chromatin structure. Dev. Cell 9, 159–165.

Searle, A.G., and Beechey, C.V. (1978). Complementation studies with mouse

translocations. Cytogenet. Cell Genet. 20, 282–303.

Sharman, G.B. (1971). Late DNA replication in the paternally derived

X chromosome of female kangaroos. Nature 230, 231–232.

Shen, Y., Matsuno, Y., Fouse, S.D., Rao, N., Root, S., Xu, R., Pellegrini, M.,

Riggs, A.D., and Fan, G. (2008). X-inactivation in female human embryonic

stem cells is in a nonrandom pattern and prone to epigenetic alterations.

Proc. Natl. Acad. Sci. USA 105, 4709–4714.

Shin, J., Bossenz, M., Chung, Y., Ma, H., Byron, M., Taniguchi-Ishigaki, N.,

Zhu, X., Jiao, B., Hall, L.L., Green, M.R., et al. (2010). Maternal Rnf12/RLIM

is required for imprinted X-chromosome inactivation in mice. Nature 467,

977–981.

Silva, S.S., Rowntree, R.K., Mekhoubad, S., and Lee, J.T. (2008). X-chromo-

some inactivation and epigenetic fluidity in human embryonic stem cells.

Proc. Natl. Acad. Sci. USA 105, 4820–4825.

Skuse, D.H. (2000). Imprinting, the X-chromosome, and the male brain: ex-

plaining sex differences in the liability to autism. Pediatr. Res. 47, 9–16.

Skuse, D.H. (2005). X-linked genes and mental functioning. Hum. Mol. Genet.

14(Spec No 1), R27–R32.

Skuse, D.H., James, R.S., Bishop, D.V., Coppin, B., Dalton, P., Aamodt-

Leeper, G., Bacarese-Hamilton, M., Creswell, C., McGurk, R., and Jacobs,

P.A. (1997). Evidence from Turner’s syndrome of an imprinted X-linked locus

affecting cognitive function. Nature 387, 705–708.

Sleutels, F., and Barlow, D.P. (2002). The origins of genomic imprinting in

mammals. In Homology Effects, J.C. Dunlap and C.-t. Wu, eds. (San Diego:

Academic Press), pp. 119–154.

Smits, G., Mungall, A.J., Griffiths-Jones, S., Smith, P., Beury, D., Matthews, L.,

Rogers, J., Pask, A.J., Shaw, G., VandeBerg, J.L., et al.; SAVOIR Consortium.

(2008). Conservation of the H19 noncoding RNA and H19-IGF2 imprinting

mechanism in therians. Nat. Genet. 40, 971–976.

Solter, D. (1988). Differential imprinting and expression of maternal and

paternal genomes. Annu. Rev. Genet. 22, 127–146.

Stadtfeld, M., Apostolou, E., Akutsu, H., Fukuda, A., Follett, P., Natesan, S.,

Kono, T., Shioda, T., and Hochedlinger, K. (2010). Aberrant silencing of

imprinted genes on chromosome 12qF1 in mouse induced pluripotent stem

cells. Nature 465, 175–181.

Stadtfeld, M., Apostolou, E., Ferrari, F., Choi, J., Walsh, R.M., Chen, T., Ooi,

S.S., Kim, S.Y., Bestor, T.H., Shioda, T., et al. (2012). Ascorbic acid prevents

loss of Dlk1-Dio3 imprinting and facilitates generation of all-iPS cell mice

from terminally differentiated B cells. Nat. Genet. 44, 398–405, S1–S2.

Starmer, J., andMagnuson, T. (2009). A newmodel for randomX chromosome

inactivation. Development 136, 1–10.

Sun, B.K., Deaton, A.M., and Lee, J.T. (2006). A transient heterochromatic

state in Xist preempts X inactivation choice without RNA stabilization. Mol.

Cell 21, 617–628.

Sun, B., Ito, M., Mendjan, S., Ito, Y., Brons, I.G., Murrell, A., Vallier, L., Fergu-

son-Smith, A.C., and Pedersen, R.A. (2012). Status of genomic imprinting in

epigenetically distinct pluripotent stem cells. Stem Cells 30, 161–168.

Sunderam, S., Kissin, D.M., Flowers, L., Anderson, J.E., Folger, S.G., Jamie-

son, D.J., and Barfield, W.D.; Centers for Disease Control and Prevention

1322 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

(CDC). (2012). Assisted reproductive technology surveillance—United States,

2009. MMWR Surveill. Summ. 61, 1–23.

Surani, M.A., Barton, S.C., and Norris, M.L. (1984). Development of reconsti-

tuted mouse eggs suggests imprinting of the genome during gametogenesis.

Nature 308, 548–550.

Tachibana, M., Ma, H., Sparman, M.L., Lee, H.-S., Ramsey, C.M., Woodward,

J.S., Sritanaudomchai, H., Masterson, K.R., Wolff, E.E., Jia, Y., and Mitalipov,

S.M. (2012). X-chromosome inactivation in monkey embryos and pluripotent

stem cells. Dev. Biol. 371, 146–155.

Tada, T., Obata, Y., Tada, M., Goto, Y., Nakatsuji, N., Tan, S., Kono, T., and

Takagi, N. (2000). Imprint switching for non-random X-chromosome inactiva-

tion during mouse oocyte growth. Development 127, 3101–3105.

Takagi, N., and Abe, K. (1990). Detrimental effects of two active X chromo-

somes on early mouse development. Development 109, 189–201.

Takagi, N., and Sasaki, M. (1975). Preferential inactivation of the paternally

derived X chromosome in the extraembryonic membranes of the mouse.

Nature 256, 640–642.

Tchieu, J., Kuoy, E., Chin, M.H., Trinh, H., Patterson, M., Sherman, S.P.,

Aimiuwu, O., Lindgren, A., Hakimian, S., Zack, J.A., et al. (2010). Female

human iPSCs retain an inactive X chromosome. Cell Stem Cell 7, 329–342.

Terranova, R., Yokobayashi, S., Stadler, M.B., Otte, A.P., van Lohuizen, M.,

Orkin, S.H., and Peters, A.H. (2008). Polycomb group proteins Ezh2 and

Rnf2 direct genomic contraction and imprinted repression in early mouse

embryos. Dev. Cell 15, 668–679.

Thorvaldsen, J.L., and Bartolomei, M.S. (2007). SnapShot: imprinted gene

clusters. Cell 130, 958.

Thorvaldsen, J.L., Duran, K.L., and Bartolomei, M.S. (1998). Deletion of the

H19 differentially methylated domain results in loss of imprinted expression

of H19 and Igf2. Genes Dev. 12, 3693–3702.

Thorvaldsen, J.L., Krapp, C., Willard, H.F., and Bartolomei, M.S. (2012).

Nonrandom X chromosome inactivation is influenced by multiple regions on

the murine X chromosome. Genetics 192, 1095–1107.

Tian, D., Sun, S., and Lee, J.T. (2010). The long noncoding RNA, Jpx, is

a molecular switch for X chromosome inactivation. Cell 143, 390–403.

Tierling, S., Dalbert, S., Schoppenhorst, S., Tsai, C.E., Oliger, S., Ferguson-

Smith, A.C., Paulsen, M., and Walter, J. (2006). High-resolution map and

imprinting analysis of the Gtl2-Dnchc1 domain on mouse chromosome 12.

Genomics 87, 225–235.

Tomoda, K., Takahashi, K., Leung, K., Okada, A., Narita, M., Yamada, N.A.,

Eilertson, K.E., Tsang, P., Baba, S., White, M.P., et al. (2012). Derivation condi-

tions impact X-inactivation status in female human induced pluripotent stem

cells. Cell Stem Cell 11, 91–99.

Turner, J.M., Mahadevaiah, S.K., Ellis, P.J., Mitchell, M.J., and Burgoyne, P.S.

(2006). Pachytene asynapsis drives meiotic sex chromosome inactivation

and leads to substantial postmeiotic repression in spermatids. Dev. Cell 10,

521–529.

Weksberg, R. (2010). Imprinted genes and human disease. Am. J. Med. Genet.

C. Semin. Med. Genet. 154C, 317–320.

Weksberg, R., Shuman, C., and Smith, A.C. (2005). Beckwith-Wiedemann

syndrome. Am. J. Med. Genet. C. Semin. Med. Genet. 137C, 12–23.

Williamson, C.M., Ball, S.T., Dawson, C., Mehta, S., Beechey, C.V., Fray, M.,

Teboul, L., Dear, T.N., Kelsey, G., and Peters, J. (2011). Uncoupling anti-

sense-mediated silencing and DNA methylation in the imprinted Gnas cluster.

PLoS Genet. 7, e1001347.

Wutz, A. (2011). Gene silencing in X-chromosome inactivation: advances in

understanding facultative heterochromatin formation. Nat. Rev. Genet. 12,

542–553.

Wutz, A., Rasmussen, T.P., and Jaenisch, R. (2002). Chromosomal silencing

and localization are mediated by different domains of Xist RNA. Nat. Genet.

30, 167–174.

Xu, N., Tsai, C.L., and Lee, J.T. (2006). Transient homologous chromosome

pairing marks the onset of X inactivation. Science 311, 1149–1152.

Xu, N., Donohoe, M.E., Silva, S.S., and Lee, J.T. (2007). Evidence that homol-

ogous X-chromosome pairing requires transcription and Ctcf protein. Nat.

Genet. 39, 1390–1396.

Yildirim, E., Kirby, J.E., Brown, D.E., Mercier, F.E., Sadreyev, R., Scadden,

D.T., and Lee, J.T. (2013). Xist RNA is a potent suppressor of hematologic

cancer in mice. Cell 152, 727–742.

Yin,Q.F., Yang, L., Zhang,Y., Xiang, J.F.,Wu,Y.W.,Carmichael,G.G., andChen,

L.L. (2012). Long noncoding RNAs with snoRNA ends. Mol. Cell 48, 219–230.

Zhao, J., Sun, B.K., Erwin, J.A., Song, J.J., and Lee, J.T. (2008). Polycomb

proteins targeted by a short repeat RNA to themouse X chromosome. Science

322, 750–756.

Zhao, J., Ohsumi, T.K., Kung, J.T., Ogawa, Y., Grau, D.J., Sarma, K., Song,

J.J., Kingston, R.E., Borowsky, M., and Lee, J.T. (2010). Genome-wide

identification of polycomb-associated RNAs by RIP-seq. Mol. Cell 40,

939–953.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1323

Leading Edge

Review

Epigenetics of Reprogrammingto Induced Pluripotency

Bernadett Papp1,2,3,4 and Kathrin Plath1,2,3,4,*1Department of Biological Chemistry, David Geffen School of Medicine2Jonsson Comprehensive Cancer Center3Bioinformatics Interdepartmental Degree Program, Molecular Biology Institute4Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research

University of California, Los Angeles, Los Angeles, CA 90095, USA

*Correspondence: kplath@mednet.ucla.eduhttp://dx.doi.org/10.1016/j.cell.2013.02.043

Reprogramming to induced pluripotent stem cells (iPSCs) proceeds in a stepwise manner withreprogramming factor binding, transcription, and chromatin states changing during transitions.Evidence is emerging that epigenetic priming events early in the process may be critical for pluri-potency induction later. Chromatin and its regulators are important controllers of reprogramming,and reprogramming factor levels, stoichiometry, and extracellular conditions influence theoutcome. The rapid progress in characterizing reprogramming is benefiting applications of iPSCsand is already enabling the rational design of novel reprogramming factor cocktails. However,recent studies have also uncovered an epigenetic instability of the X chromosome in human iPSCsthat warrants careful consideration.

Decades of research were dedicated to studies of cell fate

changes during development and led to the view that, in vivo,

differentiated cells are irreversibly committed to their fate.

However, reprogramming of somatic cells by transfer into

enucleated oocytes pioneered by John Gurdon and colleagues

in the 1950s (Gurdon et al., 1958), fusion with other cell partners

(Blau et al., 1983), and ectopic transcription factor expression

(Davis et al., 1987; Takahashi and Yamanaka, 2006) revealed

a remarkable plasticity of the differentiated state. Particularly

the exposure to ectopic transcription factors offers a powerful

and unexpectedly flexible technique to shift a somatic cell

toward alternative somatic identities or pluripotency. The

reprogramming field exploded after Takahashi and Yamanaka

established a major landmark with the generation of induced

pluripotent stem cells (iPSCs) from fibroblasts by simple ectopic

expression of Oct4 (O), Sox2 (S), cMyc (M), and Klf4 (K) (Taka-

hashi and Yamanaka, 2006). Aptly, the Nobel Prize awarded to

John Gurdon and Shinya Yamanaka in 2012 symbolizes the

extraordinary contribution that reprogramming experiments

have made (and will make) to our understanding of cellular iden-

tity and the apparently unlimited practical applications of iPSCs

and other reprogrammed cells.

This Review focuses on reprogramming to iPSCs. The beauty

of transcription-factor-induced reprogramming to iPSCs lies in

its simplicity and robustness, as many different cell types from

a wide range of species can be reprogrammed to pluripotency

by ectopic expression of OSKM (for a recent summary, see

Stadtfeld and Hochedlinger [2010]). A fundamental feature of

the resulting iPSCs is that they are, in their ideal state, function-

ally indistinguishable from embryonic stem cells (ESCs), which

are pluripotent cells derived from preimplantation embryos,

1324 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

and are capable of differentiation into cells of all three germ

layers (Bock et al., 2011; Carey et al., 2011). Consequently, re-

programming changes the transcriptome and chromatin state

of the somatic cell to that of a pluripotent cell (Chin et al.,

2009; Hawkins et al., 2010; Lister et al., 2011; Maherali et al.,

2007; Mikkelsen et al., 2008; Okita et al., 2007; Takahashi and

Yamanaka, 2006; Wernig et al., 2007). Therefore, iPSCs offer

an invaluable source of patient-specific pluripotent stem cells

for disease modeling, drug screening, toxicology tests, and

regenerative medicine (recently reviewed in Onder and Daley

[2012]; Trounson et al., 2012), and already have been employed

to unmask novel insights into human diseases (Koch et al., 2011).

Despite the extraordinary fidelity of the iPSC technology, the

induction of pluripotency upon OSKM expression typically

requires an extended latency period of around 1–2 weeks and

occurs in less than 1% of the starting cells, even when they are

genetically identical and the expression levels of the four tran-

scription factors are similar across all cells in the culture dish

(for a review, see Stadtfeld and Hochedlinger [2010]). Although

heterogeneity of the starting cell population and differentiation

state may affect reprogramming efficiency to a certain degree

(Stadtfeld and Hochedlinger, 2010), a key question has been

why only a few of a pool of seemingly equivalent OSKM-express-

ing cells induce pluripotency. Genomic approaches, RNAi

screens, and simpler genetic methods, as well as emerging

single-cell analyses, arebeginning toprovideanswersbydefining

critical reprogramming events as well as regulators and epige-

netic properties that promote or hinder reprogramming transi-

tions, which we will focus on in the first part of this Review.

Particularly the activation of pluripotency genes appears

to present a formidable task for the reprogramming factors.

Generally, transcriptional activation begins with the binding of

transcription factors to distal enhancer and promoter elements,

which initiates the recruitment of coactivators and facilitates

the binding of the general transcription machinery and the

assembly of the RNA polymerase-II-containing preinitiation

complex (PIC) at the core promoter (Green, 2005). Transcription

factors can also promote steps in the transcription process

subsequent to PIC assembly (which is of interest for the reprog-

ramming factor cMyc) (Green, 2005). Importantly, the packaging

of DNA into nucleosomes affects all aspects of transcription,

from transcription factor binding to PIC formation and transcrip-

tional elongation (Beato and Eisfeld, 1997; Li et al., 2007). The

ability of transcription factors to bind their recognition elements

is further modulated by changes in chromatin structure,

including DNA methylation, histone modifications, histone vari-

ants, or ATP-dependent chromatin remodeling. Chromatin

therefore plays a critical role in the establishment of cell-type-

specific expression patterns and is responsible for the extreme

stability of a given cellular identity under physiological condi-

tions, ensuring the stable silencing of lineage-inappropriate

genes and restricting transcription factor action to only a subset

of their target motifs in the genome (Filion et al., 2010; Gaetz

et al., 2012). In differentiated cells, pluripotency loci therefore

appear to be in an unfavorable chromatin landscape for binding

by most transcription factors. However, we will discuss the

remarkable capability of reprogramming factors to engage

closed chromatin and induce extensive chromatin changes

early in reprogramming before anymajor transcriptional changes

take place, unmasking interesting parallels between reprogram-

ming and developmental processes and highlighting the power

of the OSKM reprogramming cocktail. Together, these recent

findings have transformed the iPSC system into a powerful

model for the dissection of mechanisms underlying cell fate tran-

sitions.

The reprogramming process is most scrutinized in the mouse

system, but studies of the induced pluripotent state have been

extensively performed for both mouse and human iPSCs. Most

likely due to the fact that conventional mouse and human iPSCs

represent different states of pluripotency, these cells differ

epigenetically, as highlighted by their X chromosome inactiva-

tion state. In the second part of this Review, we will discuss

a selection of recent studies that revealed an epigenetic insta-

bility of the inactive X chromosome in female human iPSCs, remi-

niscent of processes in human ESCs, and we will focus on the

implications of these findings for the utility of iPSCs.

Steps Leading to the iPSC StateThe development of improved reprogramming techniques that

include homogeneous and inducible reprogramming factor

expression systems (summarized in Stadtfeld and Hochedlinger

[2010]) has enabled a more detailed view of the mechanism

underlying reprogramming despite the fact that only few starting

cells become iPSCs. Mouse embryonic fibroblasts are most

commonly used as a starting cell type for the dissection of the

reprogramming process due to the ease of culture and the possi-

bility of derivation from different genetic backgrounds and

mouse models. Current evidence argues that reprogramming

of these cells to iPSCs requires cell division (Hanna et al.,

2009) and is a multistep process in which the successful induc-

tion of the pluripotent state entails the transition through sequen-

tial gene expression states (or intermediates) (Figure 1). Failure to

transition through any of these steps would lead to a block in

reprogramming and would account for the low overall reprog-

ramming efficiency. Consistent with this model, it was shown

early on by the Jaenisch and Hochedlinger groups that reprog-

ramming cultures represent heterogeneous cell populations

that can be resolved based on the expression of cell surface

markers (Brambrink et al., 2008; Stadtfeld et al., 2008). Utilizing

specific surface marker combinations, cells poised to become

iPSCs can be enriched at different times of reprogramming.

This knowledge allowed the inference of a reprogramming path

in which successfully reprogramming cells first downregulate

the fibroblast-associated marker Thy1 and then transition to

a state that is positive for the embryonic marker SSEA1 and,

finally, induce the full pluripotency network (Brambrink et al.,

2008; Polo et al., 2012; Stadtfeld et al., 2008) (Figure 1). The

downregulation of Thy1 occurs in a large fraction of starting cells,

the subsequent gain of SSEA1 only in a subset of Thy1-negative

cells, and the induction of the pluripotency network in a small

subset of SSEA1-positive cells, indicating that transitions

between each of these steps occurwith lowprobability (Figure 1).

Cells that are unable to silence Thy1 relatively quickly upon

OSKM expression become refractory to the action of the reprog-

ramming factors and can yield iPSCs though with dramatic delay

and at much lower efficiency (Polo et al., 2012). Accordingly,

a single-cell cloning experiment demonstrated that virtually all

starting cells have the potential to induce pluripotency in a small

subset of their daughter cells when reprogramming is followed

over a 6 month period (Hanna et al., 2009). The intermediate

states defined by cell sorting experiments likely represent the

most favored possibilities on the path of reprogramming. Further

purification of reprogramming intermediates should be feasible

and provide insight into whether all reprogramming cells have

to pass through the same stages to induce pluripotency. Of

interest, SSEA1-positive intermediate cells are still plastic early

in reprogramming in that some of these cells can regress to

the Thy1-positive (i.e., an earlier) reprogramming state in the

presence of reprogramming factor expression. By contrast, later

in reprogramming, these cells appear to have matured and

becomemuchmore committed to progressing to the pluripotent

state (Polo et al., 2012), indicating that cellular identity is only

stabilized and locked in toward the end of the reprogramming

process.

Genome-wide transcriptional profiling was used to further

delineate the sequence of events that drive reprogramming.

Initially, cells appear to respond relatively homogeneously to

the expression of the reprogramming factors (Polo et al.,

2012) and robustly silence typical mesenchymal genes ex-

pressed in fibroblasts (such as Snai1, Snai2, Zeb1, and

Zeb2) (Li et al., 2010; Mikkelsen et al., 2008; Polo et al., 2012;

Samavarchi-Tehrani et al., 2010). These events lead to the acti-

vation of epithelial markers (such as Cdh1, Epcam, and Ocln) in

a process called mesenchymal-to-epithelial transition (MET),

which seems critical for the early reprogramming phase and is

accompanied by morphological changes, increased prolifera-

tion, and the formation of cell clusters (Li et al., 2010; Mikkelsen

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1325

Figure 1. The Generation of iPSCs Is a Multistep Process that Can Be Modulated by Extracellular Cues and Reprogramming Factor LevelsKnown events occurring in early, middle, and late phases during the OSKM-mediated reprogramming of mouse embryonic fibroblasts to iPSCs are depicted.During the final emergence of fully reprogrammed iPSCs, so-called ‘‘reprogramming-competent cells’’ appear to be inhibited by the continued expression of thefactors. The reprogramming process can be preferentially trapped in partially reprogrammed states when certain reprogramming factor levels and/or stoichi-ometries are employed (top) or can be redirected to a different cell identity, without going through the pluripotent state, by changing culture/growth factorconditions and timing of OSKM expression (bottom).

et al., 2008; Samavarchi-Tehrani et al., 2010; Smith et al., 2010).

Notably, the aforementioned transition to the SSEA1-positive

state appears to correlate with the occurrence of MET (Polo

et al., 2012; Samavarchi-Tehrani et al., 2010) (Figure 1). The

key characteristic of subsequent reprogramming phase is the

gradual activation of pluripotency-associated genes (Brambrink

et al., 2008; Buganim et al., 2012; Golipour et al., 2012;

Mikkelsen et al., 2008; Polo et al., 2012; Samavarchi-Tehrani

et al., 2010; Sridharan et al., 2009; Stadtfeld et al., 2008). For

example, the pluripotency loci Nanog and Sall4 are transcrip-

tionally upregulated at a late intermediate stage, whereas

others, such as Utf1 or endogenous Sox2, are induced even

later, closely mirroring the acquisition of the full pluripotency

expression programming (Figure 1). Although detailed time

course studies describing these transitions in reprogramming

cells still need to be performed at the single-cell level, a recent

single-cell expression study that compared the expression of

candidate genes at various reprogramming stages strongly

supports a series of consecutive pluripotency gene activation

steps late in the reprogramming process (Buganim et al.,

2012). Together, these events culminate in the establishment

of the pluripotent state that can be sustained independently of

ectopic reprogramming factor expression (Brambrink et al.,

2008; Maherali et al., 2007; Okita et al., 2007; Stadtfeld et al.,

2008; Wernig et al., 2007).

1326 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Modifying the Reprogramming ProcessEarly studies employing inducible reprogramming factor ex-

pression systems indicated that reprogramming intermediates

are dependent on continued OSKM expression to complete

the reprogramming process (Brambrink et al., 2008; Stadtfeld

et al., 2008). In addition, evidence is growing that the efficiency

of reprogramming is strongly influenced by the levels of the re-

programming factors. For example, fibroblasts engineered to

express a higher dose of OSKM in all cells have a dramatically

enhanced ability to induce pluripotency (Polo et al., 2012). A

peculiar observation is that cells that become refractory to re-

programming early on (and stay Thy1 positive) have dramatically

reduced protein levels of the four reprogramming factors

compared to cells that are able to progress toward pluripotency

(Polo et al., 2012). Because the RNA levels of the reprogramming

factors are similar between these two cell populations, these

transcription factors may be prone to increased ubiquitination

and degradation specifically in refractory cells (Buckley et al.,

2012; Polo et al., 2012). Furthermore, the inability to sustain

high reprogramming factor expression contributes strongly to

the reprogramming block in refractory cells, as a further increase

in OSKM expression specifically in these cells induces them to

convert to the next reprogramming stage and subsequently to

iPSCs more efficiently (Polo et al., 2012). Although continuity of

reprogramming factor expression is essential for driving somatic

cells toward pluripotency, a recent study pointed out that high

levels of ectopic OSKM during the final reprogramming steps

may be inhibitory to the efficient induction of the full pluripotency

network (Golipour et al., 2012) (Figure 1). This finding is consis-

tent with the observations that retrovirally expressed reprogram-

ming factors are efficiently turned off in faithfully reprogrammed

cells (Maherali et al., 2007; Okita et al., 2007; Wernig et al., 2007)

and that the activation of endogenous pluripotency regulators

during reprogramming coincides with transgene independence

(Stadtfeld et al., 2008). The reduction of ectopic reprogramming

factors at the end of reprogramming may be necessary because

even a modest increase in Oct4 levels in ESCs is detrimental to

the pluripotent state (Niwa et al., 2000).

Not just overall levels and timing, but also the specific balance

of the reprogramming factors relative to each other are critical for

the outcome of reprogramming (Figure 1). For example, many

studies agree that high Oct4 levels and low levels of Sox2

increase the efficiency of reprogramming (Nagamatsu et al.,

2012; Tiemann et al., 2011; Yamaguchi et al., 2011). High Sox2

levels have been associated with the stronger induction of devel-

opmental markers during reprogramming, whichmay guide cells

away from the path to pluripotency (Yamaguchi et al., 2011).

Moreover, even though ectopic expression of cMyc enhances

reprogramming, it also leads to emergence of a large fraction

of partially reprogrammed ESC-like colonies trapped before

the upregulation of the pluripotency program (Nakagawa et al.,

2008; Wernig et al., 2008). Remarkably, differences in reprog-

ramming factor stoichiometry appear to have consequences

for the epigenetic state and developmental potential of the

resulting iPSCs (Carey et al., 2011). This is an interesting result

in light of the ongoing debate on epigenetic differences between

iPSCs and ESCs (for a recent discussion, see Lowry, [2012]) and

suggests that at least some (and maybe all) of the observed vari-

ations between iPSCs and ESCs are not inherent to the reprog-

ramming process but are due to experimental variables that

often are not easy to control, highlighting how a better under-

standing of the mechanisms underlying reprogramming will

benefit the production of safer iPSC lines.

The efficiency of iPSC formation can also be improved by

altering media composition and growth factor conditions (Chen

et al., 2011; Esteban et al., 2010; Ichida et al., 2009; Li et al.,

2010; Samavarchi-Tehrani et al., 2010). Though it is likely that

downstream effectors of signaling pathways directly alter the

transcriptional output of their target genes, specific culture

conditions can also modulate the activity and levels of chromatin

regulators, thereby indirectly affecting OSKM functionality (Chen

et al., 2013; Marks et al., 2012; Wang et al., 2011a; Zhu et al.,

2013). To mention just one example, vitamin C (ascorbic acid)

addition to the media increases reprogramming efficiency and

potentially the quality of resulting iPSCs at least in part by influ-

encing the functionality of histone demethylases that depend on

iron (Esteban et al., 2010; Stadtfeld et al., 2012; Wang et al.,

2011a).

Notably, by supplementing OSKM-reprogramming cultures

with a growth factor cocktail normally required for the establish-

ment and maintenance of epiblast stem cells (EpiSCs), mouse

fibroblasts can be reprogrammed to an EpiSC-like state instead

of the ESC-like iPSC state (Han et al., 2011) (Figure 1). Mouse

EpiSCs and ESCs capture two different states of pluripotency,

which will be discussed in greater detail in the second part of

this Review. During the last couple of years, it has also become

clear that OSKM (or a subset of these factors) can even prompt

the establishment of various somatic cell fates, including cardio-

myocytes, blood progenitors, and neural stem cells, when over-

expressed temporally and guided by appropriate extracellular

cues, without the transition through the pluripotent state

(Figure 1) (reviewed in Sancho-Martinez et al. [2012]). The induc-

tion of various developmental regulators at intermediate stages

of reprogramming to pluripotency may explain why OSKM can

efficiently redirect the reprogramming path to other cell identities

upon exposure to suitable signaling cues and likely reflects

a function of Sox2 and Klf4 as critical regulators of various differ-

entiation paths during development (Mikkelsen et al., 2008; Polo

et al., 2012; Sridharan et al., 2009). Alternatively, and not mutu-

ally exclusive, reprogramming intermediates arising due to

OSKM expression may represent normally occurring develop-

mental progenitor states. Though the picture is emerging that

signaling cues affect the cell fate choices made during reprog-

ramming and/or lead to the stabilization of particular cell identi-

ties that arise during the process, still relatively little is known

about the exact role of signaling pathways and their downstream

regulators in reprogramming and the intersection with the re-

programming factors. Comparing the molecular dynamics of

OSKM-dependent induction of pluripotency and alternative cell

fates should demonstrate how cell fate decision processes can

be efficiently modulated and will facilitate the development of

patient-specific somatic cell populations for clinical applications.

Defining the Target Repertoire of the ReprogrammingFactorsOne approach toward a better understanding of the cascade of

molecular events underlying the establishment of pluripotency is

the definition of reprogramming factor targets at different stages

of the reprogramming process. It is generally believed that three

of the four reprogramming factors, Oct4, Sox2, and Klf4, are

necessary for the induction of pluripotency because they are

critical components of an intrinsic and highly stable pluripotency

network (Boyer et al., 2005; Chen et al., 2008; Jiang et al., 2008;

Kim et al., 2008; Loh et al., 2006; Sridharan et al., 2009). Oct4,

Sox2, and Klf4 tend to colocalize at many cell-type-specific

enhancers in ESCs, often together with additional pluripotency

transcription factors like Nanog, Esrrb, Klf2, Sall4, and Zfp42

and signaling pathway regulators such as Smad1 and Stat3

(Chen et al., 2008; Kim et al., 2008), reinforcing the importance

of OSK for the pluripotent state and the view that enhancers

are sentinels of cell-type-specific gene expression patterns (Vi-

sel et al., 2009). The integration of numerous pluripotency tran-

scription factors and signaling cues at these enhancers ensures

the expression of many genes with known roles in pluripotency

and provides stability to the ESC gene expression program.

Another important aspect of the pluripotency network is that

many pluripotency transcription factors constitute a transcrip-

tional circuit wired in a feed-forward type of regulation, as they

induce their own expression and positively regulate each other

(Boyer et al., 2005; Chen et al., 2008; Jiang et al., 2008; Kim

et al., 2008) (Figure 2A).

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1327

Figure 2. Features of OSKM in ESCs and during Reprogramming(A) In ESCs, Oct4, Sox2, and Klf4 bind their own and each others’ promoters and enhancers, as well as those of many additional ESC-specific (pluripotency)genes. Further contributing to the pluripotency circuitry, many of these ESC-specific genes are also bound by various additional pluripotency regulators, includingNanog and Esrrb, such that ESC-specific enhancers represent hot spots of pluripotency transcription factor binding.(B) In ESCs (andmany other cell types), cMyc targetsmost actively transcribed genes at the core promoter by binding high-affinity E box sequences and functionsby enhancing transcriptional elongation. Expression levels correlate with cMyc occupancy. Upon overexpression, cMyc does not appear to regulate new targetgenes but amplifies the existing gene expression pattern by binding the same genes at elevated levels and occupying additional, low-affinity E-box-likesequences in both the core promoter and enhancer regions of these genes.(C) Scheme illustrating different contributions of the reprogramming factors to the late phase of reprogramming, highlighting separable engagement of OSKand cMyc. Many genes occupied by cMyc in ESCs/iPSCs are already bound by this transcription factor and are expressed in partially reprogrammed cells,which represent a clonal, late reprogramming intermediate. By contrast, OSK bind the promoter regions of many of their ESC/iPSC-specific target genes onlylate in reprogramming, accompanying their transcriptional upregulation. This is particularly obvious for those genes that are cobound by OSK in their promoterregion in ESCs.(D) Chromatin can affect the ability of transcription factors to bind to their DNA motifs, which is thought to explain why most transcription factors bind to only asmall subset of their recognition motifs in the genome. Here, we summarize the chromatin preferences of the four reprogramming factors early in reprogramming.

By contrast, cMyc is unique among the reprogramming

factors, as it is neither a component of the core pluripotency

network (Chen et al., 2008; Kim et al., 2010) nor absolutely

necessary for reprogramming to iPSCs (Nakagawa et al., 2008;

Wernig et al., 2008). Indeed, cMyc is a central player in many

diverse biological processes, including cell growth and differen-

tiation. Two recent reports (Lin et al., 2012; Nie et al., 2012)

strongly support a model in which cMyc is not a transcription

factor that is responsible for OFF/ON switches of its target genes

as proposed for OSK. Instead, cMyc is a nonlinear amplifier of

transcriptional outputs that acts universally on active genes con-

taining the E box DNA motif. Mechanistically, cMyc promotes

transcription by regulating RNA polymerase II pause-release

and by increasing the rate of transcriptional elongation (Rahl

et al., 2010). Therefore, cMyc occupies the core promoter

regions of many active genes in ESCs/iPSCs and is typically

1328 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

not present at enhancers (Chen et al., 2008; Kim et al., 2010;

Nie et al., 2012; Soufi et al., 2012; Sridharan et al., 2009)

(Figure 2B). Analysis of cMyc binding across different inducible

expression levels in tumor cells demonstrated that cMyc

predominantly binds high-affinity E box sites at core promoters

of almost all active genes when expressed at low levels but spills

over to weaker E box sites within enhancers of the same active

genes upon higher expression, likely because promoter sites

become saturated (Figure 2B) (Lin et al., 2012; Nie et al., 2012).

Thus, the target repertoire of cMyc does not change when

cMyc is strongly expressed, but transcriptional output is

increased. The significant differences between OSK and cMyc

have important implications for the reprogramming process.

Oct4, Sox2, and Klf4 are probably crucial for specifying cell

fate change in reprogramming, whereas cMyc may simply act

by amplifying arising expression changes due to OSK action at

genes that contain E boxes, potentially helping to trap genes in

the ON state.

The low efficiency of reprogramming makes the application of

genome-wide analysis techniques of reprogramming factor

binding, such as chromatin immunoprecipitation combined

with massive parallel sequencing (ChIP-seq), challenging for

cells at intermediate stages of the reprogramming process. To

circumvent this problem, our lab initially mapped reprogramming

factor binding within promoter regions in iPSCs and in partially

reprogrammed cells—which represent a clonal, trapped late re-

programming intermediate expanded from ESC-like colonies

that arise in reprogramming cultures and fail to express pluripo-

tency regulators—and compared occupancy data with gene

expression patterns (Sridharan et al., 2009). In both cell types,

genes co-occupied by the reprogramming factors are highly ex-

pressed, indicating that an intrinsic property of reprogramming

factor cobinding is to activate genes. Interestingly, genes that

are more highly expressed in partially reprogrammed cells than

in ESCs are often more efficiently targeted by the OSKM factors

in the intermediate state than in ESCs, whereas genes more

highly expressed in ESCs are generally less bound in partially re-

programmed cells than in ESCs. Thus, many genes are more

strongly expressed in partially reprogrammed cells compared

to ESCs due to targeting of the four factors to promoter regions

that they do not normally bind in ESCs, and conversely, the

failure to activate ESC-specific genes appears to result from

the inability of the factors to bind these genes in the intermediate

state. These findings are consistent with the reprogramming

factors being directly responsible for the ‘‘ectopic’’ expression

of developmental genes in reprogramming intermediates, which

is known to hinder reprogramming (Mikkelsen et al., 2008).

Notably, the widespread lack of ESC-specific promoter binding

in partially reprogrammed cells impinges more dramatically on

Oct4, Sox2, and Klf4 than on cMyc and particularly affects

many pluripotency-related genes that are co-occupied by

combinations of Oct4, Sox2, and Klf4 in ESCs (Figure 2C). In

the case of these genes, it appears that the OSK promoter

engagement occurs only toward the very end of the reprogram-

ming process and is likely required for their transcriptional

activation (Figure 2C). These findings not only demonstrate

a separable contribution of cMyc and OSK to the activation of

various pluripotency loci and a change in the reprogramming

factor target repertoire during the reprogramming process, but

also indicate that the promoter engagement of key pluripotency

genes is a critical task for reprogramming.

Recently, Zaret and colleagues obtained a picture of the initial

chromatin engagement of the reprogramming factors by per-

forming ChIP-seq 48 hr after the induction of reprogramming

factor expression in human fibroblasts (Soufi et al., 2012),

when most cells still undergo very similar expression changes

(see above) (Polo et al., 2012). Comparing OSKM-binding

patterns between the early reprogramming stage and the plurip-

otent state, Zaret and colleagues made two interesting observa-

tions (Soufi et al., 2012). First, many more genes are bound by all

four factors early in reprogramming than in the pluripotent state,

which could be due to the high expression levels of the induced

factors. In addition, OSKM binding of apoptosis-regulating

genes early in the process suggests that the extensive cell death

apparent in reprogramming cultures (reviewed in Plath and

Lowry [2011]) is a direct consequence of reprogramming factor

binding, potentially representing a general cellular defense

mechanism against ectopic transcription factor expression

(Soufi et al., 2012). Furthermore, initial target genes of the re-

programming factors are significantly enriched for regulators of

MET, the critical early reprogramming event discussed above,

whereas pluripotency loci such as NANOG and DPPA4 are not

yet bound, corroborating that a redistribution of OSKM binding

occurs as cells move along the reprogramming path and sug-

gesting that, initially, the reprogramming factors directly target

at least some of the genes that transcriptionally change early in

the process. The second and more surprising finding is that

the reprogramming factors interact extensively with distal

genomic sites, including some known enhancers. Indeed, 85%

of all initial binding events occur distal to promoter regions (Soufi

et al., 2012). Because it appears that, in the pluripotent state, the

transcription factors have shifted to a binding pattern that

includes promoter regions much more strongly, Zaret and

colleagues proposed that the binding of the reprogramming

factors to distal elements is an early step in reprogramming

that precedes promoter binding and transcriptional activation

of many target genes (Soufi et al., 2012).

Reprogramming Factors as PioneersThe next question then is which features anticipate the recruit-

ment of ectopically expressed OSKM? The DNA motifs of the

four factors are enriched at their respective binding sites, indi-

cating that they are recruited directly through their sequence

motifs rather than randomly targeting or scanning the genome

(Soufi et al., 2012; Sridharan et al., 2009). However, transcription

factors work in a concentration-dependent manner and will, at

higher concentration, also occupy DNA sites of lower affinity,

which may be important for reprogramming, during which very

high levels of ectopic OSKM are expressed (Lin et al., 2012;

Nie et al., 2012; Soufi et al., 2012) (Figure 2B). Notably, lineage

specification factors present in the starting cell type may

contribute to the targeting of the reprogramming factors to

a subset of their DNA motifs. For example, during lineage

development, Sox transcription factors often occupy sites pre-

marked by other Sox proteins that were expressed in the

previous developmental stage (Bergsland et al., 2011). If such

lineage-specific factors are involved in the initial targeting of

the reprogramming factors, one might predict that reprogram-

ming factors will target different genomic locations in different

starting cell types.

Importantly, chromatin is thought to strongly affect the ability

of transcription factors to bind their cognate DNA motifs, and

certain chromatin states, characterized for example by the pres-

ence of specific combinations of histone modifications, may be

especially conducive to DNA binding by specific transcription

factors (Filion et al., 2010). As expected, binding of the reprog-

ramming factors does occur in open and accessible chromatin,

marked by active histone modifications such as H3K4 methyla-

tion (Soufi et al., 2012; Sridharan et al., 2009) (Figure 2D). Among

the reprogramming factors, cMYC binding is much more strictly

associated with a pre-existing active chromatin state than that of

OSK (Soufi et al., 2012; Sridharan et al., 2009), which is

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1329

consistent with active chromatin being a prerequisite for the

binding of cMyc (Guccione et al., 2006) (Figure 2D). An aston-

ishing observation by Zaret and colleagues is that the vast

majority (around 70%) of reprogramming factor binding events

early in human fibroblast reprogramming occurs within genomic

regions that display a closed chromatin state in the starting fibro-

blasts characterized by the absence of DNase hypersensitivity

and, surprisingly, any histone modifications (Soufi et al., 2012).

Thus, the reprogramming factors can efficiently access their

target sequences within genomic regions that are packed with

nucleosomes and are probably even further condensed into

higher-order structures. This is particularly true for OSK and, to

a much lesser extent, for cMYC (Soufi et al., 2012) (Figure 2D).

Indeed, the ability of cMYC to access target sites in closed chro-

matin is dependent on OSK occupancy (Soufi et al., 2012). OSK

can occupy OSKM cobound sites in the absence of ectopic

cMYC, but cMYC cannot bind when overexpressed in the

absence of ectopic OSK. In turn, ectopic cMYC enhances the

initial binding of OSK to these sites when expressed together.

These data are in agreement with cMyc potentiating the action

of the other three reprogramming factors rather than initiating

these events.

In comparison to naked DNA, nucleosomal DNA is less acces-

sible for DNA-binding factors (Beato and Eisfeld, 1997), and the

majority of transcription factors cannot bind their cognate sites

when sequestered within a nucleosome and need a structural

change in the associated nucleosome or a nucleosome-free

region for binding (Wallrath et al., 1994), highlighting an impor-

tant functionality of OSK. Cooperative binding or simultaneous

engagement of neighboring binding sites could explain the ability

of OSK to interact with nucleosomal-binding sites (Adams and

Workman, 1995). For instance, binding of one factor might

partially destabilize a nucleosome, allowing the other transcrip-

tion factor(s) to access sites that were previously buried.

However, each of the OSK-reprogramming factors alone can

also target sites in closed chromatin, i.e., without the other two

factors being detected at those sites (Soufi et al., 2012). There-

fore, Zaret and colleagues proposed that Oct4, Sox2, and Klf4

each can act as pioneer factors that are able to access closed

chromatin on their own without the help of additional transcrip-

tion factors (Soufi et al., 2012). There is additional evidence in

support of this idea. First, based on three-dimensional (3D)

structures, Oct4, Sox2, and Klf4, but not cMyc, interact with

one side of the DNA helix when bound to DNA, potentially allow-

ing them to bind DNA in the context of the nucleosome (Beato

and Eisfeld, 1997; Soufi et al., 2012). Second, a comparison of

nucleosome occupancy with binding of Oct4 and Sox2 in

ESCs genome-wide suggests that Oct4 and Sox2 can, at least

in part, interact with nucleosomal DNA (Teif et al., 2012). Third,

Sp1, a transcription factor belonging to the same family of highly

related transcription factors as Klf4, can bind nucleosomal DNA

in vitro, making it reasonable to anticipate that Klf4 will share

SP1’s capacity (Li et al., 1994). Fourth, it was found that pre-

existing nucleosomes at the enhancer and promoter regions of

the OCT4 and NANOG gene loci are displaced when OCT4 is

ectopically expressed in differentiated cells (i.e., in the absence

of any other reprogramming factors) (You et al., 2011). This chro-

matin reorganization coincided with Oct4 binding, suggesting

1330 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

that Oct4 is able to directly access DNA sites that are internal

to a nucleosome and establish a nucleosome-depleted region

(You et al., 2011).

The idea of OSK acting as pioneer factors in reprogramming

is exciting because it is reminiscent of developmental decisions,

wherein pioneer factor binding at enhancers occurs early (Gualdi

et al., 1996). The efficient activation of lineage-specific genes

during development often requires a cascade of DNA-transcrip-

tion factor interactions and chromatin changes at their enhancer

and promoter regions, which begin long before these genes are

transcribed (Zaret and Carroll, 2011). Pioneer transcription

factors initiate this series of events by accessing tissue-specific

enhancers already at a very early developmental stage and by

inducing chromatin decondensation, remodeling, and/or

a change in local chromatin modifications, thereby priming

enhancer and promoter regions for binding by additional tran-

scription factors and transcriptional activation at a later stage

of development. Thus, pioneer factors are initiator factors that

make regulatory regions competent for activation in response

to the right stimulus.

In the context of reprogramming, the binding of OSK to closed

chromatin early in reprogramming could therefore be a crucial

step for events that happen later in the process, particularly

considering that some of these distal binding events overlap

with known enhancers. One may speculate that Oct4, Sox2,

and Klf4 can engage at least some ESC-specific enhancers early

in reprogramming even though they are locked up in closed

chromatin in the starting fibroblasts, poising them for promoter

binding and transcriptional activation later in the process. In

the next section, we will provide additional evidence in support

of such epigenetic priming by focusing on chromatin changes

that occur early in the reprogramming process.

Chromatin Changes in Promoters and Enhancers Earlyin ReprogrammingAn analysis of the initial transcriptional and chromatin changes

early in mouse cell reprogramming (i.e., 24–72 hr after induction

of the reprogramming factors) revealed striking parallels to the

initial reprogramming factor binding pattern (Koche et al.,

2011). First, gene expression changes, both up and down, are

largely confined to genes with promoter regions carrying active

chromatin marks in the starting fibroblasts (i.e., in regions

marked by enrichment of H3K4me3, a modification associated

with the transcriptional start sites of active and poised genes)

(Koche et al., 2011). The restriction of expression changes to

genes that are already in an open and accessible chromatin

configuration is consistent with the fact that the perturbation of

the somatic gene expression program is the major response

early in the reprogramming process (Koche et al., 2011;

Mikkelsen et al., 2008; Polo et al., 2012; Samavarchi-Tehrani

et al., 2010; Sridharan et al., 2009).

Unexpectedly, changes in histone modifications are much

more widespread than initial changes in gene expression,

indicating that an extensive genome-wide chromatin remodeling

takes place as immediate response to reprogramming factor

expression (Koche et al., 2011). In addition to chromatin changes

associated with gene expression switches, H3K4me2 (a histone

mark associated with active or poised promoters and enhancers)

Figure 3. Chromatin Dynamics during ReprogrammingMany fibroblast-specific promoters and enhancers are decommissioned early in reprogramming (after 24–48 hr of reprogramming factor expression) by loss ofactive H3K4 methylation marks but appear to gain DNA methylation only late in reprogramming. ESC-specific enhancers and promoters can be divided into atleast two groups: those with dramatic changes in histonemodifications already early in reprogramming, long before their transcriptional activation, and those thatundergo histone modification changes only much later in the process. One key difference between these groups appears to be the DNA methylation state. Forexample, the first group includesmany pluripotency genes with CpG-dense promoter elements (indicated by higher density of circles) that are hypomethylated infibroblasts.

rapidly emerges de novo in many promoter regions in the

absence of transcriptional changes and even before any cell divi-

sion has taken place (Figure 3). Many of these promoters belong

to genes that are transcriptionally activated later in reprogram-

ming, including various pluripotency regulators like Sall4,

Pecam1, FoxD3, and Lin28. The gain of H3K4me2 is not accom-

panied by simultaneous accumulation of the H3K4me3mark and

often occurs on a nucleosome that covers the transcriptional

start site. Because nucleosomes at transcriptional start sites

are incompatible with the assembly of the basic transcriptional

machinery (Lorch et al., 1987), nucleosome depletion must be

one of the subsequent steps that allows transactivation of these

genes later in reprogramming. Interestingly, promoters with

H3K4me2 gain early in reprogramming often display a high

CpG density and are enriched for CpG islands (Koche et al.,

2011) (Figure 3), which may obviate the need for extensive chro-

matin remodeling and therefore facilitate quick changes in chro-

matin structure due to lower nucleosome occupancy (Ramirez-

Carrozzi et al., 2009).

Compared to promoters, chromatin changes at enhancers are

even more prominent early in reprogramming (Koche et al.,

2011), which is consistent with the observations that many

enhancers are active in only a single cell type and that the chro-

matin state of enhancers is more variable across cells types than

that of promoters (Heintzman et al., 2009). The systematic

mapping of enhancers is now possible genome-wide because

specific enhancer-associated chromatin signatures have been

identified that even reveal the activity of the enhancer (Creyghton

et al., 2010; Heintzman et al., 2009; Koche et al., 2011; Rada-

Iglesias et al., 2011). In the active state (i.e., when associated

with an actively transcribed gene), enhancer elements are

demarcated by domains of H3K27Ac and H3K4me1/me2, but

not H3K4me3. In association with inactive genes, enhancers

can be in one of two states: unmarked (i.e., inactive), lacking

all of the features that are associated with the active enhancer

state, or poised, carrying H3K4me1/me2 in the absence of

H3K27ac. It is thought that poised enhancers are important for

the plasticity of developmental decisions, as a subset can

acquire the signature of active enhancers upon change in

external stimuli. The specific enhancer state therefore appears

to strongly influence the ability of the cell to respond to environ-

mental or developmental stimuli. For example, immediate tran-

scriptional changes to a new signaling cue are often restricted

to genes with active and/or poised enhancers, whereas inactive

genes with unmarked (inactive) enhancers remain refractory

(Ghisletti et al., 2010; Heintzman et al., 2009).

In reprogramming, switches in enhancer states occur very

rapidly and extensively, even before the first cell division, high-

lighting an extremely quick departure from the somatic cell

identity (Koche et al., 2011). These changes go in both direc-

tions: more than 60% of fibroblast-specific enhancers are de-

commissioned, and at least 1,000 ESC-specific enhancers are

established de novo within the first 24 hr of reprogramming

factor expression, based on loss or gain of H3K4me1/2,

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1331

respectively (Figure 3). Although H3K4me1/2 on its own does not

allow one to distinguish between active and poised enhancer

states, it is likely that many of the newly marked ESC-specific

enhancers are in a poised state that will be activated at later

stages of reprogramming. Thus, extensive chromatin remodel-

ing at ESC-specific promoters and enhancers precedes the tran-

scriptional activation of many pluripotency genes.

Together, these chromatin dynamics are likely crucial for the

shutdown of the somatic expression program and the transition

toward pluripotency. During differentiation, pluripotency genes

acquire a silent state that is associated with a repressive chro-

matin environment that can include DNA methylation, histone

variants, covalent histone modifications, chromatin regulatory

proteins, and occupancy of regulatory regions by nucleosomes

(Feldman et al., 2006; Mikkelsen et al., 2008; You et al., 2011).

To activate pluripotency genes, it seems that the reprogramming

factors must surmount at least two separable obstacles:

the binding block at upstream regulatory regions (i.e., distal

enhancer and promoter elements) and a block in the transactiva-

tion of the core promoter, which prevents the assembly and acti-

vation of the RNA polymerase-II-containing basal transcription

machinery. Therefore, it may not be too surprising that the acti-

vation of pluripotency genes in reprogramming is relatively slow

and potentially requires a cascade of events. The findings

described above suggest that the formation of poised ESC-

specific enhancers early in reprogramming may be a critical first

step to orchestrate the productive engagement of the core

promoter and transcriptional activation of ESC-specific genes

later in the process when proper signals are available (Taberlay

et al., 2011). This likely requires further chromatin remodeling

and/or additional transcriptional and signaling regulators that

are unavailable early in reprogramming (for more discussion,

see the transition section below). Importantly, this epigenetic

priming does not affect all pluripotency genes early on, as

many only gain an active/poised chromatin signature at their

enhancer and promoter regions late in the process (Polo et al.,

2012; Sridharan et al., 2009) (Figure 3) (see below). Under-

standing the regulation of enhancer/promoter pairs of pluripo-

tency genes during reprogramming will be an important task

for the future and will increase our general knowledge about

the dynamics of promoter and enhancer interactions (Taberlay

et al., 2011).

Relating the extensive binding of OSK to distal sites in

unmarked, closed chromatin early in human cell reprogramming

(Soufi et al., 2012) to the epigenetic priming of many ESC-

specific enhancers early in mouse reprogramming (Koche

et al., 2011) implies that the reprogramming factors may cause

at least some of these initial epigenetic priming events directly.

To test this hypothesis, simultaneous analysis of transcription

factor binding, chromatin, and transcription states is required,

and detailed studies both in vitro and in vivo need to address

whether Oct4, Sox2, and Klf4 can indeed bind regulatory DNA

sites packaged in nucleosomes and change chromatin structure.

The ability of the reprogramming factors to engage regulatory

genomic elements in closed (silent) chromatin may be a critical

feature and may explain why OSK are such potent inducers of

pluripotency and are effective in many different somatic cell

types.

1332 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

DNA Methylation and H3K9 Methylation InfluenceReprogramming Factor BindingGiven that OSK appear to be able to efficiently engage closed

chromatin regions already early in reprogramming, it may be

surprising that many regulatory regions bound by OSKM in the

pluripotent state are not occupied early in the process

(Soufi et al., 2012; Sridharan et al., 2009). What then are the

impediments to reprogramming factor binding and action?

DNA methylation has arisen as an important factor in restricting

early reprogramming events. ESC-specific promoters and

enhancers that gain active chromatin modifications only late in

reprogramming tend to be hypermethylated in the starting fibro-

blasts and become demethylated only late in reprogramming

(Koche et al., 2011) (Figure 3). For example, hypermethylation

of key pluripotency gene promoters, including those of Nanog

and Oct4, is observed until late in reprogramming (Mikkelsen

et al., 2008; Polo et al., 2012), suggesting that demethylation of

these promoters is a rate-limiting step. By contrast, promoters

and enhancers that already gain active chromatin marks

(H3K4me2) early in reprogramming exhibit hypomethylation

throughout the entire reprogramming process (Koche et al.,

2011) (Figure 3). Thus, DNA methylation appears to limit where

histone modification changes can occur. Furthermore, Oct4

expression can establish a nucleosome-depleted region at the

distal enhancers of OCT4 and at the proximal promoter of

NANOG in somatic cells but only if these regions are un-

methylated (You et al., 2011), indicating that DNA methylation

can prevent the recruitment of the reprogramming factors

(Figure 3D). In the case of Oct4, DNA methylation must affect

binding indirectly, as its DNA motif does not contain a CpG.

Jones and colleagues proposed that DNAmethylation in flanking

sequences may stabilize the nucleosome and prevent binding

(You et al., 2011). Similarly, binding of cMyc is inhibited by

CpG methylation within its CACGTG target site (Prendergast

and Ziff, 1991). However, the binding of other transcription

factors, such as the Klf4-related transcription factor SP1, is not

affected by DNA methylation (Harrington et al., 1988), sug-

gesting that the reprogramming factors may be differentially

affected by DNA methylation. Importantly, DNA methylation is

functionally recognized as a feature that limits reprogramming

to pluripotency because interference with Dnmt1, the enzyme

responsible for the maintenance of DNA methylation (Mikkelsen

et al., 2008), promotes iPSC formation (Table 1).

Interestingly, somatic enhancers that are inactivated quickly

upon reprogramming factor expression and are typically methyl-

ated in the pluripotent state only gain hypermethylation later in

the reprogramming process (Koche et al., 2011) (Figure 3).

Thus, both the methylation of somatic genes and the demethyla-

tion of some critical pluripotency genes appear to occur only late

in reprogramming, establishing the DNA methylation pattern

characteristic of the pluripotent state, which is in contrast

to the more gradual changes in histone modifications and

transcriptional states throughout reprogramming (Koche

et al., 2011; Polo et al., 2012). This may explain, at least in

part, why reprogramming intermediates are unstable when the

reprogramming factors are withdrawn, as DNA methylation

may be required to permanently lock in a gene expression

pattern and cell identity (Koche et al., 2011). However, it needs

to be noted that reprogramming occurs normally even upon the

genetic ablation of the de novo DNAmethyltransferases Dnmt3a

and Dnmt3b, indicating that the gain of DNA methylation in

somatic promoters and enhancers may not be essential (Pawlak

and Jaenisch, 2011) (Table 1). In any case, it will be interesting to

elucidate the mechanisms underlying these bidirectional

changes of DNA methylation late in the reprogramming process.

In addition to DNA methylation, other repressive chromatin

marks affect the ability of the reprogramming factors to engage

their target sites. Indeed, Zaret and colleagues uncovered

hundreds of large regions of megabase scale that exclude re-

programming factor binding early in human cell reprogramming

even though the same regions are bound extensively by the

factors in ESCs (Soufi et al., 2012). Although gene-poor, these

regions contain various well-known pluripotency genes such

as NANOG, SOX2, and PRDM14, and almost perfectly overlap

with regions of extended H3K9me3 in the starting fibroblasts

that are in close contact with the nuclear lamina (Soufi

et al., 2012). Importantly, during reprogramming, these broad

H3K9me3 domains are erased, consistent with their absence

in human ESCs (Hawkins et al., 2010; Soufi et al., 2012; Zhu

et al., 2013), raising the possibility that the lack of OSKM binding

in these large contiguous genomic regions early in reprogram-

ming could be caused by the presence of H3K9me3.

There is currently some debate as to whether the H3K9me3

domains arise during lineage specification or are triggered in

differentiated cells in response to specific culture conditions

in vitro (Hawkins et al., 2010; Zhu et al., 2013). Regardless,

the H3K9 methyltransferase SUV39H1 is required for the

maintenance of these H3K9me3 domains, and inhibition of

TGFb signaling lowers the H3K9me3 domain signal (Soufi

et al., 2012; Zhu et al., 2013). Notably, both the suppression of

SUV39H1 and the inhibition of TGFb signaling enhance reprog-

ramming to pluripotency (Ichida et al., 2009; Onder et al., 2012;

Soufi et al., 2012) (Table 1), and inhibition of SUV39H1/2 early

in human cell reprogramming increases the access of OSKM

to sites within H3K9me3 domains (Soufi et al., 2012). Thus,

H3K9 methylation represents a barrier to the induction of plurip-

otency, at least in part, by blocking reprogramming factor

access (Figure 2D). This conclusion is supported further by the

finding that various other H3K9 methyltransferases and H3K9

demethylases control reprogramming efficiency (Chen et al.,

2013; Onder et al., 2012; Soufi et al., 2012) (Table 1). In a fasci-

nating twist, the same regions that display a shift from a broad

H3K9me3 pattern to OSKM binding during reprogramming

encompass nearly all of the 20 hot spots of aberrant epige-

netic reprogramming, which exhibit aberrant DNA methylation

patterns in human iPSCs compared to ESCs (Lister et al.,

2011; Soufi et al., 2012). Thus, the loss of H3K9me3 from these

regions may be a very inefficient process that could additionally

be influenced by the exact culture conditions used for reprog-

ramming (Zhu et al., 2013).

Transitioning between Reprogramming StepsAn important question is what exactly the rate-limiting transition

steps at various reprogramming stages are. How do reprogram-

ming cells transition fromone step to the next? Though the field is

definingmolecules that positively and negatively influence the re-

programming process (Table 1), this question is still very difficult

to address due to the inefficiency of the process. Rate-limiting

transitions are likely linked to fluctuations or inherent noise of

gene expression, chromatin state, and transcription factor bind-

ing and are further influenced by cell-cell contacts or extrinsic

signals. Single-cell gene expression studies have shown that

early reprogramming cultures and intermediate reprogramming

populations both display heterogeneity, with considerable varia-

tion in gene expression between cells (Buganim et al., 2012; Polo

et al., 2012), suggesting that stochastic gene activation events

could be an important contributor to reprogramming transitions.

Some of these expression differences are likely essential for

progression toward pluripotency, whereas others may not have

any impact on the reprogramming process or may even be inhib-

itory (Buganim et al., 2012; Polo et al., 2012).

Oct4 physically interacts with various active and repressive

chromatin complexes (Pardo et al., 2010; van den Berg et al.,

2010), raising the question of whether the activator or repressor

function of Oct4 and the other reprogramming factors is

more important for reprogramming. Recent reports in which re-

programming factors were fused to strong transcriptional activa-

tion domains (TADs) or repressor proteins indicate that activator,

but not repressor, fusions promote reprogramming (Hammachi

et al., 2012; Hirai et al., 2011; Wang et al., 2011c), suggesting

that transcriptional activation is the main action of the reprog-

ramming factors in reprogramming, and may be rate limiting.

However, not all TADs can enhance the induction of pluripo-

tency. TADs of MyoD and VP16, but not those of Mef2C and

Gata4, increase iPSC formation when fused to Oct4 (Hammachi

et al., 2012; Hirai et al., 2011, 2012; Wang et al., 2011c). Because

TADs serve as a scaffold to recruit other transcription factors,

coactivators, and specific chromatin modifiers that are required

for transcriptional activation, these findings suggest the need for

specific coregulatory proteins in pluripotency induction. In addi-

tion, a strong transcriptional activator may bypass the require-

ment for extensive chromatin remodeling at the promoter for

recruitment of the basic transcriptional machinery and preinitia-

tion complex assembly (Koutroubas et al., 2008). Of note, the

ectopic tethering of a strong transcriptional activator (the VP16

TAD) to the silent Oct4 gene in somatic cells is capable of acti-

vating this allele within 48 hr. However, this activation only

happens in a small number of cells, highlighting the need for

additional regulatory events (Hathaway et al., 2012).

Given that the reprogramming factors may act predominantly

as transcriptional activators, it may be surprising that the initial

transcriptional response includes the silencing of the somatic

expression program. However, transcriptional activators could

amplify or induce the expression of other transcriptional activa-

tors as well as repressors, which in turn could secondarily affect

gene expression patterns via emergent feedforward and feed-

back circuitries and could thereby contribute to the cell fate

change of reprogramming. High levels of strong transcription

factors may also contribute indirectly to the repression of other

genes by competing for binding at common sites on the basic

transcriptional machinery in a process referred to as squelching

(Gill and Ptashne, 1988). Additionally, not only coding genes but

also miRNAs are dynamically regulated during reprogramming

and have been implicated in the control of the reprogramming

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1333

Table 1. List of Selected Chromatin Regulators and Their Role in Reprogramming

Chromatin Mark Chromatin Regulator Reprogramming Phenotype References

H3K4me Wdr5 (MLL-HMTase subunit,

H3K4me-binding protein)

required during the initial phase of reprogramming;

interacts with Oct4

Ang et al., 2011

H3K9me Suv39h1/2; Setdb1 (ESET);

Ehmt2 (G9a) (HMTases)

depletion of Suv39h1, Suv39h2, Setdb1, or

Ehmt2 results in efficient conversion of partially

reprogrammed cells to iPSCs in themouse system;

depletion of Suv39h1/2 enhances human cell

reprogramming and allows for more efficient

binding of the reprogramming factors to domains

with broad H3K9me3 in the starting cell

Chen et al., 2013; Onder et al.,

2012; Soufi et al., 2012

Kdm3/4 (demethylases) overexpression enhances reprogramming;

knockdown reduces the conversion of

partially reprogrammed cells to iPSCs

Chen et al., 2013

H3K27me PRC2 (Ezh2, Eed) (HMTase) required for reprogramming Onder et al., 2012; Buganim

et al., 2012

Utx (demethylase) interacts with reprogramming factors; required

for reprogramming; depletion results in aberrant

and inefficient resetting of H3K27me and impairs

reactivation of pluripotency genes; depletion of

Eed rescues the reprogramming defect due

to Utx loss of function

Mansour et al., 2012

H3K36me Jhdm1a (Kdm2a); Jhdm1b

(Kdm2b) (demethylases)

knockdown impairs reprogramming;

overexpression enhances reprogramming

(requiring the demethylase activity) by affecting the

early reprogramming phase; enhances in part by

promoting cell-cycle progression and overcoming

senescence through repression of the Ink4/Arf

locus and/or facilitating the early transcriptional

response to the reprogramming factors

Wang et al., 2011a; Liang

et al., 2012

H3K79me Dot1 (HMTase) depletion in the early phase enhances

reprogramming; inhibition results in more

efficient loss of H3K79me2 from somatic

genes, thereby promoting their downregulation;

depletion allows reprogramming without ectopic

Klf4

Onder et al., 2012

Histone

acetylation

HDACs (histone deacetylases) HDAC2 knockout allows reprogramming to be

driven by the overexpression of only microRNAs;

small-molecule inhibitors of HDACs (such as VPA,

TSA, and butyrate) enhance reprogramming and

replace ectopic cMyc or Klf4

Anokye-Danso et al., 2011;

Huangfu et al., 2008; Mali

et al., 2010; Liang et al., 2010

Chromatin

remodeling

Baf155/Brg1 (ATP-dependent

chromatin-remodeling complex)

overexpression enhances reprogramming;

overexpression appears to enhance binding

of Oct4 to its pluripotency targets during

reprogramming

Singhal et al., 2010

Chd1 essential for reprogramming Gaspar-Maia et al., 2009

Histone variants macroH2A deletion enhances reprogramming to pluripotency,

overexpression prevents efficient reprogramming

of EpiSCs to naıve pluripotent cells; recruited to

regulatory region of pluripotency genes in mouse

embryonic fibroblasts, but not in ESCs

Pasque et al., 2012

DNA methylation Dnmt1 (maintenance

methyltransferase)

depletion enhances reprogramming of fibroblasts

and partially reprogrammed cells, similar to

5-azacytidine (5-AZA) treatment

Mikkelsen et al., 2008

Dnmt3a/b (de novo

methyltransferases)

dispensable for reprogramming Pawlak and Jaenisch, 2011

(Continued on next page)

1334 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Table 1. Continued

Chromatin Mark Chromatin Regulator Reprogramming Phenotype References

Others OGT (O-GlcNAc

glycosyltransferase)

blocking O-GlcNAcylation impairs reprogramming;

O-GlcNAcylation regulates the transactivation

activity of Oct4 and Sox2; O-GlcNAcylation-

defective mutant of Oct4 fails to support

reprogramming

Jang et al., 2012

Parp1 (poly ADP-ribose

polymerase)

enzymatic function and DNA-binding domain are

required for reprogramming; recruited to

pluripotency genes (e.g., Nanog promoter) in the

early phase to control 5meC levels and control

Oct4 recruitment

Doege et al., 2012

Tet2 (FeII and 2-oxoglutarate-

dependent enzyme)

required for efficient reprogramming; required

for the global as well as gene-specific (e.g., at

pluripotency gene promoters) increase in

5-hydroxymethylcytosine (5hmC) mark during

reprogramming

Doege et al., 2012

process, even allowing for the induction of pluripotency without

the ectopic expression of any transcription factor (Anokye-

Danso et al., 2011; Judson et al., 2009). miRNA expression

inversely correlates with target gene expression during reprog-

ramming (Polo et al., 2012), suggesting that miRNAs may be

critically contributing to the silencing of the somatic gene

expression program and subsequent reprogramming steps.

For example, an increase of miR-130 and miR-301 early in re-

programming enhances the process by repressing the develop-

mental regulator Meox2 (Pfaff et al., 2011), and miRNAs of the

miR-200 family are induced early and contribute to

the repression of the fibroblast regulators Zeb1 and Zeb2

(Samavarchi-Tehrani et al., 2010). The experimental depletion

of pre-existing lineage factors also promotes reprogramming

(Hanna et al., 2008) likely by facilitating the decommissioning

of somatic enhancers, thereby enabling the transition to the

next reprogramming stage.

What leads to the hierarchical pluripotency gene activation

late in reprogramming? As discussed before, their efficient

transcription requires the combinatorial and synergistic action

of multiple activators bound to the enhancer and/or distal

promoter. Enhancers can be modular, whereby each transcrip-

tion factor contributes to the transcriptional output, or nonmod-

ular, whereby each transcription factor is essential such that the

target gene is turned on only when all transcription factors are

present. Particularly considering that many ESC-specific

enhancers are bound by a large number of pluripotency tran-

scription factors in ESCs (Figure 2A), the presence of OSKM

alone is likely not sufficient for efficient binding and/or transacti-

vation. One of the factors that needs to act alongside OSK

appears to be the pluripotency transcription factor Nanog.

Nanog co-occupies many pluripotency genes together with

OSK in ESCs and targets promoter regions that fail to bind

OSK until the end of the reprogramming process (Sridharan

et al., 2009) (Figure 2A). Intriguingly, Nanog is essential for the

establishment of iPSCs (Silva et al., 2009) and becomes ex-

pressed before many other pluripotency genes during the

reprogramming process (Golipour et al., 2012), suggesting that

it could be required for their activation. Overexpression of Esrrb,

another pluripotency factor, can rescue OSKM-induced reprog-

ramming in the absence of endogenous Nanog (Festuccia et al.,

2012). Fitting with the concept of hierarchical pluripotency acti-

vation, Esrrb is a direct target of Nanog in ESCs (Festuccia

et al., 2012). Therefore, a critical function of Nanog in reprogram-

ming may be to activate Esrrb, which in turn directly interacts

with the general transcriptional machinery and also co-occupies

many pluripotency loci with OSK and Nanog (Percharde

et al., 2012; van den Berg et al., 2010). Interestingly, a recent

RNAi screen identified various chromatin regulators, including

Morc1, as regulators of the final reprogramming steps, which

have not yet directly been implicated in the maintenance of

pluripotency (Golipour et al., 2012), indicating that in addition

to transcriptional activation an extensive chromatin remodeling

may be required at the late reprogramming stage.

Today, we are just beginning to discover how chromatin limits

but also guides reprogramming factors and how the factors

overcome chromatin barriers. Direct interactions of the reprog-

ramming factors with chromatin regulators may be important.

For example, Oct4 can interact with subunits of the BAF

chromatin-remodeling complex (Pardo et al., 2010; van den

Berg et al., 2010), which enhances reprogramming and could

stimulate the binding of transcription factors to nucleosomal

sites (Singhal et al., 2010; Utley et al., 1997). Similarly, the activity

of the reprogramming factors can be modulated by posttransla-

tional modifications such asO-GlcNAc, which in the case of Oct4

is required for activation of target genes in ESCs and for Oct4’s

full functionality in reprogramming (Jang et al., 2012).

Recent studies have identified additional chromatin regulators

that are essential for the process (for a summary, see Table 1).

For example, the H3K27me demethylase Utx also interacts

with OSK and is critical for the removal of this repressive

H3K37me3 from pluripotency loci (Mansour et al., 2012).

Similarly, decreasing the levels of histone marks associated

with transcriptional elongation promotes the downregulation

of the somatic gene expression program and suppression of

senescence regulators (Liang et al., 2012; Onder et al.,

2012; Wang et al., 2011a). While additional regulatory factors

likely need to function alongside OSKM to allow for binding

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1335

Figure 4. X Chromosome States in Mouse and Human Pluripotent Cells(A) X chromosome inactivation and reactivation cycles in the mouse system, highlighting the association of naive pluripotency with the XaXa state and of primedpluripotency with the XiXa state. Xa, active X chromosome; Xi, inactive X chromosome.(B) Drift and hierarchy of X chromosome states in female human ESCs during long-term culture. Xe, eroded Xi. The box marks the only X chromosome state thatallows de novo X inactivation upon induction of differentiation.(C) Xi reactivation does not occur when female human somatic cells are reprogrammed to primed iPSCs (under bFGF reprogramming conditions). Whilefibroblasts are mosaic for which X is inactivated (Xp, paternal X; Xm, maternal X), each early passage iPSC line carries the X-inactivation state of the differentiatedcell that initiated the reprogramming event. This state is subsequently maintained upon differentiation.(D) As in (B) but for the drift and hierarchy of X chromosome states in female human iPSCs during long-term culture.

to repressed pluripotency genes (Doege et al., 2012), such an

opportunity may normally arise during every cell division, imme-

diately following DNA replication before nucleosome assembly

(Wolffe, 1991). It remains to be determined whether replication

(i.e., cell proliferation) is required for changing gene expression

patterns at every stage of the reprogramming process.

X Chromosome State in Differentiation andReprogramming in the Mouse ModelIn the remaining sections of this Review, we will focus on the

characterization of the induced pluripotent state in both mouse

and human iPSCs, highlighting differences and parallels

between these two cell types particularly as they relate to the

epigenetic state of the X chromosome. In mammals, X chromo-

some inactivation (XCI) leads to the transcriptional silencing of

one X chromosome in female (XX) cells, equalizing gene dosage

to XY males. This epigenetic process has been very powerful in

revealing that the typical reprogramming experiment with human

and mouse cells leads to different developmental states. XCI

1336 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

involves several noncoding RNAs and a dramatic reorganization

of chromatin with various epigenetic layers of regulation such as

DNAmethylation, histone modifications, and late replication in S

phase (reviewed in Wutz [2011]). In the mouse, X chromosome

silencing is established very early in embryonic development,

in the epiblast cells of the implanting blastocyst, which will give

rise to the embryo proper. XCI can therefore be recapitulated

in vitro in differentiating mouse ESCs, the in vivo counterpart of

the epiblast cells of the preimplantation blastocyst. Differentia-

tion induces expression of the noncoding RNA Xist, which then

quickly spreads to coat the chromosome in cis, mediating

silencing of X-linked genes and inducing a repressive

chromatin character along the entire chromosome (Wutz, 2011)

(Figure 4A). This process is random such that the paternally

and maternally inherited X chromosome (Xp and Xm, respec-

tively) become silenced with equal chance. However, in the

mouse system, two states of pluripotency exist in vivo and

in vitro. ESCs and the epiblast cells of the preimplantation

blastocyst represent the naive pluripotent state. By contrast,

primed pluripotent cells are isolated from the epithelialized

epiblast of the postimplantation embryo as mouse epiblast

stem cells (EpiSC) and represent a developmentally advanced

pluripotent state (Brons et al., 2007; Tesar et al., 2007). Conse-

quently, EpiSCs are distinct from ESCs in gene expression,

growth factor dependence, morphology, and the ability to

contribute to blastocyst chimeras, although various core plurip-

otency regulators are present in both mouse ESCs and EpiSCs

and both cell types are capable of multilineage differentiation

in vitro (reviewed in Nichols and Smith [2009]). Importantly,

EpiSCs are post X-inactivation, i.e., are XiXISTXa, mirroring the

state of the epithelialized epiblast in vivo (Pasque et al., 2011)

(Figure 4A). Therefore, in the mouse system, the XaXa state

appears to be a hallmark specifically of naive pluripotency.

Because XCI represents one of the most dramatic events of

facultative heterochromatin formation in mammalian develop-

ment, the question arises of how the somatically silent X chromo-

some is regulated during reprogramming. In the mouse system,

the typical reprogramming experiment establishes naive pluripo-

tency, i.e., iPSCs that are equivalent to leukemia inhibitory factor

(LIF)-dependent, naive ESCs. Our lab demonstrated that female

mouse iPSCs, like female mouse ESCs, carry two active X chro-

mosomes (XaXa), indicating that the Xi is reactivated during

reprogramming to naive pluripotency (Maherali et al., 2007)

(Figure 4A). The activation of genes on the Xi is accompanied

by the loss of all known heterochromatic chromatin marks and

the silencing of Xist (Maherali et al., 2007). Together, these

events enable random X-inactivation upon induction of differen-

tiation, indicating that there is no epigenetic memory for the prior

Xi left behind. Xi reactivation occurs very late in the reprogram-

ming process at around the time of pluripotency gene expression

(Stadtfeld et al., 2008). In contrast to iPSCs, induced EpiSCs

(iEpiSCs) —generated by OSKM expression and culture condi-

tions required for support of the primed pluripotent state

(bFGF/activin) —are XiXISTXa (Han et al., 2011) (Figure 4A).

EpiSCs can be reprogrammed to the ESC-like state with various

transcription factors and a switch in culture environment, estab-

lishing the XaXa state (Nichols and Smith, 2009) (Figure 4A).

Together, these findings establish the X chromosome state as

a sensitive indicator of the developmental state in the mouse

system, both in differentiation and reprogramming processes,

and demonstrate that the XaXa state is indisputably only associ-

ated with the naive state of pluripotency in this system.

X Chromosome Status in Human ESCs and iPSCsThe analysis of human ESCs led to the puzzling observation that

various ESC lines differ in their X chromosome status (Hoffman

et al., 2005; Shen et al., 2008; Silva et al., 2008) (Figure 4B). (1)

They can be XaXa and undergo XCI upon differentiation, compa-

rable to mouse ESCs. (2) Some human ESC lines have already

undergone XCI and display a heterochromatic Xi with XIST

RNA coating the undifferentiated state (XiXISTXa). (3) Many

human ESCs have a silent Xi that lacks XIST expression

(Xiw/oXISTXa). Currently it is thought that newly derived human

ESCs start in the XaXa state and subsequently drift toward XCI

and later loss of XIST RNA with additional time in culture

(Figure 4B). The strongest support for this model comes from

the fact that the XaXa state can be stabilized in newly derived

ESCs under physiological oxygen conditions, whereas chronic

exposure to atmospheric oxygen concentrations irreversibly

induces XCI (Lengner et al., 2010). Regardless of the X chromo-

some state, human ESCs generally share more features with the

primed pluripotent state of the mouse than with mouse ESCs

(Nichols and Smith, 2009). Therefore, the XaXa state is not

restricted to naive pluripotency in the human system and can

also mark the primed pluripotent state. To date, the occurrence

of the XaXa state and the instability of the X have not been

described for mouse EpiSCs and, in fact, for any other cell type.

Given the different states of the X in human ESCs, an inter-

esting question was whether reprogramming of female human

cells to iPSCs, which recapitulate the primed pluripotent state

of human ESCs, would result in Xi reactivation. Originally, our

group demonstrated that female human iPSC lines carry an

XISTRNA-coated Xi (XiXISTXa) when they are first derived (Tchieu

et al., 2010) (Figure 4C). In contrast to somatic cell populations,

which are mosaic with respect to which X chromosome is

inactivated, iPSC lines display a nonrandom pattern of XCI that

is maintained upon induction of differentiation (Tchieu et al.,

2010). As a result, two types of iPSC lines can be derived—those

expressing only the Xp (XmiXpa) and those expressing only the

Xm (XmaXpi) (Tchieu et al., 2010) (Figure 4C). Therefore, reprog-

ramming to human iPSCs does not elicit Xi reactivation, and

iPSCs inherit the Xi of the particular somatic cell in the culture

dish that underwent a successful reprogramming event (Pomp

et al., 2011; Tchieu et al., 2010). Although subsequent reports

confirmed this conclusion (Cheung et al., 2011; Pomp et al.,

2011), other groups obtained conflicting results and argued

that Xi reactivation is prevalent in iPSCs (Kim et al., 2011;

Marchetto et al., 2010).

Recent reports help to reconcile these apparently con-

tradictory conclusions and confirm that the silent state of the X

is faithfully maintained through the reprogramming process but

unravels with the time that iPSCs spend in culture (Anguera

et al., 2012; Mekhoubad et al., 2012; Nazor et al., 2012; Tchieu

et al., 2010; Tomoda et al., 2012). Similar to human ESCs, human

iPSCs are prone to undergo XIST silencing upon prolonged

passaging, yielding Xiw/oXISTXa lines and accordingly losing all

XIST RNA-dependent repressive chromatin marks such as

H3K27me3 (Pomp et al., 2011; Tchieu et al., 2010) (Figure 4D).

Reprogramming experiments with female fibroblasts heterozy-

gous for a mutation of the X-linked gene HPRT combined with

an elegant drug selection system that can distinguish between

the expression of wild-type or mutant HPRT revealed that spon-

taneous loss of XIST RNA coating coincides with re-expression

of the HPRT allele from the Xi (Mekhoubad et al., 2012). Thus,

XiHPRTwtXaHPRTmut iPSCs express only the mutant HPRT allele

at early passage but activate the wild-type HPRT allele upon

XIST RNA loss. Importantly, the activation of Xi-linked genes is

not limited to this one gene but appears to affect the Xi more

broadly, as demonstrated by global expression and DNA meth-

ylation profiles of female iPSC lines (Anguera et al., 2012;

Mekhoubad et al., 2012; Nazor et al., 2012). Specifically, in early

passage XiXISTXa iPSCs, X-linked genes are expressed at the

level of male (XaY) iPSCs and display DNA methylation in

promoters of Xi-linked genes (Mekhoubad et al., 2012; Nazor

et al., 2012). By contrast, higher-passage female iPSCs with no

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1337

XIST RNA (Xiw/oXISTXa) are often characterized by higher expres-

sion of various X-linked genes and hypomethylation of a subset

of Xi-linked promoters, suggesting that the loss of DNA methyl-

ation contributes to the activation of Xi-linked genes.

Importantly, the activation of X-linked genes does not appear

to affect the entire X chromosome. Eggan and colleagues coined

the partial reactivation of the Xi ‘‘erosion of dosage compensa-

tion,’’ yielding an eroded Xi, the Xe (Mekhoubad et al., 2012)

(Figure 4D). Even with long-term culturing, none of the female

human iPSC lines reach the low DNA methylation level along

the entire X that is typical for male iPSCs (with their single Xa),

indicating that even in the worst case the activation of genes

on the Xi is limited in range (Nazor et al., 2012). Across many

female human iPSC lines, the X chromosome is affected to

varying degrees, but the loss of DNA methylation appears to

target similar large, noncontiguous regions of the X chromo-

some, indicating that certain parts of the X can effectively main-

tain proper silencing while others are more prone to reactivation

(Nazor et al., 2012). The patchy erasure of DNA methylation

along the X, along with loss of gene silencing and XIST RNA

coating, cannot be corrected upon differentiation, nor upon

a repeated round of reprogramming (Mekhoubad et al., 2012;

Nazor et al., 2012). Together, these findings are most consistent

with amodel in which reprogramming sustains the XiXISTXa state,

but continued passaging of iPSCs results in XIST silencing

(Xiw/oXISTXa), which then triggers partial reactivation of the Xi

(Xew/oXISTXa) (Figure 4D). Notably, one could argue that these

X-related events are a consequence of continued reprogram-

ming processes, particularly given that continuous passaging

of iPSCs reduces gene expression differences compared to

ESCs (Chin et al., 2010; Polo et al., 2010). However, the erosion

of the X has also recently been observed in many human ESC

lines upon XIST RNA loss, very similar in extent to iPSCs (Nazor

et al., 2012) (Figure 4C). Importantly, iPSCs with an eroded Xi still

depend on FGF/Activin signaling to maintain pluripotency

(Mekhoubad et al., 2012), confirming that the erosion of the X

chromosome occurs in the context of primed pluripotency and

is likely not associated with a change in cell identity to naive plu-

ripotency. Thus, for human pluripotent cells (iPSCs and ESCs),

dosage compensation erosion appears to be a problem of cell

culture, particularly given that it remains a feature of the differen-

tiated progeny, necessitating the development of improved

culturing methods for these cell types (see below).

Why are XIST expression and the silent state of the X unstable

upon long-term culturing? A few relevant observations have

been made. iPSC lines obtained from the same reprogramming

experiment (i.e., the same fibroblast population) typically display

widely different X states at the same passage, with some lines

being able to maintain the XiXISTXa state and others being on

the path of erosion (Anguera et al., 2012; Kim et al., 2011;

Mekhoubad et al., 2012; Nazor et al., 2012; Tchieu et al.,

2010). Similarly, any given iPSC and ESC line can be heteroge-

neous regarding its X chromosome state (Anguera et al., 2012;

Mekhoubad et al., 2012; Silva et al., 2008; Tchieu et al., 2010;

Tomoda et al., 2012). These findings, combined with the fact

that no genomic abnormalities were found in iPSC lines with an

eroded Xi, suggest that epigenetic, but not genetic, changes

are responsible for the instability of the X chromosome (Anguera

1338 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

et al., 2012; Mekhoubad et al., 2012). Consistently, complete

methylation of the XIST promoter correlates with the loss of the

RNA in iPSCs (Tchieu et al., 2010), implying that de novo meth-

ylation contributes to its silencing. Interestingly, in mouse fibro-

blasts, experimentally induced loss of Xist by itself does not

induce the reactivation of candidate X-linked genes (Csan-

kovszki et al., 2001). However, when Xist loss is combined with

the deletion of Dnmt1 and loss of DNA methylation, a dramatic

reactivation of the Xi occurs inmouse somatic cells (Csankovszki

et al., 2001). This parallels what happens when the Xi erodes in

human iPSCs, suggesting that deregulation of the DNA methyl-

ation machinery may directly contribute to this process.

An interesting observation is that the propagation of XiXISTXa

iPSCs in media containing bFGF and IGF2 and on feeder cells

expressing LIF predictably induces XIST RNA loss and activates

genes of the Xi after only a few passages. In this case, silencing is

reinitiated upon differentiation, suggesting that complete Xi

reactivation occurred, establishing an XaXa state in human

iPSCs, rather than an Xe (Tomoda et al., 2012) (Figure 4D).

Based on cell morphology, it appears that these XaXa cells still

maintain the primed pluripotent state (Tomoda et al., 2012). A

somewhat surprising observation is that XIST RNA was not de-

tected at the endpoint of differentiation (Tomoda et al., 2012).

More work will be needed to test whether XIST is upregulated

earlier in the differentiation process, as X inactivation without

XIST expression would be a highly unexpected possibility

(Figure 4D). In any case, this study re-emphasizes that culture

conditions can have a dramatic impact on the epigenetic state

of the X in human iPSCs and enhance transition between X chro-

mosome states.

A comparison of the X states in female human ESCs and iPSCs

highlights two key differences. The XaXa state appears to be the

most ‘‘immature’’ state for primed human ESCs (Lengner et al.,

2010) (Figure 4B, boxed), but it is a downstream state in the

hierarchy of X states in primed human iPSCs (Tomoda et al.,

2012) (Figure 4D, boxed). Hypoxic conditions or the addition of

HDAC inhibitors, which appear to promote the generation and

maintenance of XaXa hESCs (Lengner et al., 2010; Ware et al.,

2009), do not enhance the establishment of XeXa or XaXa iPSCs

(Anguera et al., 2012; Kim et al., 2011; Mekhoubad et al., 2012;

Pomp et al., 2011; Tchieu et al., 2010). One reason for the differ-

ence in X-state hierarchies between human iPSCs and ESCs

may be that the cells are of very different origin—iPSCs are

derived from somatic XiXa cells and ESCs from XaXa cells of

the female human blastocyst (Okamoto et al., 2011). Under-

standing the behavior of the human X in ESCs and iPSCs will

be an important contribution to the ongoing debate about poten-

tial transcriptional, epigenetic, and genetic differences between

various iPSC and ESC lines and their relevance (Lowry, 2012).

It is important to realize that human pluripotent cells that re-

semble the naive, mouse ESC state can be established in vitro

via transcription-factor-induced reprogramming methods. For

example, the overexpression of OCT4 and KLF4 or KLF4 and

KLF2 in primed human ESCs/iPSCs or OSKM in fibroblasts,

combined with specific culture conditions that support the naive

state, allows the establishment of human naive iPSCs (Hanna

et al., 2010). However, the naive state is still relatively difficult

to establish and maintain (Hanna et al., 2010; Pomp et al.,

Figure 5. Effects of X Chromosome Instability on Disease ModelingReprogramming of differentiated cells from females heterozygous for anX-linked mutation results in iPSC lines that express either the mutant or thewild-type allele from the Xa at early passage due to nonrandom X inactivation.These cell lines represent pairs of experimental and control cells ideal formodeling X-linked diseases on an isogenic background. However, upon XISTloss and Xi erosion, the allele from the Xi can become re-expressed, resultingin the loss or modulation of the disease phenotype.

2011; Wang et al., 2011b). When derived from XiXISTXa iPSCs,

naive human pluripotent cells become XIST negative but display

XIST RNA coating in virtually all cells upon differentiation (Hanna

et al., 2010). Despite the fact that the analysis of the X chromo-

some state in naive human cells is still in its infancy, these data

argue strongly that the mouse ESC-like XaXa state, which allows

XIST-dependent induction of X inactivation during differentia-

tion, can be established in human cells upon reprogramming to

the naive state. Naive human pluripotent cells may therefore

represent an excellent model to study the regulation of human

XCI and may get around problems associated with the instability

of the X in primed pluripotent cells. However, the existence of

human naive (mouse ESC-like) pluripotent cells in vivo remains

unclear, and their derivation from preimplantation embryos has

not yet been accomplished (Kuijk et al., 2012; Roode et al.,

2012).

Instability of the Human X, Differentiation, and DiseaseModelingiPSCs can be derived for specific diseases and can differentiate

into any cell type of the human body. Therefore, they offer an

unprecedented opportunity to examine disease states and

develop novel drugs (Onder and Daley, 2012; Trounson et al.,

2012). The nonrandom X inactivation in early passage XiXISTXa

iPSCs has an interesting consequence for the modeling of

X-linkeddiseases.Considering females heterozygous for amuta-

tion in an X-linked gene, iPSCs can be derived that express either

the wild-type or the mutant form of the protein, which represent

an interesting experimental system for the investigation of

disease phenotypes, as both wild-type and mutant cell lines

are on the same genetic background (Tchieu et al., 2010) (Fig-

ure 5). To date, X-linked diseases such as Rett syndrome and

Lesch-Nyhan syndrome (LNS) have been modeled by such

matched iPSCs (Cheung et al., 2011; Kim et al., 2011; Mekhou-

bad et al., 2012). For example, mutations in the X-linked gene

HPRT cause LNS, which leads to behavioral and neurological

symptoms in males but is typically nonsymptomatic in heterozy-

gous females because of random X inactivation (Figure 5). From

these heterozygous females, XiHPRTwtXaHPRTmut iPSCs can be

obtained that, at early passage, exhibit the LNS phenotype

upon differentiation into neurons in vitro, whereas iPSCs with

the opposite X-inactivation pattern (XiHPRTmutXaHPRTwt) behave

normally (Mekhoubad et al., 2012). However, at higher passage,

erosion of the Xi in XiHPRT wtXaHPRTmut iPSCs leads to the expres-

sion of the wild-type HPRT allele and loss of the disease pheno-

type (Mekhoubad et al., 2012) (Figure 5). The interpretation of X-

linked disease studies therefore requires caution and a careful

assessment of the X chromosome state.

Problems caused by the erosion of the Xi in human iPSCs and

ESCs do not only apply to studies of X-linked diseases but

should also be taken seriously for the modeling of autosomal

diseases or, in fact, any differentiation process, as the erosion

of the Xi in long-term culture can also alter the expression of

some autosomal genes in addition to increasing X-linked gene

expression (Anguera et al., 2012). Furthermore, female iPSC

lines without XIST expression grow faster in culture, survive

better in routine culturing, and appear to form only poorly differ-

entiating teratomas, which may be associated with the upre-

gulation of several X-linked oncogenes (Anguera et al., 2012),

indicating that the erosion of the X affects the behavior of female

iPSCs and ESCs more broadly. Importantly, all recent studies

agree that loss of XIST RNA coating is closely associated with

the erosion of the Xi under conventional culture conditions

(Anguera et al., 2012; Mekhoubad et al., 2012; Nazor et al.,

2012; Tchieu et al., 2010; Tomoda et al., 2012). Thus, currently

female human iPSCs with XIST RNA coating should be preferen-

tially used for any downstream application, as these cells are in

the well-defined XiXa state. Accordingly, Lee and colleagues

proposed that XIST RNA coating of the Xi and the accumulation

of XIST-dependent chromatin marks such as H3K27me3 can be

considered biomarkers, as they appear to directly identify the

stable XiXa state (Anguera et al., 2012).

OutlookThe improved mechanistic understanding of the path to pluripo-

tency has already enabled the establishment of non-OSK-

containing reprogramming cocktails (Buganim et al., 2012;

Mansour et al., 2012) and allowed for the replacement of essen-

tial endogenous proteins by downstream targets (Festuccia

et al., 2012). Currently, we are learning only by analyzing a few

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1339

snapshots of the reprogramming process. However, more and

more snapshots will eventually become a continuous epigenetic

movie of the cell fate change that underlies reprogramming to

pluripotency, through which we can virtually watch how the

epigenetic landscape is reset. The 2006 era showcased the

potency of diverse transcription factors in converting cell fates.

It now seems likely that it may eventually be possible to generate

any cell type by forced expression of the appropriate transcrip-

tion factor(s). Continued dissection of the reprogramming

process holds the promise that, at some point in the future, we

will be able to predict exactly which transcription factors are

most potent as reprogramming factors. Finally, other fields

such as tumor biologywill benefit from the insight gained through

reprogramming studies given that, for example, mutations that

prevent senescence have been shown to increase both reprog-

ramming efficiency and tumor development.

ACKNOWLEDGMENTS

Our work is supported by the NIH (DP2OD001686 and P01 GM099134), by the

CIRM (RN1-00564 and RB3-05080), and by the Eli and Edythe Broad Center of

Regenerative Medicine and Stem Cell Research at UCLA. We apologize to all

authors whose work could not be cited due to space limitations. We thank Dr.

Sanjeet Patel for critical reading of the manuscript.

REFERENCES

Adams, C.C., and Workman, J.L. (1995). Binding of disparate transcriptional

activators to nucleosomal DNA is inherently cooperative. Mol. Cell. Biol. 15,

1405–1421.

Ang, Y.S., Tsai, S.Y., Lee, D.F., Monk, J., Su, J., Ratnakumar, K., Ding, J., Ge,

Y., Darr, H., Chang, B., et al. (2011). Wdr5 mediates self-renewal and reprog-

ramming via the embryonic stem cell core transcriptional network. Cell 145,

183–197.

Anguera, M.C., Sadreyev, R., Zhang, Z., Szanto, A., Payer, B., Sheridan, S.D.,

Kwok, S., Haggarty, S.J., Sur, M., Alvarez, J., et al. (2012). Molecular signa-

tures of human induced pluripotent stem cells highlight sex differences and

cancer genes. Cell Stem Cell 11, 75–90.

Anokye-Danso, F., Trivedi, C.M., Juhr, D., Gupta, M., Cui, Z., Tian, Y., Zhang,

Y., Yang, W., Gruber, P.J., Epstein, J.A., and Morrisey, E.E. (2011). Highly effi-

cient miRNA-mediated reprogramming of mouse and human somatic cells to

pluripotency. Cell Stem Cell 8, 376–388.

Beato, M., and Eisfeld, K. (1997). Transcription factor access to chromatin.

Nucleic Acids Res. 25, 3559–3563.

Bergsland, M., Ramskold, D., Zaouter, C., Klum, S., Sandberg, R., and Muhr,

J. (2011). Sequentially acting Sox transcription factors in neural lineage

development. Genes Dev. 25, 2453–2464.

Blau, H.M., Chiu, C.P., and Webster, C. (1983). Cytoplasmic activation of

human nuclear genes in stable heterocaryons. Cell 32, 1171–1180.

Bock, C., Kiskinis, E., Verstappen, G., Gu, H., Boulting, G., Smith, Z.D., Ziller,

M., Croft, G.F., Amoroso, M.W., Oakley, D.H., et al. (2011). Reference Maps of

human ES and iPS cell variation enable high-throughput characterization of

pluripotent cell lines. Cell 144, 439–452.

Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Zucker, J.P.,

Guenther, M.G., Kumar, R.M., Murray, H.L., Jenner, R.G., et al. (2005). Core

transcriptional regulatory circuitry in human embryonic stem cells. Cell 122,

947–956.

Brambrink, T., Foreman, R., Welstead, G.G., Lengner, C.J., Wernig, M., Suh,

H., and Jaenisch, R. (2008). Sequential expression of pluripotency

markers during direct reprogramming of mouse somatic cells. Cell Stem Cell

2, 151–159.

1340 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Brons, I.G., Smithers, L.E., Trotter, M.W., Rugg-Gunn, P., Sun, B., Chuva de

Sousa Lopes, S.M., Howlett, S.K., Clarkson, A., Ahrlund-Richter, L., Pedersen,

R.A., and Vallier, L. (2007). Derivation of pluripotent epiblast stem cells from

mammalian embryos. Nature 448, 191–195.

Buckley, S.M., Aranda-Orgilles, B., Strikoudis, A., Apostolou, E., Loizou, E.,

Moran-Crusio, K., Farnsworth, C.L., Koller, A.A., Dasgupta, R., Silva, J.C.,

et al. (2012). Regulation of pluripotency and cellular reprogramming by the

ubiquitin-proteasome system. Cell Stem Cell 11, 783–798.

Buganim, Y., Faddah, D.A., Cheng, A.W., Itskovich, E., Markoulaki, S., Ganz,

K., Klemm, S.L., van Oudenaarden, A., and Jaenisch, R. (2012). Single-cell

expression analyses during cellular reprogramming reveal an early stochastic

and a late hierarchic phase. Cell 150, 1209–1222.

Carey, B.W., Markoulaki, S., Hanna, J.H., Faddah, D.A., Buganim, Y., Kim, J.,

Ganz, K., Steine, E.J., Cassady, J.P., Creyghton, M.P., et al. (2011). Reprog-

ramming factor stoichiometry influences the epigenetic state and biological

properties of induced pluripotent stem cells. Cell Stem Cell 9, 588–598.

Chen, X., Xu, H., Yuan, P., Fang, F., Huss,M., Vega, V.B.,Wong, E., Orlov, Y.L.,

Zhang, W., Jiang, J., et al. (2008). Integration of external signaling pathways

with the core transcriptional network in embryonic stem cells. Cell 133,

1106–1117.

Chen, J., Liu, J., Chen, Y., Yang, J., Chen, J., Liu, H., Zhao, X.,Mo, K., Song, H.,

Guo, L., et al. (2011). Rational optimization of reprogramming culture condi-

tions for the generation of induced pluripotent stem cells with ultra-high

efficiency and fast kinetics. Cell Res. 21, 884–894.

Chen, J., Liu, H., Liu, J., Qi, J., Wei, B., Yang, J., Liang, H., Chen, Y., Chen, J.,

Wu, Y., et al. (2013). H3K9 methylation is a barrier during somatic cell reprog-

ramming into iPSCs. Nat. Genet. 45, 34–42. Published online December 2,

2012. http://dx.doi.org/10.1038/ng.2491.

Cheung, A.Y., Horvath, L.M., Grafodatskaya, D., Pasceri, P., Weksberg, R.,

Hotta, A., Carrel, L., and Ellis, J. (2011). Isolation of MECP2-null Rett Syndrome

patient hiPS cells and isogenic controls through X-chromosome inactivation.

Hum. Mol. Genet. 20, 2103–2115.

Chin, M.H., Mason, M.J., Xie, W., Volinia, S., Singer, M., Peterson, C., Ambart-

sumyan, G., Aimiuwu, O., Richter, L., Zhang, J., et al. (2009). Induced pluripo-

tent stem cells and embryonic stem cells are distinguished by gene expression

signatures. Cell Stem Cell 5, 111–123.

Chin, M.H., Pellegrini, M., Plath, K., and Lowry, W.E. (2010). Molecular

analyses of human induced pluripotent stem cells and embryonic stem cells.

Cell Stem Cell 7, 263–269.

Creyghton, M.P., Cheng, A.W., Welstead, G.G., Kooistra, T., Carey, B.W.,

Steine, E.J., Hanna, J., Lodato, M.A., Frampton, G.M., Sharp, P.A., et al.

(2010). Histone H3K27ac separates active from poised enhancers and

predicts developmental state. Proc. Natl. Acad. Sci. USA 107, 21931–21936.

Csankovszki, G., Nagy, A., and Jaenisch, R. (2001). Synergism of Xist RNA,

DNA methylation, and histone hypoacetylation in maintaining X chromosome

inactivation. J. Cell Biol. 153, 773–784.

Davis, R.L., Weintraub, H., and Lassar, A.B. (1987). Expression of a single

transfected cDNA converts fibroblasts to myoblasts. Cell 51, 987–1000.

Doege, C.A., Inoue, K., Yamashita, T., Rhee, D.B., Travis, S., Fujita, R., Guar-

nieri, P., Bhagat, G., Vanti, W.B., Shih, A., et al. (2012). Early-stage epigenetic

modification during somatic cell reprogramming by Parp1 and Tet2. Nature

488, 652–655.

Esteban, M.A., Wang, T., Qin, B., Yang, J., Qin, D., Cai, J., Li, W., Weng, Z.,

Chen, J., Ni, S., et al. (2010). Vitamin C enhances the generation of mouse

and human induced pluripotent stem cells. Cell Stem Cell 6, 71–79.

Feldman, N., Gerson, A., Fang, J., Li, E., Zhang, Y., Shinkai, Y., Cedar, H., and

Bergman, Y. (2006). G9a-mediated irreversible epigenetic inactivation of

Oct-3/4 during early embryogenesis. Nat. Cell Biol. 8, 188–194.

Festuccia, N., Osorno, R., Halbritter, F., Karwacki-Neisius, V., Navarro, P.,

Colby, D., Wong, F., Yates, A., Tomlinson, S.R., and Chambers, I. (2012). Esrrb

is a direct Nanog target gene that can substitute for Nanog function in

pluripotent cells. Cell Stem Cell 11, 477–490.

Filion, G.J., van Bemmel, J.G., Braunschweig, U., Talhout, W., Kind, J., Ward,

L.D., Brugman, W., de Castro, I.J., Kerkhoven, R.M., Bussemaker, H.J., and

van Steensel, B. (2010). Systematic protein location mapping reveals five prin-

cipal chromatin types in Drosophila cells. Cell 143, 212–224.

Gaetz, J., Clift, K.L., Fernandes, C.J., Mao, F.F., Lee, J.H., Zhang, L., Baker,

S.W., Looney, T.J., Foshay, K.M., Yu, W.H., et al. (2012). Evidence for a critical

role of gene occlusion in cell fate restriction. Cell Res. 22, 848–858.

Gaspar-Maia, A., Alajem, A., Polesso, F., Sridharan, R., Mason, M.J., Heiders-

bach, A., Ramalho-Santos, J., McManus, M.T., Plath, K., Meshorer, E., and

Ramalho-Santos, M. (2009). Chd1 regulates open chromatin and pluripotency

of embryonic stem cells. Nature 460, 863–868.

Ghisletti, S., Barozzi, I., Mietton, F., Polletti, S., De Santa, F., Venturini, E.,

Gregory, L., Lonie, L., Chew, A., Wei, C.L., et al. (2010). Identification and char-

acterization of enhancers controlling the inflammatory gene expression

program in macrophages. Immunity 32, 317–328.

Gill, G., and Ptashne, M. (1988). Negative effect of the transcriptional activator

GAL4. Nature 334, 721–724.

Golipour, A., David, L., Liu, Y., Jayakumaran, G., Hirsch, C.L., Trcka, D., and

Wrana, J.L. (2012). A late transition in somatic cell reprogramming requires

regulators distinct from the pluripotency network. Cell Stem Cell 11, 769–782.

Green, M.R. (2005). Eukaryotic transcription activation: right on target. Mol.

Cell 18, 399–402.

Gualdi, R., Bossard, P., Zheng, M., Hamada, Y., Coleman, J.R., and Zaret, K.S.

(1996). Hepatic specification of the gut endoderm in vitro: cell signaling and

transcriptional control. Genes Dev. 10, 1670–1682.

Guccione, E., Martinato, F., Finocchiaro, G., Luzi, L., Tizzoni, L., Dall’ Olio, V.,

Zardo, G., Nervi, C., Bernard, L., and Amati, B. (2006). Myc-binding-site recog-

nition in the human genome is determined by chromatin context. Nat. Cell Biol.

8, 764–770.

Gurdon, J.B., Elsdale, T.R., and Fischberg, M. (1958). Sexually mature individ-

uals of Xenopus laevis from the transplantation of single somatic nuclei. Nature

182, 64–65.

Hammachi, F., Morrison, G.M., Sharov, A.A., Livigni, A., Narayan, S., Papape-

trou, E.P., O’Malley, J., Kaji, K., Ko, M.S., Ptashne, M., and Brickman, J.M.

(2012). Transcriptional activation by Oct4 is sufficient for the maintenance

and induction of pluripotency. Cell Rep. 1, 99–109.

Han, D.W., Greber, B., Wu, G., Tapia, N., Arauzo-Bravo, M.J., Ko, K., Berne-

mann, C., Stehling, M., and Scholer, H.R. (2011). Direct reprogramming of

fibroblasts into epiblast stem cells. Nat. Cell Biol. 13, 66–71.

Hanna, J., Markoulaki, S., Schorderet, P., Carey, B.W., Beard, C., Wernig, M.,

Creyghton, M.P., Steine, E.J., Cassady, J.P., Foreman, R., et al. (2008). Direct

reprogramming of terminally differentiated mature B lymphocytes to pluripo-

tency. Cell 133, 250–264.

Hanna, J., Saha, K., Pando, B., van Zon, J., Lengner, C.J., Creyghton, M.P.,

van Oudenaarden, A., and Jaenisch, R. (2009). Direct cell reprogramming is

a stochastic process amenable to acceleration. Nature 462, 595–601.

Hanna, J., Cheng, A.W., Saha, K., Kim, J., Lengner, C.J., Soldner, F., Cassady,

J.P., Muffat, J., Carey, B.W., and Jaenisch, R. (2010). Human embryonic stem

cells with biological and epigenetic characteristics similar to those of mouse

ESCs. Proc. Natl. Acad. Sci. USA 107, 9222–9227.

Harrington, M.A., Jones, P.A., Imagawa, M., and Karin, M. (1988). Cytosine

methylation does not affect binding of transcription factor Sp1. Proc. Natl.

Acad. Sci. USA 85, 2066–2070.

Hathaway, N.A., Bell, O., Hodges, C., Miller, E.L., Neel, D.S., and Crabtree,

G.R. (2012). Dynamics and memory of heterochromatin in living cells. Cell

149, 1447–1460.

Hawkins, R.D., Hon, G.C., Lee, L.K., Ngo, Q., Lister, R., Pelizzola, M., Edsall,

L.E., Kuan, S., Luu, Y., Klugman, S., et al. (2010). Distinct epigenomic land-

scapes of pluripotent and lineage-committed human cells. Cell Stem Cell 6,

479–491.

Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P., Stark, A., Harp,

L.F., Ye, Z., Lee, L.K., Stuart, R.K., Ching, C.W., et al. (2009). Histone modifi-

cations at human enhancers reflect global cell-type-specific gene expression.

Nature 459, 108–112.

Hirai, H., Tani, T., Katoku-Kikyo, N., Kellner, S., Karian, P., Firpo,M., and Kikyo,

N. (2011). Radical acceleration of nuclear reprogramming by chromatin re-

modeling with the transactivation domain of MyoD. StemCells 29, 1349–1361.

Hirai, H., Katoku-Kikyo, N., Karian, P., Firpo, M., and Kikyo, N. (2012). Efficient

iPS cell production with the MyoD transactivation domain in serum-free

culture. PLoS ONE 7, e34149.

Hoffman, L.M., Hall, L., Batten, J.L., Young, H., Pardasani, D., Baetge, E.E.,

Lawrence, J., and Carpenter, M.K. (2005). X-inactivation status varies in

human embryonic stem cell lines. Stem Cells 23, 1468–1478.

Huangfu, D., Maehr, R., Guo, W., Eijkelenboom, A., Snitow, M., Chen, A.E.,

and Melton, D.A. (2008). Induction of pluripotent stem cells by defined

factors is greatly improved by small-molecule compounds. Nat. Biotechnol.

26, 795–797.

Ichida, J.K., Blanchard, J., Lam, K., Son, E.Y., Chung, J.E., Egli, D., Loh, K.M.,

Carter, A.C., Di Giorgio, F.P., Koszka, K., et al. (2009). A small-molecule inhib-

itor of tgf-Beta signaling replaces sox2 in reprogramming by inducing nanog.

Cell Stem Cell 5, 491–503.

Jang, H., Kim, T.W., Yoon, S., Choi, S.Y., Kang, T.W., Kim, S.Y., Kwon, Y.W.,

Cho, E.J., and Youn, H.D. (2012). O-GlcNAc regulates pluripotency and

reprogramming by directly acting on core components of the pluripotency

network. Cell Stem Cell 11, 62–74.

Jiang, J., Chan, Y.S., Loh, Y.H., Cai, J., Tong, G.Q., Lim, C.A., Robson, P.,

Zhong, S., and Ng, H.H. (2008). A core Klf circuitry regulates self-renewal of

embryonic stem cells. Nat. Cell Biol. 10, 353–360.

Judson, R.L., Babiarz, J.E., Venere, M., and Blelloch, R. (2009). Embryonic

stem cell-specific microRNAs promote induced pluripotency. Nat. Biotechnol.

27, 459–461.

Kim, J., Chu, J., Shen, X., Wang, J., and Orkin, S.H. (2008). An extended

transcriptional network for pluripotency of embryonic stem cells. Cell 132,

1049–1061.

Kim, J., Woo, A.J., Chu, J., Snow, J.W., Fujiwara, Y., Kim, C.G., Cantor, A.B.,

and Orkin, S.H. (2010). A Myc network accounts for similarities between

embryonic stem and cancer cell transcription programs. Cell 143, 313–324.

Kim, K.Y., Hysolli, E., and Park, I.H. (2011). Neuronal maturation defect in

induced pluripotent stem cells from patients with Rett syndrome. Proc. Natl.

Acad. Sci. USA 108, 14169–14174.

Koch, P., Breuer, P., Peitz, M., Jungverdorben, J., Kesavan, J., Poppe, D.,

Doerr, J., Ladewig, J., Mertens, J., Tuting, T., et al. (2011). Excitation-induced

ataxin-3 aggregation in neurons from patients with Machado-Joseph disease.

Nature 480, 543–546.

Koche, R.P., Smith, Z.D., Adli, M., Gu, H., Ku, M., Gnirke, A., Bernstein, B.E.,

and Meissner, A. (2011). Reprogramming factor expression initiates wide-

spread targeted chromatin remodeling. Cell Stem Cell 8, 96–105.

Koutroubas, G., Merika, M., and Thanos, D. (2008). Bypassing the require-

ments for epigenetic modifications in gene transcription by increasing

enhancer strength. Mol. Cell. Biol. 28, 926–938.

Kuijk, E.W., van Tol, L.T., Van de Velde, H., Wubbolts, R., Welling, M., Geijsen,

N., and Roelen, B.A. (2012). The roles of FGF and MAP kinase signaling in the

segregation of the epiblast and hypoblast cell lineages in bovine and human

embryos. Development 139, 871–882.

Lengner, C.J., Gimelbrant, A.A., Erwin, J.A., Cheng, A.W., Guenther, M.G.,

Welstead, G.G., Alagappan, R., Frampton, G.M., Xu, P., Muffat, J., et al.

(2010). Derivation of pre-X inactivation human embryonic stem cells under

physiological oxygen concentrations. Cell 141, 872–883.

Li, B., Adams, C.C., and Workman, J.L. (1994). Nucleosome binding by the

constitutive transcription factor Sp1. J. Biol. Chem. 269, 7756–7763.

Li, B., Carey, M., and Workman, J.L. (2007). The role of chromatin during

transcription. Cell 128, 707–719.

Li, R., Liang, J., Ni, S., Zhou, T., Qing, X., Li, H., He,W., Chen, J., Li, F., Zhuang,

Q., et al. (2010). Amesenchymal-to-epithelial transition initiates and is required

for the nuclear reprogramming of mouse fibroblasts. Cell Stem Cell 7, 51–63.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1341

Liang, G., Taranova, O., Xia, K., and Zhang, Y. (2010). Butyrate promotes

induced pluripotent stem cell generation. J. Biol. Chem. 285, 25516–25521.

Liang, G., He, J., and Zhang, Y. (2012). Kdm2b promotes induced pluripotent

stem cell generation by facilitating gene activation early in reprogramming.

Nat. Cell Biol. 14, 457–466.

Lin, C.Y., Loven, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Bradner, J.E., Lee,

T.I., and Young, R.A. (2012). Transcriptional amplification in tumor cells with

elevated c-Myc. Cell 151, 56–67.

Lister, R., Pelizzola, M., Kida, Y.S., Hawkins, R.D., Nery, J.R., Hon, G.,

Antosiewicz-Bourget, J., O’Malley, R., Castanon, R., Klugman, S., et al.

(2011). Hotspots of aberrant epigenomic reprogramming in human induced

pluripotent stem cells. Nature 471, 68–73.

Loh, Y.H., Wu, Q., Chew, J.L., Vega, V.B., Zhang, W., Chen, X., Bourque, G.,

George, J., Leong, B., Liu, J., et al. (2006). The Oct4 and Nanog transcription

network regulates pluripotency in mouse embryonic stem cells. Nat. Genet.

38, 431–440.

Lorch, Y., LaPointe, J.W., and Kornberg, R.D. (1987). Nucleosomes inhibit the

initiation of transcription but allow chain elongation with the displacement of

histones. Cell 49, 203–210.

Lowry, W.E. (2012). Does transcription factor induced pluripotency accurately

mimic embryo derived pluripotency? Curr. Opin. Genet. Dev. 22, 429–434.

Maherali, N., Sridharan, R., Xie, W., Utikal, J., Eminli, S., Arnold, K., Stadtfeld,

M., Yachechko, R., Tchieu, J., Jaenisch, R., et al. (2007). Directly reprog-

rammed fibroblasts show global epigenetic remodeling and widespread tissue

contribution. Cell Stem Cell 1, 55–70.

Mali, P., Chou, B.K., Yen, J., Ye, Z., Zou, J., Dowey, S., Brodsky, R.A., Ohm,

J.E., Yu, W., Baylin, S.B., et al. (2010). Butyrate greatly enhances derivation

of human induced pluripotent stem cells by promoting epigenetic re-

modeling and the expression of pluripotency-associated genes. Stem Cells

28, 713–720.

Mansour, A.A., Gafni, O., Weinberger, L., Zviran, A., Ayyash, M., Rais, Y.,

Krupalnik, V., Zerbib, M., Amann-Zalcenstein, D., Maza, I., et al. (2012). The

H3K27 demethylase Utx regulates somatic and germ cell epigenetic reprog-

ramming. Nature 488, 409–413.

Marchetto, M.C., Carromeu, C., Acab, A., Yu, D., Yeo, G.W., Mu, Y., Chen, G.,

Gage, F.H., andMuotri, A.R. (2010). Amodel for neural development and treat-

ment of Rett syndrome using human induced pluripotent stem cells. Cell 143,

527–539.

Marks, H., Kalkan, T., Menafra, R., Denissov, S., Jones, K., Hofemeister, H.,

Nichols, J., Kranz, A., Stewart, A.F., Smith, A., and Stunnenberg, H.G.

(2012). The transcriptional and epigenomic foundations of ground state plurip-

otency. Cell 149, 590–604.

Mekhoubad, S., Bock, C., de Boer, A.S., Kiskinis, E., Meissner, A., and Eggan,

K. (2012). Erosion of dosage compensation impacts human iPSC disease

modeling. Cell Stem Cell 10, 595–609.

Mikkelsen, T.S., Hanna, J., Zhang, X., Ku, M., Wernig, M., Schorderet, P.,

Bernstein, B.E., Jaenisch, R., Lander, E.S., andMeissner, A. (2008). Dissecting

direct reprogramming through integrative genomic analysis. Nature 454,

49–55.

Nagamatsu, G., Saito, S., Kosaka, T., Takubo, K., Kinoshita, T., Oya, M.,

Horimoto, K., and Suda, T. (2012). Optimal ratio of transcription factors for

somatic cell reprogramming. J. Biol. Chem. 287, 36273–36282.

Nakagawa, M., Koyanagi, M., Tanabe, K., Takahashi, K., Ichisaka, T., Aoi, T.,

Okita, K., Mochiduki, Y., Takizawa, N., and Yamanaka, S. (2008). Generation of

induced pluripotent stem cells without Myc from mouse and human fibro-

blasts. Nat. Biotechnol. 26, 101–106.

Nazor, K.L., Altun, G., Lynch, C., Tran, H., Harness, J.V., Slavin, I., Garitaonan-

dia, I., Muller, F.J., Wang, Y.C., Boscolo, F.S., et al. (2012). Recurrent

variations in DNA methylation in human pluripotent stem cells and their differ-

entiated derivatives. Cell Stem Cell 10, 620–634.

Nichols, J., and Smith, A. (2009). Naive and primed pluripotent states. Cell

Stem Cell 4, 487–492.

1342 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Nie, Z., Hu, G., Wei, G., Cui, K., Yamane, A., Resch, W., Wang, R., Green, D.R.,

Tessarollo, L., Casellas, R., et al. (2012). c-Myc is a universal amplifier of

expressed genes in lymphocytes and embryonic stem cells. Cell 151, 68–79.

Niwa, H., Miyazaki, J., and Smith, A.G. (2000). Quantitative expression of Oct-

3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nat.

Genet. 24, 372–376.

Okamoto, I., Patrat, C., Thepot, D., Peynot, N., Fauque, P., Daniel, N., Diaban-

gouaya, P., Wolf, J.P., Renard, J.P., Duranthon, V., and Heard, E. (2011).

Eutherian mammals use diverse strategies to initiate X-chromosome inactiva-

tion during development. Nature 472, 370–374.

Okita, K., Ichisaka, T., and Yamanaka, S. (2007). Generation of germline-

competent induced pluripotent stem cells. Nature 448, 313–317.

Onder, T.T., and Daley, G.Q. (2012). New lessons learned from disease

modeling with induced pluripotent stem cells. Curr. Opin. Genet. Dev. 22,

500–508.

Onder, T.T., Kara, N., Cherry, A., Sinha, A.U., Zhu, N., Bernt, K.M., Cahan, P.,

Marcarci, B.O., Unternaehrer, J., Gupta, P.B., et al. (2012). Chromatin-

modifying enzymes as modulators of reprogramming. Nature 483, 598–602.

Pardo, M., Lang, B., Yu, L., Prosser, H., Bradley, A., Babu, M.M., and

Choudhary, J. (2010). An expanded Oct4 interaction network: implications

for stem cell biology, development, and disease. Cell Stem Cell 6, 382–395.

Pasque, V., Gillich, A., Garrett, N., and Gurdon, J.B. (2011). Histone variant

macroH2A confers resistance to nuclear reprogramming. EMBO J. 30,

2373–2387.

Pasque, V., Radzisheuskaya, A., Gillich, A., Halley-Stott, R.P., Panamarova,

M., Zernicka-Goetz, M., Surani, M.A., and Silva, J.C. (2012). Histone variant

macroH2A marks embryonic differentiation in vivo and acts as an epigenetic

barrier to induced pluripotency. J. Cell Sci. Published October 17, 2012.

http://dx.doi.org/10.1242/jcs.113019.

Pawlak, M., and Jaenisch, R. (2011). De novo DNAmethylation by Dnmt3a and

Dnmt3b is dispensable for nuclear reprogramming of somatic cells to a plurip-

otent state. Genes Dev. 25, 1035–1040.

Percharde, M., Lavial, F., Ng, J.H., Kumar, V., Tomaz, R.A., Martin, N., Yeo,

J.C., Gil, J., Prabhakar, S., Ng, H.H., et al. (2012). Ncoa3 functions as an essen-

tial Esrrb coactivator to sustain embryonic stem cell self-renewal and reprog-

ramming. Genes Dev. 26, 2286–2298.

Pfaff, N., Fiedler, J., Holzmann, A., Schambach, A., Moritz, T., Cantz, T., and

Thum, T. (2011). miRNA screening reveals a new miRNA family stimulating

iPS cell generation via regulation of Meox2. EMBO Rep. 12, 1153–1159.

Plath, K., and Lowry, W.E. (2011). Progress in understanding reprogramming

to the induced pluripotent state. Nat. Rev. Genet. 12, 253–265.

Polo, J.M., Liu, S., Figueroa, M.E., Kulalert, W., Eminli, S., Tan, K.Y., Aposto-

lou, E., Stadtfeld, M., Li, Y., Shioda, T., et al. (2010). Cell type of origin influ-

ences the molecular and functional properties of mouse induced pluripotent

stem cells. Nat. Biotechnol. 28, 848–855.

Polo, J.M., Anderssen, E., Walsh, R.M., Schwarz, B.A., Nefzger, C.M., Lim,

S.M., Borkent, M., Apostolou, E., Alaei, S., Cloutier, J., et al. (2012). A molec-

ular roadmap of reprogramming somatic cells into iPS cells. Cell 151, 1617–

1632.

Pomp, O., Dreesen, O., Leong, D.F., Meller-Pomp, O., Tan, T.T., Zhou, F., and

Colman, A. (2011). Unexpected X chromosome skewing during culture and re-

programming of human somatic cells can be alleviated by exogenous telome-

rase. Cell Stem Cell 9, 156–165.

Prendergast, G.C., and Ziff, E.B. (1991). Methylation-sensitive sequence-

specific DNA binding by the c-Myc basic region. Science 251, 186–189.

Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S.A., Flynn, R.A., and

Wysocka, J. (2011). A unique chromatin signature uncovers early develop-

mental enhancers in humans. Nature 470, 279–283.

Rahl, P.B., Lin, C.Y., Seila, A.C., Flynn, R.A., McCuine, S., Burge, C.B., Sharp,

P.A., and Young, R.A. (2010). c-Myc regulates transcriptional pause release.

Cell 141, 432–445.

Ramirez-Carrozzi, V.R., Braas, D., Bhatt, D.M., Cheng, C.S., Hong, C., Doty,

K.R., Black, J.C., Hoffmann, A., Carey, M., and Smale, S.T. (2009). A unifying

model for the selective regulation of inducible transcription by CpG islands and

nucleosome remodeling. Cell 138, 114–128.

Roode,M., Blair, K., Snell, P., Elder, K., Marchant, S., Smith, A., and Nichols, J.

(2012). Human hypoblast formation is not dependent on FGF signalling. Dev.

Biol. 361, 358–363.

Samavarchi-Tehrani, P., Golipour, A., David, L., Sung, H.K., Beyer, T.A., Datti,

A., Woltjen, K., Nagy, A., and Wrana, J.L. (2010). Functional genomics reveals

a BMP-driven mesenchymal-to-epithelial transition in the initiation of somatic

cell reprogramming. Cell Stem Cell 7, 64–77.

Sancho-Martinez, I., Baek, S.H., and Izpisua Belmonte, J.C. (2012). Lineage

conversion methodologies meet the reprogramming toolbox. Nat. Cell Biol.

14, 892–899.

Shen, Y., Matsuno, Y., Fouse, S.D., Rao, N., Root, S., Xu, R., Pellegrini, M.,

Riggs, A.D., and Fan, G. (2008). X-inactivation in female human embryonic

stem cells is in a nonrandom pattern and prone to epigenetic alterations.

Proc. Natl. Acad. Sci. USA 105, 4709–4714.

Silva, S.S., Rowntree, R.K., Mekhoubad, S., and Lee, J.T. (2008). X-chromo-

some inactivation and epigenetic fluidity in human embryonic stem cells.

Proc. Natl. Acad. Sci. USA 105, 4820–4825.

Silva, J., Nichols, J., Theunissen, T.W., Guo, G., van Oosten, A.L., Barrandon,

O., Wray, J., Yamanaka, S., Chambers, I., and Smith, A. (2009). Nanog is the

gateway to the pluripotent ground state. Cell 138, 722–737.

Singhal, N., Graumann, J., Wu, G., Arauzo-Bravo, M.J., Han, D.W., Greber, B.,

Gentile, L., Mann, M., and Scholer, H.R. (2010). Chromatin-Remodeling

Components of the BAF Complex Facilitate Reprogramming. Cell 141,

943–955.

Smith, Z.D., Nachman, I., Regev, A., and Meissner, A. (2010). Dynamic single-

cell imaging of direct reprogramming reveals an early specifying event. Nat.

Biotechnol. 28, 521–526.

Soufi, A., Donahue, G., and Zaret, K.S. (2012). Facilitators and impediments of

the pluripotency reprogramming factors’ initial engagement with the genome.

Cell 151, 994–1004.

Sridharan, R., Tchieu, J., Mason, M.J., Yachechko, R., Kuoy, E., Horvath, S.,

Zhou, Q., and Plath, K. (2009). Role of the murine reprogramming factors in

the induction of pluripotency. Cell 136, 364–377.

Stadtfeld, M., and Hochedlinger, K. (2010). Induced pluripotency: history,

mechanisms, and applications. Genes Dev. 24, 2239–2263.

Stadtfeld, M., Maherali, N., Breault, D.T., and Hochedlinger, K. (2008). Defining

molecular cornerstones during fibroblast to iPS cell reprogramming in mouse.

Cell Stem Cell 2, 230–240.

Stadtfeld, M., Apostolou, E., Ferrari, F., Choi, J., Walsh, R.M., Chen, T., Ooi,

S.S., Kim, S.Y., Bestor, T.H., Shioda, T., et al. (2012). Ascorbic acid prevents

loss of Dlk1-Dio3 imprinting and facilitates generation of all-iPS cell mice

from terminally differentiated B cells. Nat. Genet. 44, 398–405.

Taberlay, P.C., Kelly, T.K., Liu, C.C., You, J.S., De Carvalho, D.D., Miranda,

T.B., Zhou, X.J., Liang, G., and Jones, P.A. (2011). Polycomb-repressed genes

have permissive enhancers that initiate reprogramming. Cell 147, 1283–1294.

Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells

from mouse embryonic and adult fibroblast cultures by defined factors. Cell

126, 663–676.

Tchieu, J., Kuoy, E., Chin, M.H., Trinh, H., Patterson, M., Sherman, S.P.,

Aimiuwu, O., Lindgren, A., Hakimian, S., Zack, J.A., et al. (2010). Female

human iPSCs retain an inactive X chromosome. Cell Stem Cell 7, 329–342.

Teif, V.B., Vainshtein, Y., Caudron-Herger,M., Mallm, J.P., Marth, C., Hofer, T.,

andRippe, K. (2012). Genome-wide nucleosome positioning during embryonic

stem cell development. Nat. Struct. Mol. Biol. 19, 1185–1192.

Tesar, P.J., Chenoweth, J.G., Brook, F.A., Davies, T.J., Evans, E.P., Mack,

D.L., Gardner, R.L., and McKay, R.D. (2007). New cell lines from mouse

epiblast share defining features with human embryonic stem cells. Nature

448, 196–199.

Tiemann, U., Sgodda, M., Warlich, E., Ballmaier, M., Scholer, H.R., Scham-

bach, A., and Cantz, T. (2011). Optimal reprogramming factor stoichiometry

increases colony numbers and affects molecular characteristics of murine

induced pluripotent stem cells. Cytometry A 79, 426–435.

Tomoda, K., Takahashi, K., Leung, K., Okada, A., Narita, M., Yamada, N.A., Ei-

lertson, K.E., Tsang, P., Baba, S., White, M.P., et al. (2012). Derivation condi-

tions impact X-inactivation status in female human induced pluripotent stem

cells. Cell Stem Cell 11, 91–99.

Trounson, A., Shepard, K.A., and DeWitt, N.D. (2012). Human disease

modeling with induced pluripotent stem cells. Curr. Opin. Genet. Dev. 22,

509–516.

Utley, R.T., Cote, J., Owen-Hughes, T., and Workman, J.L. (1997). SWI/SNF

stimulates the formation of disparate activator-nucleosome complexes

but is partially redundant with cooperative binding. J. Biol. Chem. 272,

12642–12649.

van den Berg, D.L., Snoek, T., Mullin, N.P., Yates, A., Bezstarosti, K.,

Demmers, J., Chambers, I., and Poot, R.A. (2010). An Oct4-centered protein

interaction network in embryonic stem cells. Cell Stem Cell 6, 369–381.

Visel, A., Blow, M.J., Li, Z., Zhang, T., Akiyama, J.A., Holt, A., Plajzer-Frick, I.,

Shoukry, M., Wright, C., Chen, F., et al. (2009). ChIP-seq accurately predicts

tissue-specific activity of enhancers. Nature 457, 854–858.

Wallrath, L.L., Lu, Q., Granok, H., and Elgin, S.C. (1994). Architectural

variations of inducible eukaryotic promoters: preset and remodeling chromatin

structures. Bioessays 16, 165–170.

Wang, T., Chen, K., Zeng, X., Yang, J., Wu, Y., Shi, X., Qin, B., Zeng, L.,

Esteban, M.A., Pan, G., and Pei, D. (2011a). The histone demethylases

Jhdm1a/1b enhance somatic cell reprogramming in a vitamin-C-dependent

manner. Cell Stem Cell 9, 575–587.

Wang, W., Yang, J., Liu, H., Lu, D., Chen, X., Zenonos, Z., Campos, L.S., Rad,

R., Guo, G., Zhang, S., et al. (2011b). Rapid and efficient reprogramming

of somatic cells to induced pluripotent stem cells by retinoic acid re-

ceptor gamma and liver receptor homolog 1. Proc. Natl. Acad. Sci. USA

108, 18283–18288.

Wang, Y., Chen, J., Hu, J.L., Wei, X.X., Qin, D., Gao, J., Zhang, L., Jiang, J., Li,

J.S., Liu, J., et al. (2011c). Reprogramming of mouse and human somatic cells

by high-performance engineered factors. EMBO Rep. 12, 373–378.

Ware, C.B., Wang, L., Mecham, B.H., Shen, L., Nelson, A.M., Bar, M., Lamba,

D.A., Dauphin, D.S., Buckingham, B., Askari, B., et al. (2009). Histone deace-

tylase inhibition elicits an evolutionarily conserved self-renewal program in

embryonic stem cells. Cell Stem Cell 4, 359–369.

Wernig, M., Meissner, A., Foreman, R., Brambrink, T., Ku, M., Hochedlinger,

K., Bernstein, B.E., and Jaenisch, R. (2007). In vitro reprogramming of fibro-

blasts into a pluripotent ES-cell-like state. Nature 448, 318–324.

Wernig, M., Meissner, A., Cassady, J.P., and Jaenisch, R. (2008). c-Myc is

dispensable for direct reprogramming of mouse fibroblasts. Cell Stem Cell

2, 10–12.

Wolffe, A.P. (1991). Implications of DNA replication for eukaryotic gene expres-

sion. J. Cell Sci. 99, 201–206.

Wutz, A. (2011). Gene silencing in X-chromosome inactivation: advances in

understanding facultative heterochromatin formation. Nat. Rev. Genet. 12,

542–553.

Yamaguchi, S., Hirano, K., Nagata, S., and Tada, T. (2011). Sox2 expression

effects on direct reprogramming efficiency as determined by alternative

somatic cell fate. Stem Cell Res. (Amst.) 6, 177–186.

You, J.S., Kelly, T.K., De Carvalho, D.D., Taberlay, P.C., Liang, G., and Jones,

P.A. (2011). OCT4 establishes and maintains nucleosome-depleted regions

that provide additional layers of epigenetic regulation of its target genes.

Proc. Natl. Acad. Sci. USA 108, 14497–14502.

Zaret, K.S., and Carroll, J.S. (2011). Pioneer transcription factors: establishing

competence for gene expression. Genes Dev. 25, 2227–2241.

Zhu, J., Adli, M., Zou, J.Y., Verstappen, G., Coyne, M., Zhang, X., Durham, T.,

Miri, M., Deshpande, V., De Jager, P.L., et al. (2013). Genome-wide Chromatin

State Transitions Associated with Developmental and Environmental Cues.

Cell 152, 642–654.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1343

Leading Edge

Review

Chromatin Remodelingat DNA Double-Strand Breaks

Brendan D. Price1 and Alan D. D’Andrea1,*1Division of Genomic Stability and DNA Repair, Department of Radiation Oncology, Dana-Farber Cancer Institute, Harvard Medical School,

450 Brookline Avenue, Boston, MA 02215, USA*Correspondence: alan_dandrea@dfci.harvard.edu

http://dx.doi.org/10.1016/j.cell.2013.02.011

DNA double-strand breaks (DSBs) can arise from multiple sources, including exposure to ionizingradiation. The repair of DSBs involves both posttranslational modification of nucleosomes andconcentration of DNA-repair proteins at the site of damage. Consequently, nucleosome packingand chromatin architecture surrounding the DSBmay limit the ability of the DNA-damage responseto access and repair the break. Here, we review early chromatin-based events that promote theformation of open, relaxed chromatin structures at DSBs and that allow the DNA-repair machineryto access the spatially confined region surrounding the DSB, thereby facilitating mammalian DSBrepair.

DNA Double-Strand Breaks and CancerMaintaining the integrity of genetic information is critical both

for normal cellular functions and for suppressing mutagenic

events that can lead to cancer. Damage to DNA can arise

from external sources, such as exposure to ionizing radiation

(IR), ultraviolet radiation (UV), or environmental toxins, or from

endogenous sources, such as reactive oxygen species or

errors during DNA replication. These events can generate

a wide range of DNA lesions, including modified bases or sugar

residues, the formation of DNA adducts, crosslinking of the

DNA strands, and production of single- and double-strand

breaks (DSBs). Consequently, cells have evolved at least six

different DNA-repair pathways to deal with these distinct types

of DNA damage (Kennedy and D’Andrea, 2006). Among these

lesions, DNA DSBs are particularly lethal because they result

in physical cleavage of the DNA backbone. DSBs can occur

through replication-fork collapse, during the processing of

interstrand crosslinks, or following exposure to IR (Ciccia and

Elledge, 2010; Jackson and Bartek, 2009; Kennedy and D’An-

drea, 2006). Because IR (radiation therapy) is widely used to

treat cancer, understanding how cells repair DSBs created

by IR, and how this process is altered in tumors, is of high

significance.

Chromatin Structure and DSB RepairDSB repair takes place within the complex organization of the

chromatin, and it is clear from work in many model systems

that chromatin structure and nucleosome organization represent

a significant barrier to the efficient detection and repair of DSBs.

Mammalian cells contain a diverse array of specialized chro-

matin structures, such as active genes, telomeres, replication

forks, intergenic regions, and compact heterochromatin. These

structures are distinguished by specific patterns of histone

modifications, unique histone variants, arrays of chromatin-

binding proteins, and the density of nucleosome packing (de

1344 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Wit and van Steensel, 2009; Grewal and Jia, 2007; Peng and

Karpen, 2008). This complexity and diversity in chromatin orga-

nization present a series of challenges to the DSB-repair

machinery. The impact of chromatin on DNA repair was first

described in the ‘‘access-repair-restore’’ model (Smerdon,

1991; reviewed in Soria et al., 2012). This model proposed the

minimal steps needed to reorganize the chromatin and repair

DNA damage. Broadly, the DSB-repair machinery must be

able to (1) detect DNA damage in different chromatin structures;

(2) remodel the local chromatin architecture to provide access to

the site of damage; (3) reorganize the nucleosome-DNA template

for processing and repair of the damage; and, importantly, (4)

restore the local chromatin organization after repair has been

completed. Since this model was first put forward in 1991, we

now know many of the remodeling factors and histone-modi-

fying enzymes that act to create open chromatin structures

and promote DNA repair, as well as factors such as histone

chaperones, deacetylases, and phosphatases that reassemble

the chromatin after repair is complete. Here, we will focus on

the ‘‘access’’ component of the ‘‘access-repair-restore’’ model,

reviewing some of the early (seconds–minutes) remodeling

events that occur after DNA damage and that are required to

create open chromatin structures. Although the ‘‘access-

repair-restore’’ model is likely applicable to the repair of all

types of DNA damage, we will focus our discussion specifically

on the repair of DNA DSBs. In particular, we will examine three

broad chromatin-based events that occur during the first

seconds-to-minutes after production of DSBs: (1) the formation

of open chromatin structures at DSBs through acetylation of

histone H4; (2) the importance of kap-1 in promoting chromatin

relaxation in heterochromatin; and (3) the rapid polyADP-ribosy-

lation (PARylation) of the chromatin by the polyADP-ribose poly-

merase (Parp) family, which promotes the transient recruitment

of chromatin-remodeling enzymes and heterochromatin factors

to the DSB.

Figure 1. The Mechanism of DSB RepairTop: ATM phosphorylates H2AX at DSBs, creating a binding site for the mdc1protein. ATM-MRN complexes then associate with mdc1, promoting thespreading of gH2AX along the chromatin for hundreds of kilobases.Bottom: mdc1 recruits multiple DSB-repair proteins, including the RNF8/RNF168 ubiquitin ligases, to sites of damage. Chromatin ubiquitination thenfacilitates loading of the brca1 complex and 53BP1 DSB-repair proteins.P = phosphorylation, Ub = ubiquitination, MRN =mre11-rad50-nbs1 complex.

DSB Repair in Mammalian CellsThe mammalian DSB-repair pathway is a complex signaling

mechanism that regulates the two key responses to DSBs—

the rapid activation of cell-cycle checkpoints and the recruitment

of DNA-repair proteins onto the chromatin at the DSB (Figure 1).

The MRN complex, consisting of the mre11, rad50, and nbs1

proteins, is first recruited to DSBs, where it functions to recruit

and activate the ATM protein kinase (Lavin, 2008; Sun et al.,

2010). Activated ATM has been shown to phosphorylate

hundreds of proteins (Matsuoka et al., 2007), including proteins

involved in checkpoint activation (e.g., p53 and chk2) and

DNA-repair proteins such as brca1 and 53BP1 (Ciccia and

Elledge, 2010; Jackson and Bartek, 2009; Kennedy and D’An-

drea, 2006). A critical target for ATM is phosphorylation of the

C terminus of the histone variant H2AX. Phosphorylated H2AX

(referred to as gH2AX) creates a binding site for the BRCT

domains of the mdc1 protein (Lou et al., 2006; Stucki et al.,

2005) (Figure 1). Positioning of mdc1 at the DSB creates a dock-

ing site for additional DSB-repair proteins, including the MRN-

ATM complex (Chapman and Jackson, 2008; Melander et al.,

2008). Consequently, phosphorylation of H2AX by ATM spreads

away from the DSB, creating gH2AX domains that extend for

hundreds of kilobases along the chromatin from the DSB (Bon-

ner et al., 2008; Rogakou et al., 1999). The mdc1 protein also

recruits late-acting effector proteins, including the RNF8 and

RNF168 ubiquitin ligases, which ubiquitinate the chromatin and

promote loading of the brca1 and 53BP1 proteins (Doil et al.,

2009; Kolas et al., 2007). Similar to gH2AX spreading, chromatin

ubiquitination can also spread for tens of kilobases from the DSB

(Xu et al., 2010). This extension of chromatin ubiquitination is

opposed by the activity of the two E3 ligases, TRIP12 and

UBR5, which promote the ubiquitin-dependent degradation of

RNF168 (Gudjonsson et al., 2012). DSB repair therefore involves

the sequential recruitment and concentration of thousands of

copies of individual DSB-repair proteins onto the chromatin, as

well as extensive posttranslational modification of the nucleo-

somes.

DSB Repair by HR and NHEJThe actual repair of DSBs can proceed through two distinct

mechanisms: the error-prone nonhomologous end-joining

(NHEJ) pathway and the error-free homologous recombination

(HR) pathway (Huertas, 2010; Jackson and Bartek, 2009).

NHEJ involves minimal processing of the damaged DNA by

nucleases, followed by direct re-ligation of the DNA ends.

NHEJ requires the Ku70/80 DNA-binding complex and the

DNA-PKcs kinase. In contrast, HR requires the generation of

single-stranded DNA (ssDNA) intermediates, which are used

for homology searching within adjacent sister chromatids. The

production of ssDNA requires the initial nuclease activity of the

CtIP-MRN complex (Sartori et al., 2007), followed by further

end processing by additional nucleases to produce ssDNA

intermediates (Symington and Gautier, 2011). This ssDNA is

then used for homology searching in sister chromatids, which

then provide the template for accurate repair of DSBs by HR.

Importantly, because sister chromatids are only present during

the S and G2 phases of the cell cycle, HR repair is restricted to

this part of the cell cycle. Consequently, NHEJ predominates

in G1 and HR in S and G2 phases. However, how cells regulate

the choice between HR and NHEJ repair pathways is not well

understood, although both the 53BP1 and brca1 proteins can

play a key role in this choice (Bothmer et al., 2010; Bunting

et al., 2010).

Influence of Chromatin Organization on GenomicStabilityThe nucleosome is the basic functional unit of chromatin and

consists of 147 bp of DNA wrapped around a histone octamer

(Campos and Reinberg, 2009). Nucleosomes form linear 10 nm

beads-on-a-string structures that pack together to form 30 nm

arrays and other higher-order structures. The core of each nucle-

osome contains two H3-H4 dimers and two H2A-H2B dimers.

The N-terminal tails of histones extend out from the nucleosome

and contain conserved lysine residues that can be modified by

acetylation, methylation, or ubiquitination. These modifications

can function to attract specific chromatin complexes that can

then alter nucleosome function. In addition to histone posttrans-

lational modifications, chromatin organization is also regulated

by multisubunit remodeling complexes built around a large

motor ATPase. Four major ATPase families, including the SWI/

SNF, CHD, INO80, and ISWI families, have been identified in

eukaryotes (Clapier and Cairns, 2009). These remodeling com-

plexes utilize the energy fromATP hydrolysis to (1) remove nucle-

osomes from the chromatin and create open DNA sequences;

(2) shift the position of the nucleosome relative to the DNA by

exposing (or burying) a DNA sequence (nucleosome sliding);

or (3) exchange pre-existing histones for specialized histone

variants. Chromatin-remodeling complexes and histone modifi-

cations can alter the interaction within or between adjacent

nucleosomes and recruit chromatin-binding proteins to specific

regions (Cairns, 2005; Campos and Reinberg, 2009). Nucleo-

somes can therefore be envisaged as dynamic hubs to which

chromatin-modifying proteins and specific modifications attach

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1345

and that regulate the function and packing of the DNA in the

chromatin.

The importance of chromatin organization in maintaining

genomic stability is underscored by studies demonstrating that

mutation rates are not even across the human genome.

Sequencing of multiple cancer genomes has revealed that muta-

tions accumulate at much higher levels in compact, H3K9me3-

rich heterochromatin domains (Schuster-Bockler and Lehner,

2012), consistent with the slower rates of DNA repair reported

in heterochromatin (Goodarzi et al., 2008; Noon et al., 2010).

Further, inserts and deletions are depleted around nucleosomes,

whereas mutations tend to cluster on the nucleosomal DNA

(Chen et al., 2012; Sasaki et al., 2009; Tolstorukov et al., 2011),

and both can be influenced by the presence of specific epige-

netic modifications on the nucleosome (Schuster-Bockler and

Lehner, 2012; Tolstorukov et al., 2011). Some of these differ-

ences in mutation rates may accrue by negative selection (for

example, selection against mutations in coding regions) or

through protection of the DNA from mutagens by association

with nucleosomes. However, the elevated mutation rates

in compact, transcriptionally silent heterochromatin domains

(Schuster-Bockler and Lehner, 2012) imply that chromatin

packing may impact the detection or repair of damage by the

DNA-repair machinery. That is, the ability of the DNA-repair

machinery to access the DNA can have a significant impact on

genomic stability within specific regions.

DSBs Promote Rapid Histone H4 AcetylationOne of the best of the best characterized changes in chromatin

organization is the rapid formation of open chromatin structures

at DSBs. Several groups have demonstrated that this process is

associated with increased acetylation of histones H2A andH4 on

nucleosomes at DSBs (Downs et al., 2004; Jha et al., 2008;

Kusch et al., 2004; Murr et al., 2006). This acetylation extends

for hundreds of kilobases away from the break (Downs et al.,

2004; Murr et al., 2006; Xu et al., 2010), similar to the spreading

of gH2AX (Figure 1). The acetylation of histone H4 at DSBs

is dependent on the Tip60 acetyltransferase, a haploinsufficient

tumor-suppressor protein that is required for the repair of DSBs

(Doyon and Cote, 2004; Gorrini et al., 2007; Sun et al., 2010).

Tip60 is rapidly recruited to DSBs, where it can acetylatemultiple

DDR proteins, including histones H2A and H4, the ATM kinase,

p53, and other repair proteins (Bird et al., 2002; Ikura et al.,

2007; Jha et al., 2008; Sun et al., 2005, 2010; Sykes et al.,

2006). Tip60 functions in DSB repair as a subunit of the human

NuA4 (hNuA4) remodeling complex. hNuA4 contains at least

16 subunits (Doyon and Cote, 2004), of which 4 posses catalytic

activity—the Tip60 acetyltransferase, the p400 motor ATPase,

and the Ruvbl1 and Ruvbl2 helicase-like proteins. Multiple sub-

units of hNuA4, including Tip60 (Sun et al., 2009), Trrap (Downs

et al., 2004; Kusch et al., 2004; Murr et al., 2006), p400 (Xu et al.,

2010), and ruvbl1 and ruvbl2 (Jha et al., 2008) are corecruited to

DSBs, suggesting that these proteins are recruited together as

components of hNuA4.

Interestingly, hNuA4 is a fusion of two separate yeast com-

plexes—the smaller yeast NuA4 (yNuA4) complex, which con-

tains the Tip60 homolog esa1, and the ySWR1 complex, which

contains the Swr1 ATPase and the yeast Ruvbl1 and Ruvbl2

1346 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

homologs (Clapier and Cairns, 2009; Doyon and Cote, 2004).

Both yNuA4 (Downs et al., 2004) and ySWR1 complexes (Papa-

michos-Chronakis et al., 2006; van Attikum et al., 2007) are re-

cruited to enzymatically generated DSBs in yeast. However,

whereas yNuA4 and SWR1 are recruited to DSBs through direct

interaction with gH2AX (Downs et al., 2004; van Attikum et al.,

2007), hNuA4 is loaded onto chromatin through interaction

with the mdc1 protein (Xu and Price, 2011; Xu et al., 2010).

However, in both yeast and mammalian cells, loading of either

yNuA4 or hNuA4 at DSBs leads to the rapid acetylation of the

N-terminal tail of histone H4 by Tip60 (Downs et al., 2004; Ikura

et al., 2007; Murr et al., 2006; Sun et al., 2009; Xu et al., 2010).

Inactivation of Tip60 (Bird et al., 2002; Downs et al., 2004; Ikura

et al., 2000; Murr et al., 2006) blocks H4 acetylation and

increases sensitivity to DNA damage. Finally, mutation of the

Tip60 acetylation sites on H4 in yeast increases sensitivity to

DNA damage similar to that seen following Tip60 inactivation

(Bird et al., 2002; Downs et al., 2004). Although mutation of the

N-terminal tail of H4 is not possible in mammalian cells, the

results from both yeast and mammalian systems indicate that

the rapid recruitment of NuA4 complexes containing Tip60 to

DSBs leads to the increased acetylation of histone H4 and

H2A adjacent to the DSB.

HistoneAcetylationCreatesOpenChromatin StructuresIt is well-established that open chromatin conformations at

actively transcribed genes are associated with acetylation of

histone H4 (Campos and Reinberg, 2009; de Wit and van Steen-

sel, 2009). The N-terminal tail of histone H4 can interact with the

acidicpatchon the surfaceofH2A-H2Bdimersof adjacent nucle-

osomes (Luger et al., 2012). Disruption of this interaction by

acetylation of H4 on lysine 16 (Robinson et al., 2008; Shogren-

Knaak et al., 2006) inhibits packing of 30 nm fibers and leads to

chromatin decompaction. The increase in acetylation of histones

H2A and H4 at DSBs may therefore promote chromatin unpack-

inganddirect the formationof open, relaxedchromatin structures

detected at DSBs (Kruhlak et al., 2006). In fact, several studies

have demonstrated that chromatin at DSBs undergoes a transi-

tion to a more open, less compact conformation. For example,

the sensitivity of DNA to nuclease digestion increases after DNA

damage (Smerdon et al., 1978; Ziv et al., 2006), indicating that

linker DNA between nucleosomes is more accessible. Depletion

of histone H1, which binds to linker DNA and promotes nucleo-

some packing, promotes chromatin relaxation and facilitates

DSB repair (Murga et al., 2007). Histones at DSBs are susceptible

to extraction in low salt (Xu et al., 2010), implying a weaker inter-

action between DNA and histones at DSBs. Further, biophysical

studies demonstrate that DSBs lead to a localized chromatin

expansion at DSBs (Kruhlak et al., 2006). Finally, inactivation of

Tip60 (Murr et al., 2006; Xu et al., 2010, 2012) blocked the forma-

tion of open chromatin structures at DSBs, consistent with acet-

ylation of histone H4 by Tip60 playing a central role in creating

open, flexible chromatin structures at DSBs.

The p400 ATPase of hNuA4 Catalyzes H2A.Z Exchangeat DSBsIn addition to Tip60, the hNuA4 complex also contains the

p400 motor ATPase. p400 is a member of the Ino80 family of

Figure 2. H2A.Z Exchange Drives H4 AcetylationExchange of H2A for H2A.Z alters interaction between the N-terminal tail of H4 and adjacent nucleosomes, exposing the tail to acetylation by Tip60. Thecombination of H2A.Z exchange and H4 acetylation functions to shift chromatin into the open, relaxed conformation required for DSB repair. H4 = histone H4 tail,Ac = acetylation.

chromatin-remodeling ATPases, which includes two yeast

proteins—yIno80 and ySwr1. yIno80 and ySwr1 are both re-

cruited to DSBs in yeast, and loss of either component leads

to significant defects in both checkpoint activation and DSB

repair (Downs et al., 2004; Papamichos-Chronakis et al., 2006;

van Attikum et al., 2007). Members of the Ino80 family, including

the mammalian p400 ATPase, can exchange histone H2A for

the H2A variant H2A.Z (Fuchs et al., 2001; Gevry et al., 2007;

Kusch et al., 2004), suggesting that Ino80 family members may

regulate H2A.Z exchange during DSB repair. Indeed, in yeast,

loss of H2A.Z leads to increased sensitivity to DNA-damaging

agents (Morillo-Huesca et al., 2010; Papamichos-Chronakis

et al., 2011) and defective repair of DSBs (Kalocsay et al.,

2009). Although a transient increase in H2A.Z deposition at

DSBs in yeast has been reported (Kalocsay et al., 2009), other

studies suggest that Ino80 and Swr1 may function antagonisti-

cally to regulate ormaintain H2A.Z at DSBs (Papamichos-Chron-

akis et al., 2006; van Attikum et al., 2007), with no overall increase

in H2A.Z exchange at DSBs in yeast (van Attikum et al., 2007).

However, in mammalian cells, the hNuA4 complex promotes

not only H4 acetylation by the Tip60 subunit but also the rapid

exchange of H2A for H2A.Z at DSBs (Figure 2) (Xu et al., 2012).

H2A.Z exchange requires the ATPase activity of the p400 motor

protein and creates chromatin domains containing H2A.Z nucle-

osomes that extend away from the DSB. Surprisingly, H2A.Z

precedes, and is required for, both the acetylation of histone

H4 by Tip60 and the creation of open chromatin domains at

DSBs (Downs et al., 2004; Murr et al., 2007; Xu et al., 2010).

The exchange of H2A.Z onto nucleosomes at DSBs leads to

an increase in the salt solubility of the histones (Xu et al.,

2012), indicating the formation of open chromatin at the site of

damage. This is consistent with published work demonstrating

that H2A.Z nucleosomes are less stable than H2A nucleosomes

and are more sensitive to extraction at low-salt concentrations

(Henikoff et al., 2009; Jin and Felsenfeld, 2007; Weber et al.,

2010; Zhang et al., 2005). However, other studies have shown

that H2A.Z stabilizes nucleosomes (Fan et al., 2004; Park et al.,

2004). These opposing effects of H2A.Z on nucleosome struc-

ture have been extensively reviewed by others (Billon and

Cote, 2012; Zlatanova and Thakar, 2008). However, it has been

noted that the ability of H2A.Z to reduce nucleosome stability

is dependent on both histone modifications and the presence

of additional histone variants, including histone H3.3, on the

nucleosome (Henikoff et al., 2009; Jin and Felsenfeld, 2007;

Jin et al., 2009). The ability of H2A.Z to destabilize nucleosomes

at DSBs may therefore depend on both the presence of

additional histone variants (such as H3.3) and histone post-

translational modifications on nucleosomes. Consistent with

this, the ability of H2A.Z to create open chromatin structures

at DSBs requires both the presence of H2A.Z and acetylation

of histone H4 tails by the Tip60 acetyltransferase (Xu et al.,

2012) (Figure 2). That is, H2A.Z appears to only be capable

of destabilizing nucleosomes at DSBs in the context of an

acetylated H4 tail.

How the presence of H2A.Z promotes the acetylation of the

N-terminal tail of H4 by Tip60 is less clear. Nucleosomes contain-

ing H2A.Z exhibit only subtle differences in structure from H2A

nucleosomes (Suto et al., 2000). The N-terminal tail of histone

H4 interacts with an acidic patch on the surface of the nucleo-

some and promotes packing into 30 nm fibers (Robinson et al.,

2008; Shogren-Knaak et al., 2006). In H2A.Z, this acidic patch

is extended in length, and it has been proposed that this ex-

tended acidic region stabilizes the interaction between H2A.Z

and H4, promoting packing of nucleosome fibers (Fan et al.,

2004). This would tend to restrict the ability of Tip60 to acetylate

the N-terminal tail of H4. However, as discussed above, the

ability of H2A.Z to impact chromatin organization can be modu-

lated by the presence of histone H3.3 or by additional histone

modifications within the nucleosome (Jin and Felsenfeld, 2007;

Jin et al., 2009; Zlatanova and Thakar, 2008) (Figure 2). H2A.Z

exchange may therefore be only part of the equation, with the

potential for exchange of H3.3, specific acetylation of H2A.Z,

or additional remodeling motor ATPases contributing to acetyla-

tion of histone H4 in response to DSBs. Unraveling these early

events will provide new insight into H2A.Z-mediated shifts in

chromatin structure at the DSB.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1347

Figure 3. H2A.Z Exchange Drives Chromatin Changes that Direct

Chromatin Modification at DSBsH2A.Z exchange promotes H4 acetylation by Tip60, which in turn directsubiquitination of the chromatin by the RNF8/RNF168 ubiquitin ligases. 53BP1is then recruited to chromatin through interaction with H4K20me2. 53BP1mayutilize pre-existing H4K20me2 or require de novo methylation by MMSET.Whether ubiquitination promotes access to H4K20me2 is not yet known.Association of NuA4-Tip60 with mdc1 omitted for clarity. P = phosphorylation,Ac = H4 acetylation, Ub = ubiquitination, Me = H4K20me2.

Rapid Chromatin Remodeling Promotes OrderedChromatin ModificationThe NuA4-driven changes in chromatin organization (Figure 2)

have a significant impact on the mechanism of DSB repair. In

particular, the formation of open chromatin domains through

H2A.Z exchange and H4 acetylation facilitates further DNA-

damage-dependent modification of the chromatin by both ubiq-

uitination andmethylation of histone H4 (Figure 3). Inactivation of

components of hNuA4, including p400, Tip60, or Trrap, blocks

the ubiquitination of histone H2A/H2AX by RNF8/RNF168 and

inhibits the subsequent loading of several effector proteins,

including brca1, 53BP1, and rad51, onto chromatin (Figure 3)

(Courilleau et al., 2012; Murr et al., 2006; Xu et al., 2010, 2012).

Brca1 recruitment requires interaction between the RAP80

subunit of the brca1 complex and ubiquitinated chromatin at

DSBs (Sobhian et al., 2007). The NuA4-dependent shift in chro-

matin structure at DSBs may therefore reveal cryptic sites for

H2A/H2AX ubiquitination by RNF8/RNF168 and drive loading

1348 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

of brca1. The recruitment of 53BP1, a DNA-repair protein that

regulates NHEJ (Bunting et al., 2010), is complex and can also

be regulated by RNF8/RNF168-mediated chromatin ubiquitina-

tion (Doil et al., 2009; Huen et al., 2007). However, 53BP1 does

not possess an identifiable ubiquitin-binding motif. It has also

been shown that 53BP1 recruitment to DSBs requires H4 acety-

lation (Murr et al., 2006; Xu et al., 2010) and H4K20 methylation

(Botuyan et al., 2006). In fact, 53BP1’s tudor domain can bind to

histone H4 dimethylated on lysine 20 (H4K20me2) (Botuyan

et al., 2006). Because a significant fraction (>80%) of H4K20 is

dimethylated in mammalian cells, the increased acetylation of

histone H4 at DSBs may function to both unpack closely

opposed chromatin fibers and reveal H4K20me2 for 53BP1

binding. Also, H2A/H2AX ubiquitination by RNF8 and RNF168

may further promote 53BP1 loading by altering the accessibility

of 53BP1 to H4K20me2 (Figure 3). Interestingly, mice lacking

both of the suv4-20h H4K20me2 methyltransferases have

almost no H4K20me2 and display increased genomic instability

yet maintain normal recruitment of 53BP1 to DSBs (Schotta

et al., 2008). Although this may suggest that H4K20me2 is

dispensable for 53BP1 recruitment to DSBs, it has recently

been reported that the methyltransferase MMSET is recruited

to DSBs and promotes the formation of H4K20me2 (Pei et al.,

2011). Recruitment of MMSET may provide the mechanism for

methylation of the small fraction of H4K20 that is not constitu-

tively methylated and may partially compensate for loss of

constitutive H4K20me2 in the suv4-20h1/suv4-20h2 double-

knockout mice. In fact, 53BP1 has been reported to promote

long-range interactions between DNA ends (Difilippantonio

et al., 2008), suggesting that 53BP1 binding may itself play

a role in regulating or stabilizing chromatin structure after DNA

damage (Noon et al., 2010). Thus the initial change in nucleo-

some function imposed by H2A.Z exchange promotes an

ordered series of histone modifications, including acetylation

of histone H4 and ubiquitination of the chromatin (Figure 3).

This may then either unmask H4K20me2 buried within the nucle-

osome structure and/or promote H4K20 methylation by MMSET

and thereby facilitate loading of both 53BP1 and brca1 com-

plexes onto the chromatin. The early remodeling events there-

fore play a critical role in directing the ordered recruitment of

DSB-repair proteins to the site of damage.

Impact of H2A.Z on DSB RepairCells lacking H2A.Z or components of NuA4 are hypersensitive

to IR and have defects in both NHEJ- and HR-directed repair

(Downs et al., 2004; Ikura et al., 2000; Murr et al., 2006; Xu

et al., 2010, 2012). This wide range of defects reflects the early

and critical role of hNuA4 in promoting access to sites of

damage and reflects both the failure to create open chromatin

structures and the lack of recruitment of brca1, which is essential

for HR-mediated DSB repair. Intriguingly, when H2A.Z exchange

at DSBs is inhibited, cells undergo unrestricted end resection,

leading to accumulation of excess ssDNA and the loss of

Ku70/80 binding (Xu et al., 2012). Further, this defect can be

reversed by depletion of CtIP, suggesting that H2A.Z exchange

functions to restrain or restrict the ability of the CtIP-MRN

nuclease complex to initiate end resection of the DSB. In

yeast, loss of the ySwr1 ATPase also leads to defects in Ku70

recruitment and defects in error-free NHEJ (van Attikum et al.,

2007), although this is not directly linked to H2A.Z exchange.

Recent work on the role of H2A.Z at transcriptional start sites

(TSS) provides some potential insight into how H2A.Z may

restrict end resection. The TSS of many genes are flanked by

H2A.Z nucleosomes (Jin et al., 2009; Zhang et al., 2005), which

may function to fix the positions of nucleosomes on either side

of the TSS and thereby maintain nucleosome-free DNA for tran-

scription-factor binding. Nucleosomes are also lost at DSBs,

creating nucleosome-free regions (Tsukuda et al., 2005). The

placement of H2A.Z nucleosomes on either side of nucleo-

some-free regions at the DSB therefore creates a structure

similar to that reported at the TSS of genes. Positioning of

H2A.Z on either side of the DSB may therefore define the limits

of the nucleosome-free region and create a chromatin template

that restricts or limits end resection by the CtIP-MRN complex.

The early remodeling of the chromatin at DSBs through H2A.Z

exchange and H4 acetylation is therefore critical for setting the

scene for further processing and eventual repair of the DSB

through either NHEJ or HR pathways.

Accessing DSBs in HeterochromatinHow cells access and repair DSBs within the higher-order

chromatin environment of heterochromatin has been the sub-

ject of recent studies. Heterochromatin is classically described

as condensed, densely staining regions of DNA that contain

few active genes but are enriched for repetitive sequences.

Mammalian heterochromatin is characterized by high levels

of the histone modifications H3K9me3 and H3K27me3 and

low levels of histone acetylation. Heterochromatin is main-

tained by a dense array of specific chromatin-binding proteins,

including members of the HP1 family (which bind to methylated

H3K9), kap-1, histone deacetylases (HDACs), and histone

methyltransferases. From the perspective of DSB repair, it

is important to determine whether the dense packing and

unique array of heterochromatin-binding proteins present a

specific barrier to the DSB-repair machinery. Further, the pres-

ence of repetitive DNA within heterochromatin may provide

a significant challenge for HR-mediated repair, requiring more

stringent control of HR to prevent inappropriate recombination

events.

kap-1 is a repressor protein that interacts with HP1, HDACs,

and histone methyltransferases and functions to maintain

heterochromatin (Iyengar and Farnham, 2011). In response to

DSBs, kap-1 is phosphorylated by ATM (Goodarzi et al., 2008;

Ziv et al., 2006), promoting a general relaxation of the chromatin

structure. Repair of DSBs (as measured by loss of gH2AX foci)

is significantly slower within heterochromatin regions and is

dependent on phosphorylation of kap-1 by ATM. Further,

kap-1 phosphorylation promotes release of the CHD3-remodel-

ing ATPase from heterochromatin (Goodarzi et al., 2011), a

process required for efficient repair. It is currently unclear how

loss of CHD3 or phosphorylation of kap-1 (which remains asso-

ciated with the DSB regions) impacts overall chromatin structure

at DSBs. In addition to kap-1 phosphorylation, HP1 proteins

(including HP1a, b, and g) can repress heterochromatin repair.

Depletion of HP1 proteins (or depletion of the H3K9methyltrans-

ferases) can decondense heterochromatin and promote repair of

DSBs even in the absence of ATM kinase activity (Chiolo et al.,

2011; Goodarzi et al., 2008, 2011). Further, there is some

evidence to suggest that HP1 proteins are actively ejected

from the chromatin during DNA repair (Ayoub et al., 2008).

These observations are consistent with the idea that the dense

packing of nucleosomes and the presence of specific hetero-

chromatin-binding complexes are a significant barrier to repair

of heterochromatic DSBs. Further, these results indicate a

critical role for phosphorylation of kap-1 by the ATM kinase in

promoting the unpacking of heterochromatin and thereby facili-

tating repair of heterochromatic DSBs. Currently, it is unclear

whether, for example, the NuA4-Tip60 complex acetylates his-

tones at heterochromatic DSBs or whether the phosphorylation

of kap-1 within heterochromatin is sufficient to create the

required open chromatin structure. Further, given that H2A.Z is

found at heterochromatin boundaries, it will be interesting to

determine whether this histone variant is important for hetero-

chromatic DSB repair as well.

Spacing of H2AX Nucleosomes and HeterochromatinStudies on DSB repair in heterochromatin utilize microscopy to

monitor the appearance of gH2AX foci and either DAPI (to detect

dense chromatin domains) or antibodies to locate regions of

heterochromatin (Chiolo et al., 2011; Goodarzi et al., 2008;

Noon et al., 2010). Several studies indicate that gH2AX foci pref-

erentially assemble in euchromatin or are predominantly located

at the boundary of the heterochromatin (Goodarzi et al., 2008;

Kim et al., 2007; Noon et al., 2010). However, studies with enzy-

matically generated DSBs coupled with chromatin immunopre-

cipitation indicate that gH2AX does not spread uniformly along

the chromosome (Iacovoni et al., 2010; Meier et al., 2007; Savic

et al., 2009), and the size of the gH2AX domain varies between

different chromatin locations (Xu et al., 2012). Further, in yeast,

gH2AX does not spread through heterochromatin regions (Kim

et al., 2007). H2AX is unique compared to other DSB-repair

proteins because it is prepositioned on nucleosomes rather

than recruited to DSBs. To function as a DSB detector, and to

allow for gH2AX propagation along the chromatin, it would be

expected that H2AX should be evenly deposited along the chro-

matin. However, the amount of H2AX in cells can vary from 2% to

20% of the total H2A (Rogakou et al., 1998). That is, in some

cells, 1 in 2.5 nucleosomes contain H2AX, whereas in other cells,

as few as 1 in 25 nucleosomes may contain H2AX. In fact, high-

resolution microscopy indicates that H2AX is concentrated in

specific domains (Bewersdorf et al., 2006), and chromatin immu-

noprecipitation combined with sequencing (ChIP-Seq) analysis

indicates that H2AX is concentrated within gene-rich regions

(Iacovoni et al., 2010). This raises the possibility that H2AX

density or distribution within heterochromatin is significantly

lower than in other domains. The failure to detect gH2AX foci

in heterochromatin withmicroscopymay therefore reflect altered

H2AX distribution in heterochromatin and a reduced need for

H2AX function in heterochromatin.

In addition to differential H2AX distribution in heterochro-

matin, recent work in Drosophila has provided an alternative

explanation for why gH2AX foci are only detected at the

periphery of the heterochromatin. This work demonstrates that

phosphorylation of H2AX and initial recruitment of DSB-repair

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1349

Figure 4. Creating Access to DSBsProposed chronological sequence of steps in remodeling of a DSB. InitialPARylation by PARP1 leads to rapid recruitment of NuRD and ALC1 (throughinteraction with PAR) and kap-1/HP1 complexes (possibly through interactionwith PAR). Deacetylation of histones (including H2A, H3, and H4) by NuRDand proposed H3K9 methylation (by HP1/kap1-associated lysine methyl-transferases [KMTs] including suv39h1 and G9a) create a temporary repres-sive chromatin structure with low histone acetylation and high density ofH3K9me3. Subsequently, the HP1/kap1, ALC1, and NuRD complexes arerapidly released from the chromatin, potentially through dePARylation byPARG. Phosphorylation of gH2AX then recruits NuA4-Tip60, promoting theordered remodeling of the chromatin through H2A.Z exchange, acetylation ofhistone H4 (H4Ac), chromatin ubiquitination, and modulation of H4K20me2.This creates a common chromatin template for DSB repair by either NHEJ- orHR-mediated repair.

proteins to the break occur normally within the heterochro-

matin. However, these heterochromatic DSBs rapidly migrate

out of the heterochromatin; hence the actual DSB repair is

carried out within euchromatin (Chiolo et al., 2011; Jakob

et al., 2011). Further, this relocation of the DSB is only partly

dependent on ATM, indicating that phosphorylation of kap-1

by ATM does not contribute to this process. Moving the

DSB out of the heterochromatin may limit recombination with

repetitive sequences and allow increased mobility and easier

access to the DSB. However, it should be noted that experi-

ments in mammalian cells have indicated only limited mobility

for DSBs, so it will be important to explore DSB mobility in the

heterochromatin of mammalian cells (Krawczyk et al., 2012;

Soutoglou et al., 2007). Finally, it is interesting to note that, in

yeast, exchange of H2A.Z into the chromatin is required for

relocalization of persistent DSBs to the nuclear periphery (Kaloc-

say et al., 2009). The NuA4-mediated exchange of H2A.Z

at heterochromatin DSBs (Figure 2) may potentially promote

relocation of DSBs out of the heterochromatin. Clearly, our

understanding of the mechanism of DSB repair within hetero-

chromatin is limited. Developing new approaches, such as

coupling synthetic nucleases to create DSBs in heterochro-

matin with ChIP-Seq approaches, may provide a more directed

approach to understanding DSB repair within specific chromatin

domains.

Early Recruitment Events: HP1It is now clear that additional chromatin based events occur prior

to the NuA4-mediated chromatin relaxation. In particular, 2

heterochromatin-associated proteins, HP1 and kap-1, partici-

pate in the early response to DSBs in euchromatin. HP1a and

kap-1 are rapidly recruited to DSBs within seconds to minutes

after damage induction ((Baldeyron et al., 2011; Luijsterburg

et al., 2009) reviewed in (Soria et al., 2012)). The recruitment of

HP1a and kap-1 is essential for loading 53BP1 and brca1 and

for HR directed repair. Kap-1 and HP1 proteins may be recruited

to DSBs as a single complex, althoughHP1a loading requires the

histone chaperone ASF1 (Baldeyron et al., 2011). Importantly,

HP1 and kap-1 recruitment to euchromatin is transient, with

both proteins dissociating from the break a few minutes after

damage induction (Baldeyron et al., 2011). It is currently unclear

if HP1 and kap-1 have distinct roles in heterochromatin and

euchromatin during DSB repair, and why transient recruitment

and release of HP1 is important remains to be investigated.

One potential explanation is that kap-1 exists as a complex

with repressive factors including HDACs and H3K9 methyltrans-

ferases (Iyengar and Farnham, 2011). Recruitment of repressive

kap-1 complexes may rapidly ‘‘heterochromatinize’’ the DSB

region, preventing transcription and stabilizing the chromatin

structure. Further, since the Tip60 sub-unit of NuA4 requires

interaction with H3K9me3 for stimulation of its acetyltransferase

activity (Sun et al., 2009), recruitment of kap-1/HP1 complexes

may provide a mechanism for the rapid methylation of H3K9

and therefore facilitate the activity of both Tip60 and the NuA4

complex. The transient accumulation of kap-1 and HP1 com-

plexes may rewrite local histone modification signatures,

thereby increasing available H3K9me3 and promoting the

activity of the Tip60 sub-unit of NuA4 and other factors.

1350 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Early Recruitment of NuRD and ALC1 Complexesthrough PARylationSimilarly to recruitment of kap-1/HP1, there is also a rapid and

transient accumulation of the NuRD (Chou et al., 2010; Larsen

et al., 2010; Polo et al., 2010; Smeenk et al., 2010) and ALC1

(Ahel et al., 2009) remodeling complexes at DSBs (Figure 4).

NuRD complexes contain either the CHD3 or CHD4 ATPase,

HDAC1 or HDAC2, and associated regulatory subunits (Clapier

and Cairns, 2009). NuRD is a repressive complex that maintains

higher-order chromatin structure. Inactivation of NuRD or ALC1

leads to defects in DSB repair and increased sensitivity to DNA

damage (Ahel et al., 2009; Chou et al., 2010; Polo et al., 2010;

Smeenk et al., 2010). NuRD regulates the acetylation of p53

and thereby controls the extent of G1-S arrest following DNA

damage (Larsen et al., 2010; Polo et al., 2010). Second, NuRD,

like NuA4, is required for chromatin ubiquitination by RNF8/

RNF168 and for loading of brca1 (Larsen et al., 2010; Smeenk

et al., 2010). The recruitment of NuRD complexes to DSBs

requires PARylation of the chromatin by PARP1 (Chou et al.,

2010; Polo et al., 2010). PARP1 belongs to a family of Parps

that play a central role in both transcription and DNA repair

(Gibson and Kraus, 2012). Chromatin at DSBs is rapidly and tran-

siently PARylated (Figure 4), and it is this modification, rather

than gH2AX or ATM signaling, that localizes NuRD at the DSB

(Chou et al., 2010; Polo et al., 2010).

Similarly, ALC1, a remodeling ATPase that functions to reposi-

tion nucleosomes on the chromatin, is also rapidly recruited to

DSBs through direct interaction with PAR chains on the chro-

matin (Ahel et al., 2009; Gottschalk et al., 2009). ALC1 loading

is rapid and transient after DNA damage and may favor the

formation of open chromatin (Ahel et al., 2009). Thus at least

three remodeling complexes, HP1/kap-1, NuRD, and ALC1,

are rapidly, but transiently, recruited to DSBs (Figure 4). Because

PARylation of the chromatin is transient yet independent of

gH2AX formation, the recruitment of HP1/kap-1, NuRD, and

ALC1 likely precedes the recruitment and loading of the NuA4-

Tip60 complex (Figure 4). However, whether these complexes

work sequentially or in parallel is not yet known. For example,

whether the recruitment of NuA4-Tip60 or H2A.Z exchange

requires prior processing of the chromatin by either ALC1 or

NuRD or is dependent on chromatin PARylation is not known.

Further, it remains to be seen whether the HP1/kap-1 complex

is recruited to DSBs through PARylation or some other mecha-

nism. Finally, the rapid release of ALC1, NuRD, and HP1/kap-1

complexes may be brought about by dePARylation of the chro-

matin by polyADP-ribose glycohydrolases (PARGs) (Figure 4).

Understanding the regulation of PARGsmay provide new insight

into some of the earliest events occurring during DSB repair.

The HP1/kap-1, ALC1, and NuRD complexes deploy a wide

range of chromatin-remodeling activities, including HDACs

(NuRD), methyltransferases (HP1/kap-1), and remodeling

ATPase activities (NuRD and ALC1) at the DSB. Because these

complexes are only retained at the DSBs for a short time period

(minutes), they must play a critical role in the initial detection and

processing of the chromatin at the DSB. This role could include

the rapid termination of local transcription by promoting histone

deacetylation (NuRD) and/or the formation of repressive chro-

matin through histone methylation and loading of kap-1/HP1

complexes. By erasing previous histone acetylation marks,

NuRD and the other complexes may prime the chromatin for

uniform acetylation by the NuA4-Tip60 complex. Further, ALC1

may function to reposition nucleosomes at the DSB and to stabi-

lize the chromatin and facilitate further processing and repair.

These events may rapidly and transiently stabilize the local chro-

matin structure by creating a temporary, compacted, repressive

chromatin environment at the DSB. Subsequently, DSB sig-

naling, including gH2AX formation and ATM activation, leads to

the ordered recruitment of DSB-repair proteins to the chromatin

at DSBs. The transient creation of PAR chains at DSBs by

PARP1, which allows the rapid recruitment of NuRD, ALC1,

and potentially kap-1/HP1, is therefore a critical early event in

the DNA-damage response.

Conclusions and Future DirectionsA eukaryotic cell must integrate classical DSB-repair signaling

and repair by NHEJ and HR pathways with the complexity of

the local chromatin architecture. Functional chromatin domains,

such as replication forks, genes, or heterochromatin, differ

significantly in the patterns of histone modifications, the types

of chromatin-binding proteins, and the degree of nucleosome

packing. Each of these domains may therefore require unique

chromatin-remodeling complexes to alter the local chromatin

architecture at individual DSBs. Identifying the protein-remodel-

ing complexes that are essential for repair in specific chromatin

structures is therefore of key importance. Such processes may

be critical for reshaping the local chromatin structure and for

creating a common DNA template that can be presented to the

DSB-repair machinery. It is clear that some of the earliest events

in DSB repair occurring in the first few minutes after damage can

have a profound impact on processing of the damaged chro-

matin template. However, in addition to these early events, there

are many additional steps in DSB repair that require chromatin

remolding, such as homology searching during HR-directed

repair or regulation of end resection during repair. In addition,

resetting the chromatin structure and restoring the original

epigenetic code to the repaired chromatin are vital to ensure

that normal functionality is restored to the damaged chromatin.

ACKNOWLEDGMENTS

This work was supported by NIH grants CA64585 and CA93602 to B.D.P. and

grant RO1-DK43889 to A.D.D.

REFERENCES

Ahel, D., Horejsı, Z., Wiechens, N., Polo, S.E., Garcia-Wilson, E., Ahel, I., Flynn,

H., Skehel, M., West, S.C., Jackson, S.P., et al. (2009). Poly(ADP-ribose)-

dependent regulation of DNA repair by the chromatin remodeling enzyme

ALC1. Science 325, 1240–1243.

Ayoub, N., Jeyasekharan, A.D., Bernal, J.A., and Venkitaraman, A.R. (2008).

HP1-beta mobilization promotes chromatin changes that initiate the DNA

damage response. Nature 453, 682–686.

Baldeyron, C., Soria, G., Roche, D., Cook, A.J., and Almouzni, G. (2011).

HP1alpha recruitment to DNA damage by p150CAF-1 promotes homologous

recombination repair. J. Cell Biol. 193, 81–95.

Bewersdorf, J., Bennett, B.T., and Knight, K.L. (2006). H2AX chromatin struc-

tures and their response to DNA damage revealed by 4Pi microscopy. Proc.

Natl. Acad. Sci. USA 103, 18137–18142.

Billon, P., andCote, J. (2012). Precise deposition of histone H2A.Z in chromatin

for genome expression and maintenance. Biochim. Biophys. Acta 1819,

290–302.

Bird, A.W., Yu, D.Y., Pray-Grant, M.G., Qiu, Q., Harmon, K.E., Megee, P.C.,

Grant, P.A., Smith, M.M., and Christman, M.F. (2002). Acetylation of histone

H4 by Esa1 is required for DNA double-strand break repair. Nature 419,

411–415.

Bonner, W.M., Redon, C.E., Dickey, J.S., Nakamura, A.J., Sedelnikova, O.A.,

Solier, S., and Pommier, Y. (2008). GammaH2AX and cancer. Nat. Rev. Cancer

8, 957–967.

Bothmer, A., Robbiani, D.F., Feldhahn, N., Gazumyan, A., Nussenzweig, A.,

andNussenzweig, M.C. (2010). 53BP1 regulates DNA resection and the choice

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1351

between classical and alternative end joining during class switch recombina-

tion. J. Exp. Med. 207, 855–865.

Botuyan, M.V., Lee, J., Ward, I.M., Kim, J.E., Thompson, J.R., Chen, J., and

Mer, G. (2006). Structural basis for the methylation state-specific recognition

of histone H4-K20 by 53BP1 and Crb2 in DNA repair. Cell 127, 1361–1373.

Bunting, S.F., Callen, E., Wong, N., Chen, H.T., Polato, F., Gunn, A., Bothmer,

A., Feldhahn, N., Fernandez-Capetillo, O., Cao, L., et al. (2010). 53BP1 inhibits

homologous recombination in Brca1-deficient cells by blocking resection of

DNA breaks. Cell 141, 243–254.

Cairns, B.R. (2005). Chromatin remodeling complexes: strength in diversity,

precision through specialization. Curr. Opin. Genet. Dev. 15, 185–190.

Campos, E.I., and Reinberg, D. (2009). Histones: annotating chromatin. Annu.

Rev. Genet. 43, 559–599.

Chapman, J.R., and Jackson, S.P. (2008). Phospho-dependent interactions

between NBS1 and MDC1 mediate chromatin retention of the MRN complex

at sites of DNA damage. EMBO Rep. 9, 795–801.

Chen, X., Chen, Z., Chen, H., Su, Z., Yang, J., Lin, F., Shi, S., and He, X. (2012).

Nucleosomes suppress spontaneous mutations base-specifically in eukary-

otes. Science 335, 1235–1238.

Chiolo, I., Minoda, A., Colmenares, S.U., Polyzos, A., Costes, S.V., and

Karpen, G.H. (2011). Double-strand breaks in heterochromatin move outside

of a dynamic HP1a domain to complete recombinational repair. Cell 144,

732–744.

Chou, D.M., Adamson, B., Dephoure, N.E., Tan, X., Nottke, A.C., Hurov, K.E.,

Gygi, S.P., Colaiacovo, M.P., and Elledge, S.J. (2010). A chromatin localization

screen reveals poly (ADP ribose)-regulated recruitment of the repressive poly-

comb and NuRD complexes to sites of DNA damage. Proc. Natl. Acad. Sci.

USA 107, 18475–18480.

Ciccia, A., and Elledge, S.J. (2010). The DNA damage response: making it safe

to play with knives. Mol. Cell 40, 179–204.

Clapier, C.R., and Cairns, B.R. (2009). The biology of chromatin remodeling

complexes. Annu. Rev. Biochem. 78, 273–304.

Courilleau, C., Chailleux, C., Jauneau, A., Grimal, F., Briois, S., Boutet-Rob-

inet, E., Boudsocq, F., Trouche, D., and Canitrot, Y. (2012). The chromatin

remodeler p400 ATPase facilitates Rad51-mediated repair of DNA double-

strand breaks. J. Cell Biol. 199, 1067–1081.

de Wit, E., and van Steensel, B. (2009). Chromatin domains in higher eukary-

otes: insights from genome-wide mapping studies. Chromosoma 118, 25–36.

Difilippantonio, S., Gapud, E., Wong, N., Huang, C.Y., Mahowald, G., Chen,

H.T., Kruhlak, M.J., Callen, E., Livak, F., Nussenzweig, M.C., et al. (2008).

53BP1 facilitates long-range DNA end-joining during V(D)J recombination.

Nature 456, 529–533.

Doil, C., Mailand, N., Bekker-Jensen, S., Menard, P., Larsen, D.H., Pepperkok,

R., Ellenberg, J., Panier, S., Durocher, D., Bartek, J., et al. (2009). RNF168

binds and amplifies ubiquitin conjugates on damaged chromosomes to allow

accumulation of repair proteins. Cell 136, 435–446.

Downs, J.A., Allard, S., Jobin-Robitaille, O., Javaheri, A., Auger, A., Bouchard,

N., Kron, S.J., Jackson, S.P., and Cote, J. (2004). Binding of chromatin-modi-

fying activities to phosphorylated histone H2A at DNA damage sites. Mol. Cell

16, 979–990.

Doyon, Y., and Cote, J. (2004). The highly conserved andmultifunctional NuA4

HAT complex. Curr. Opin. Genet. Dev. 14, 147–154.

Fan, J.Y., Rangasamy, D., Luger, K., and Tremethick, D.J. (2004). H2A.Z alters

the nucleosome surface to promote HP1alpha-mediated chromatin fiber

folding. Mol. Cell 16, 655–661.

Fuchs, M., Gerber, J., Drapkin, R., Sif, S., Ikura, T., Ogryzko, V., Lane, W.S.,

Nakatani, Y., and Livingston, D.M. (2001). The p400 complex is an essential

E1A transformation target. Cell 106, 297–307.

Gevry, N., Chan, H.M., Laflamme, L., Livingston, D.M., and Gaudreau, L.

(2007). p21 transcription is regulated by differential localization of histone

H2A.Z. Genes Dev. 21, 1869–1881.

1352 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Gibson, B.A., and Kraus, W.L. (2012). New insights into the molecular and

cellular functions of poly(ADP-ribose) and PARPs. Nat. Rev. Mol. Cell Biol.

13, 411–424.

Goodarzi, A.A., Noon, A.T., Deckbar, D., Ziv, Y., Shiloh, Y., Lobrich, M., and

Jeggo, P.A. (2008). ATM signaling facilitates repair of DNA double-strand

breaks associated with heterochromatin. Mol. Cell 31, 167–177.

Goodarzi, A.A., Kurka, T., and Jeggo, P.A. (2011). KAP-1 phosphorylation

regulates CHD3 nucleosome remodeling during the DNA double-strand break

response. Nat. Struct. Mol. Biol. 18, 831–839.

Gorrini, C., Squatrito, M., Luise, C., Syed, N., Perna, D., Wark, L., Martinato, F.,

Sardella, D., Verrecchia, A., Bennett, S., et al. (2007). Tip60 is a haplo-insuffi-

cient tumour suppressor required for an oncogene-induced DNA damage

response. Nature 448, 1063–1067.

Gottschalk, A.J., Timinszky, G., Kong, S.E., Jin, J., Cai, Y., Swanson, S.K.,

Washburn, M.P., Florens, L., Ladurner, A.G., Conaway, J.W., and Conaway,

R.C. (2009). Poly(ADP-ribosyl)ation directs recruitment and activation of an

ATP-dependent chromatin remodeler. Proc. Natl. Acad. Sci. USA 106,

13770–13774.

Grewal, S.I., and Jia, S. (2007). Heterochromatin revisited. Nat. Rev. Genet. 8,

35–46.

Gudjonsson, T., Altmeyer, M., Savic, V., Toledo, L., Dinant, C., Grofte,M., Bart-

kova, J., Poulsen, M., Oka, Y., Bekker-Jensen, S., et al. (2012). TRIP12 and

UBR5 suppress spreading of chromatin ubiquitylation at damaged chromo-

somes. Cell 150, 697–709.

Henikoff, S., Henikoff, J.G., Sakai, A., Loeb, G.B., and Ahmad, K. (2009).

Genome-wide profiling of salt fractionsmaps physical properties of chromatin.

Genome Res. 19, 460–469.

Huen, M.S., Grant, R., Manke, I., Minn, K., Yu, X., Yaffe, M.B., and Chen, J.

(2007). RNF8 transduces the DNA-damage signal via histone ubiquitylation

and checkpoint protein assembly. Cell 131, 901–914.

Huertas, P. (2010). DNA resection in eukaryotes: deciding how to fix the break.

Nat. Struct. Mol. Biol. 17, 11–16.

Iacovoni, J.S., Caron, P., Lassadi, I., Nicolas, E., Massip, L., Trouche, D., and

Legube, G. (2010). High-resolution profiling of gammaH2AX around DNA

double strand breaks in the mammalian genome. EMBO J. 29, 1446–1457.

Ikura, T., Ogryzko, V.V., Grigoriev, M., Groisman, R., Wang, J., Horikoshi, M.,

Scully, R., Qin, J., and Nakatani, Y. (2000). Involvement of the TIP60 histone

acetylase complex in DNA repair and apoptosis. Cell 102, 463–473.

Ikura, T., Tashiro, S., Kakino, A., Shima, H., Jacob, N., Amunugama, R., Yoder,

K., Izumi, S., Kuraoka, I., Tanaka, K., et al. (2007). DNA damage-dependent

acetylation and ubiquitination of H2AX enhances chromatin dynamics. Mol.

Cell. Biol. 27, 7028–7040.

Iyengar, S., and Farnham, P.J. (2011). KAP1 protein: an enigmatic master

regulator of the genome. J. Biol. Chem. 286, 26267–26276.

Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in human

biology and disease. Nature 461, 1071–1078.

Jakob, B., Splinter, J., Conrad, S., Voss, K.O., Zink, D., Durante, M., Lobrich,

M., and Taucher-Scholz, G. (2011). DNA double-strand breaks in heterochro-

matin elicit fast repair protein recruitment, histone H2AX phosphorylation and

relocation to euchromatin. Nucleic Acids Res. 39, 6489–6499.

Jha, S., Shibata, E., and Dutta, A. (2008). Human Rvb1/Tip49 is required for the

histone acetyltransferase activity of Tip60/NuA4 and for the downregulation of

phosphorylation on H2AX after DNA damage. Mol. Cell. Biol. 28, 2690–2700.

Jin, C., and Felsenfeld, G. (2007). Nucleosome stability mediated by histone

variants H3.3 and H2A.Z. Genes Dev. 21, 1519–1529.

Jin, C., Zang, C.,Wei, G., Cui, K., Peng,W., Zhao, K., and Felsenfeld, G. (2009).

H3.3/H2A.Z double variant-containing nucleosomes mark ‘nucleosome-free

regions’ of active promoters and other regulatory regions. Nat. Genet. 41,

941–945.

Kalocsay, M., Hiller, N.J., and Jentsch, S. (2009). Chromosome-wide Rad51

spreading and SUMO-H2A.Z-dependent chromosome fixation in response

to a persistent DNA double-strand break. Mol. Cell 33, 335–343.

Kennedy, R.D., and D’Andrea, A.D. (2006). DNA repair pathways in clinical

practice: lessons from pediatric cancer susceptibility syndromes. J. Clin.

Oncol. 24, 3799–3808.

Kim, J.A., Kruhlak, M., Dotiwala, F., Nussenzweig, A., and Haber, J.E. (2007).

Heterochromatin is refractory to gamma-H2AX modification in yeast and

mammals. J. Cell Biol. 178, 209–218.

Kolas, N.K., Chapman, J.R., Nakada, S., Ylanko, J., Chahwan, R., Sweeney,

F.D., Panier, S., Mendez, M., Wildenhain, J., Thomson, T.M., et al. (2007).

Orchestration of the DNA-damage response by the RNF8 ubiquitin ligase.

Science 318, 1637–1640.

Krawczyk, P.M., Borovski, T., Stap, J., Cijsouw, T., ten Cate, R., Medema, J.P.,

Kanaar, R., Franken, N.A., and Aten, J.A. (2012). Chromatin mobility is

increased at sites of DNA double-strand breaks. J. Cell Sci. 125, 2127–2133.

Kruhlak, M.J., Celeste, A., Dellaire, G., Fernandez-Capetillo, O., Muller, W.G.,

McNally, J.G., Bazett-Jones, D.P., and Nussenzweig, A. (2006). Changes in

chromatin structure and mobility in living cells at sites of DNA double-strand

breaks. J. Cell Biol. 172, 823–834.

Kusch, T., Florens, L., Macdonald, W.H., Swanson, S.K., Glaser, R.L., Yates,

J.R., 3rd, Abmayr, S.M., Washburn, M.P., and Workman, J.L. (2004). Acetyla-

tion by Tip60 is required for selective histone variant exchange at DNA lesions.

Science 306, 2084–2087.

Larsen, D.H., Poinsignon, C., Gudjonsson, T., Dinant, C., Payne, M.R., Hari,

F.J., Rendtlew Danielsen, J.M., Menard, P., Sand, J.C., Stucki, M., et al.

(2010). The chromatin-remodeling factor CHD4 coordinates signaling and

repair after DNA damage. J. Cell Biol. 190, 731–740.

Lavin, M.F. (2008). Ataxia-telangiectasia: from a rare disorder to a paradigm for

cell signalling and cancer. Nat. Rev. Mol. Cell Biol. 9, 759–769.

Lou, Z., Minter-Dykhouse, K., Franco, S., Gostissa, M., Rivera, M.A., Celeste,

A., Manis, J.P., van Deursen, J., Nussenzweig, A., Paull, T.T., et al. (2006).

MDC1 maintains genomic stability by participating in the amplification of

ATM-dependent DNA damage signals. Mol. Cell 21, 187–200.

Luger, K., Dechassa, M.L., and Tremethick, D.J. (2012). New insights into

nucleosome and chromatin structure: an ordered state or a disordered affair?

Nat. Rev. Mol. Cell Biol. 13, 436–447.

Luijsterburg, M.S., Dinant, C., Lans, H., Stap, J., Wiernasz, E., Lagerwerf, S.,

Warmerdam, D.O., Lindh, M., Brink, M.C., Dobrucki, J.W., et al. (2009).

Heterochromatin protein 1 is recruited to various types of DNA damage.

J. Cell Biol. 185, 577–586.

Matsuoka, S., Ballif, B.A., Smogorzewska, A., McDonald, E.R., 3rd, Hurov,

K.E., Luo, J., Bakalarski, C.E., Zhao, Z., Solimini, N., Lerenthal, Y., et al.

(2007). ATM and ATR substrate analysis reveals extensive protein networks

responsive to DNA damage. Science 316, 1160–1166.

Meier, A., Fiegler, H., Munoz, P., Ellis, P., Rigler, D., Langford, C., Blasco,M.A.,

Carter, N., and Jackson, S.P. (2007). Spreading of mammalian DNA-damage

response factors studied by ChIP-chip at damaged telomeres. EMBO J. 26,

2707–2718.

Melander, F., Bekker-Jensen, S., Falck, J., Bartek, J., Mailand, N., and Lukas,

J. (2008). Phosphorylation of SDT repeats in the MDC1 N terminus triggers

retention of NBS1 at the DNA damage-modified chromatin. J. Cell Biol. 181,

213–226.

Morillo-Huesca, M., Clemente-Ruiz, M., Andujar, E., and Prado, F. (2010). The

SWR1 histone replacement complex causes genetic instability and genome-

wide transcription misregulation in the absence of H2A.Z. PLoS ONE 5,

e12143.

Murga, M., Jaco, I., Fan, Y., Soria, R., Martinez-Pastor, B., Cuadrado, M.,

Yang, S.M., Blasco, M.A., Skoultchi, A.I., and Fernandez-Capetillo, O.

(2007). Global chromatin compaction limits the strength of the DNA damage

response. J. Cell Biol. 178, 1101–1108.

Murr, R., Loizou, J.I., Yang, Y.G., Cuenin, C., Li, H., Wang, Z.Q., and Herceg, Z.

(2006). Histone acetylation by Trrap-Tip60 modulates loading of repair

proteins and repair of DNA double-strand breaks. Nat. Cell Biol. 8, 91–99.

Murr, R., Vaissiere, T., Sawan, C., Shukla, V., and Herceg, Z. (2007). Orches-

tration of chromatin-based processes: mind the TRRAP. Oncogene 26, 5358–

5372.

Noon, A.T., Shibata, A., Rief, N., Lobrich, M., Stewart, G.S., Jeggo, P.A., and

Goodarzi, A.A. (2010). 53BP1-dependent robust localized KAP-1 phosphory-

lation is essential for heterochromatic DNA double-strand break repair. Nat.

Cell Biol. 12, 177–184.

Papamichos-Chronakis, M., Krebs, J.E., and Peterson, C.L. (2006). Interplay

between Ino80 and Swr1 chromatin remodeling enzymes regulates cell cycle

checkpoint adaptation in response to DNA damage. Genes Dev. 20, 2437–

2449.

Papamichos-Chronakis, M., Watanabe, S., Rando, O.J., and Peterson, C.L.

(2011). Global regulation of H2A.Z localization by the INO80 chromatin-remod-

eling enzyme is essential for genome integrity. Cell 144, 200–213.

Park, Y.J., Dyer, P.N., Tremethick, D.J., and Luger, K. (2004). A new fluores-

cence resonance energy transfer approach demonstrates that the histone

variant H2AZ stabilizes the histone octamer within the nucleosome. J. Biol.

Chem. 279, 24274–24282.

Pei, H., Zhang, L., Luo, K., Qin, Y., Chesi, M., Fei, F., Bergsagel, P.L., Wang, L.,

You, Z., and Lou, Z. (2011). MMSET regulates histone H4K20 methylation and

53BP1 accumulation at DNA damage sites. Nature 470, 124–128.

Peng, J.C., and Karpen, G.H. (2008). Epigenetic regulation of heterochromatic

DNA stability. Curr. Opin. Genet. Dev. 18, 204–211.

Polo, S.E., Kaidi, A., Baskcomb, L., Galanty, Y., and Jackson, S.P. (2010).

Regulation of DNA-damage responses and cell-cycle progression by the chro-

matin remodelling factor CHD4. EMBO J. 29, 3130–3139.

Robinson, P.J., An, W., Routh, A., Martino, F., Chapman, L., Roeder, R.G., and

Rhodes, D. (2008). 30 nm chromatin fibre decompaction requires both H4-K16

acetylation and linker histone eviction. J. Mol. Biol. 381, 816–825.

Rogakou, E.P., Boon, C., Redon, C., and Bonner, W.M. (1999). Megabase

chromatin domains involved in DNA double-strand breaks in vivo. J. Cell

Biol. 146, 905–916.

Rogakou, E.P., Pilch, D.R., Orr, A.H., Ivanova, V.S., and Bonner, W.M. (1998).

DNA double-stranded breaks induce histone H2AX phosphorylation on serine

139. J. Biol. Chem. 273, 5858–5868.

Sartori, A.A., Lukas, C., Coates, J., Mistrik, M., Fu, S., Bartek, J., Baer, R., Lu-

kas, J., and Jackson, S.P. (2007). Human CtIP promotes DNA end resection.

Nature 450, 509–514.

Sasaki, S., Mello, C.C., Shimada, A., Nakatani, Y., Hashimoto, S., Ogawa, M.,

Matsushima, K., Gu, S.G., Kasahara, M., Ahsan, B., et al. (2009). Chromatin-

associated periodicity in genetic variation downstream of transcriptional start

sites. Science 323, 401–404.

Savic, V., Yin, B., Maas, N.L., Bredemeyer, A.L., Carpenter, A.C., Helmink,

B.A., Yang-Iott, K.S., Sleckman, B.P., and Bassing, C.H. (2009). Formation

of dynamic gamma-H2AX domains along broken DNA strands is distinctly

regulated by ATM and MDC1 and dependent upon H2AX densities in chro-

matin. Mol. Cell 34, 298–310.

Schotta, G., Sengupta, R., Kubicek, S., Malin, S., Kauer, M., Callen, E.,

Celeste, A., Pagani, M., Opravil, S., De La Rosa-Velazquez, I.A., et al. (2008).

A chromatin-wide transition to H4K20 monomethylation impairs genome

integrity and programmed DNA rearrangements in the mouse. Genes Dev.

22, 2048–2061.

Schuster-Bockler, B., and Lehner, B. (2012). Chromatin organization is amajor

influence on regional mutation rates in human cancer cells. Nature 488,

504–507.

Shogren-Knaak, M., Ishii, H., Sun, J.M., Pazin, M.J., Davie, J.R., and Peterson,

C.L. (2006). Histone H4-K16 acetylation controls chromatin structure and

protein interactions. Science 311, 844–847.

Smeenk, G., Wiegant, W.W., Vrolijk, H., Solari, A.P., Pastink, A., and van Atti-

kum, H. (2010). The NuRD chromatin-remodeling complex regulates signaling

and repair of DNA damage. J. Cell Biol. 190, 741–749.

Smerdon, M.J. (1991). DNA repair and the role of chromatin structure. Curr.

Opin. Cell Biol. 3, 422–428.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1353

Smerdon, M.J., Tlsty, T.D., and Lieberman, M.W. (1978). Distribution of ultra-

violet-inducedDNA repair synthesis in nuclease sensitive and resistant regions

of human chromatin. Biochemistry 17, 2377–2386.

Sobhian, B., Shao, G., Lilli, D.R., Culhane, A.C., Moreau, L.A., Xia, B., Living-

ston, D.M., and Greenberg, R.A. (2007). RAP80 targets BRCA1 to specific

ubiquitin structures at DNA damage sites. Science 316, 1198–1202.

Soria, G., Polo, S.E., and Almouzni, G. (2012). Prime, repair, restore: the active

role of chromatin in the DNA damage response. Mol. Cell 46, 722–734.

Soutoglou, E., Dorn, J.F., Sengupta, K., Jasin, M., Nussenzweig, A., Ried, T.,

Danuser, G., and Misteli, T. (2007). Positional stability of single double-strand

breaks in mammalian cells. Nat. Cell Biol. 9, 675–682.

Stucki, M., Clapperton, J.A., Mohammad, D., Yaffe, M.B., Smerdon, S.J., and

Jackson, S.P. (2005). MDC1 directly binds phosphorylated histone H2AX to

regulate cellular responses to DNA double-strand breaks. Cell 123, 1213–

1226.

Sun, Y., Jiang, X., Chen, S., Fernandes, N., and Price, B.D. (2005). A role for the

Tip60 histone acetyltransferase in the acetylation and activation of ATM. Proc.

Natl. Acad. Sci. USA 102, 13182–13187.

Sun, Y., Jiang, X., Xu, Y., Ayrapetov, M.K., Moreau, L.A., Whetstine, J.R., and

Price, B.D. (2009). Histone H3methylation links DNA damage detection to acti-

vation of the tumour suppressor Tip60. Nat. Cell Biol. 11, 1376–1382.

Sun, Y., Jiang, X., and Price, B.D. (2010). Tip60: connecting chromatin to DNA

damage signaling. Cell Cycle 9, 930–936.

Suto, R.K., Clarkson, M.J., Tremethick, D.J., and Luger, K. (2000). Crystal

structure of a nucleosome core particle containing the variant histone

H2A.Z. Nat. Struct. Biol. 7, 1121–1124.

Sykes, S.M., Mellert, H.S., Holbert, M.A., Li, K., Marmorstein, R., Lane, W.S.,

and McMahon, S.B. (2006). Acetylation of the p53 DNA-binding domain regu-

lates apoptosis induction. Mol. Cell 24, 841–851.

1354 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Symington, L.S., and Gautier, J. (2011). Double-strand break end resection

and repair pathway choice. Annu. Rev. Genet. 45, 247–271.

Tolstorukov, M.Y., Volfovsky, N., Stephens, R.M., and Park, P.J. (2011).

Impact of chromatin structure on sequence variability in the human genome.

Nat. Struct. Mol. Biol. 18, 510–515.

Tsukuda, T., Fleming, A.B., Nickoloff, J.A., and Osley, M.A. (2005). Chromatin

remodelling at a DNA double-strand break site in Saccharomyces cerevisiae.

Nature 438, 379–383.

van Attikum, H., Fritsch, O., and Gasser, S.M. (2007). Distinct roles for SWR1

and INO80 chromatin remodeling complexes at chromosomal double-strand

breaks. EMBO J. 26, 4113–4125.

Weber, C.M., Henikoff, J.G., and Henikoff, S. (2010). H2A.Z nucleosomes en-

riched over active genes are homotypic. Nat. Struct. Mol. Biol. 17, 1500–1507.

Xu, Y., and Price, B.D. (2011). Chromatin dynamics and the repair of DNA

double strand breaks. Cell Cycle 10, 261–267.

Xu, Y., Sun, Y., Jiang, X., Ayrapetov, M.K., Moskwa, P., Yang, S., Weinstock,

D.M., and Price, B.D. (2010). The p400 ATPase regulates nucleosome stability

and chromatin ubiquitination during DNA repair. J. Cell Biol. 191, 31–43.

Xu, Y., Ayrapetov, M.K., Xu, C., Gursoy-Yuzugullu, O., Hu, Y., and Price, B.D.

(2012). Histone H2A.Z controls a critical chromatin remodeling step required

for DNA double-strand break repair. Mol. Cell 48, 723–733.

Zhang, H., Roberts, D.N., and Cairns, B.R. (2005). Genome-wide dynamics of

Htz1, a histone H2A variant that poises repressed/basal promoters for activa-

tion through histone loss. Cell 123, 219–231.

Ziv, Y., Bielopolski, D., Galanty, Y., Lukas, C., Taya, Y., Schultz, D.C., Lukas,

J., Bekker-Jensen, S., Bartek, J., and Shiloh, Y. (2006). Chromatin relaxation

in response to DNA double-strand breaks is modulated by a novel ATM- and

KAP-1 dependent pathway. Nat. Cell Biol. 8, 870–876.

Zlatanova, J., and Thakar, A. (2008). H2A.Z: view from the top. Structure 16,

166–179.

Leading Edge

Review

Chromatin Movement in theMaintenance of Genome Stability

Vincent Dion1 and Susan M. Gasser1,2,*1Friedrich Miescher Institute for Biomedical Research, Maulbeerstrasse 66, CH-4058 Basel, Switzerland2Faculty of Natural Sciences, University of Basel, CH-4056 Basel, Switzerland*Correspondence: susan.gasser@fmi.ch

http://dx.doi.org/10.1016/j.cell.2013.02.010

Mechanistic analyses based on improved imaging techniques have begun to explore the biologicalimplications of chromatin movement within the nucleus. Studies in both prokaryotes and eukary-otes have shed light on what regulates the mobility of DNA over long distances. Interestingly, ineukaryotes, genomic loci increase their movement in response to double-strand break induction.Break mobility, in turn, correlates with the efficiency of repair by homologous recombination.We review here the source and regulation of DNA mobility and discuss how it can both contributeto and jeopardize genome stability.

The Not-Quite-Random Walk of ChromatinChromatin is often depicted as a static entity comprising DNA

wrapped around histone octamers organized in the form of

arrays. Constant changes in the composition of nucleosomes,

posttranscriptional modifications of histones, and shifts in

nucleosome positioning (Campos and Reinberg, 2009; Segal

and Widom, 2009) ensure that chromatin is dynamic. In

addition, recent studies argue that the physical movement of

the chromatin fiber itself is an important element of chromatin

dynamics. Indeed, chromatin in the interphase nucleus moves

constantly, not only due to temperature-dependent Brownian

motion. Here, we review new findings that shed light on the

mechanisms that promote DNA movement as well as its bio-

logical implications.

In the 1990s, the development of a nonmultimerizing green

fluorescent protein (GFP)-Lac repressor (Lacl) fusion that could

bind lacO arrays integrated in the yeast genome opened the

door to microscopic analysis of the position of chromosomal

loci in living cells (Robinett et al., 1996). The LacI-lacO system

was followed by development of a TetR-TetO tagging pair (Mi-

chaelis et al., 1997) and the coupling of these to a GFP-pore

protein fusion (Heun et al., 2001a, 2001b). This made it possible

to track the movement of tagged chromosomal loci accurately,

independent of nuclear movement. In these systems, the GFP-

fused repressors concentrate at their cognate operator sites,

generating a visible fluorescent spot. Other methods that track

chromatin in living cells rely on the incorporation of fluorescently

labeled deoxy- or ribo-NTP analogs (for example, Zink et al.,

1998) or on the expression of photoactivatable fluorescent

proteins linked to histones (Kruhlak et al., 2006; Wiesmeijer

et al., 2008). Although these avoid the use of bacterial operator

arrays, they do not allow one to score the dynamics of specific

chromosomal loci.

Once the movement of a lacO-tagged locus is captured by

time-lapse microscopy, the character of the movement can be

quantified using a mean-squared displacement (MSD) analysis

(Berg, 1993). Multiple time-lapse series of a given locus are

acquired and are used to calculate the average of the squared

distance covered by that locus, which is, in turn, plotted against

increasing time intervals (Figure 1). In brief, MSD = < (xt – xt+Dt)2 >

where t is time and x is the position of a moving fluorescent spot.

This method of analysis is highly robust as it averages a large

number of data points to generate quantitative movement

parameters such as the diffusion coefficient and radius of

constraint (Rc).

The diffusion coefficient of a particle moving in a random

Brownian walk is directly proportional to the initial slope on an

MSD graph, and it scales with time (Berg, 1993). However, as

time intervals increase, the mean square of the movement

(MSD) curve will plateau because of the constraint or confine-

ment imposed by the nuclear sphere (that is, a moving chromo-

somal locus will not move beyond the confinement of the nuclear

envelope, regardless of the time interval queried) (Figure 1). From

the plateau reached by the MSD curve over time, one can calcu-

late the radius of the constrained volume within which the

particle moves.

Using this model for single genomic locus movement, early

experiments suggested that thediffusion coefficient of chromatin

movement ranges from 10�4 to 10�3 mm2/s, which—remark-

ably—seemed to hold true for bacteria and yeast as well as Dro-

sophila and mammalian cells, regardless of the precise tracking

method used or the range of nuclear sizes (Bornfleth et al.,

1999; Chubb et al., 2002; Heun et al., 2001b; Marshall

et al., 1997; Neumann et al., 2012; Vazquez et al., 2001; Weber

et al., 2012). In a pioneering work, Marshall et al. (1997) showed

that chromatin movement appears to be a constrained random

walk.More recent studies indicate that themovement of chromo-

somal loci in bacteria, yeast, and mammalian cells does not fully

recapitulate aBrownian randomwalk (Bornfleth et al., 1999;Neu-

mann et al., 2012; Weber et al., 2010, 2012). Both intrinsic and

external constraints appear to restrict movement, causing it to

appear nonrandom. On the other hand, the movement of an

excised, extrachromosomal ring of yeast chromatin is indistin-

guishable from a random walk trajectory (Neumann et al., 2012).

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1355

Figure 1. MSD AnalysisMSD values are derived from determining the distance moved by a particleover increasing time intervals, Dt. In other words, (Xt – Xt+Dt), where X isthe position at time t. The top depicts a characteristic MSD plot for a randomwalk where the slope (m) equals the diffusion coefficient (D) times twice thenumber of dimensions in which movement is measured (d). The middle panelshows the shape of a MSD graph in cases where the motion is directional. Themobility of a particle moving according to Brownian motion within confinedspace will generate a curve that levels off at larger time intervals (bottom).In this case, the plateau (p) that the curve reaches is equal to the square rootof 2/5 times the number of dimensions (d) times the radius of constraint(Rc) (Neumann et al., 2012).

1356 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

The Stokes-Einstein equation:D= kBT=6pha, in which kB is the

Boltzmann constant, h is the viscosity of the liquid, and a is the

size of the moving particle, dictates that, if chromatin movement

were Brownian, the diffusion coefficient (D) would be directly

proportional to the temperature (T) in degrees Kelvin (Weber

et al., 2012). This has been tested in both yeast and bacteria

by determining the diffusion coefficient at different temperatures

and checking for a linear relationship. Movement of a bacterial

locus has been determined at temperatures ranging from 10�C(283 K) to 30�C (303 K). The expected change in diffusion coef-

ficient should be around 7% for Brownian motion within this

temperature range, yet a 2-fold increase is calculated (Weber

et al., 2012). This argues that DNA motion is superthermal. Simi-

larly, unexpectedly large changes in mobility are scored for

tagged loci in yeast for temperature shifts from 25�C to 37�C(Neumann et al., 2012). This, coupled with the fact that locus

movement in yeast is significantly affected by the level of glucose

in themedia and by the presence of protonophores that collapse

mitochondrial and plasma membrane potentials (Gartenberg

et al., 2004; Heun et al., 2001b; Marshall et al., 1997), argues

strongly that ATP is likely to be involved in chromatin movement.

This effect of ATP depletion is also observed in mammalian

cells (Chubb et al., 2002).

In addition to the nonlinear effects of temperature, which argue

against pure Brownian motion, studies in bacteria identify a drag

on moving genomic loci that cannot be explained by the princi-

ples of Brownian motion (Weber et al., 2010). Earlier, eukaryotic

loci had been observed to undergo spring-like—and thus

nonrandom—movements, visualized as ‘‘large’’ unidirectional

steps (>0.5 mm in <10 s) in yeast, that are often followed by

similar movement in the opposite direction (Heun et al.,

2001b). The analysis of chromatin dynamics in Drosophila sper-

matocytes revealed that tagged loci have a tendency to move in

one direction and then return to their previous location (Vazquez

et al., 2001). Based on such observations and on computer

simulations, it has been proposed that chromosomal movement

is best explained by fractional Langevin motion, in which an

elastic, semiviscous milieu (i.e., the nucleoplasm) ‘‘pushes

back’’ on the moving particle, possibly accounting for this irreg-

ular, spring-like movement (Weber et al., 2010).

A further source of drag on chromatin diffusion comes from the

contiguity of the chromatin fiber itself. As mentioned above,

when a chromosomal domain is excised from a chromosome

forming an extrachromosomal ring of chromatin, the diffusion

coefficient doubles and the Rc becomes identical to the radius

of the nucleus (Gartenberg et al., 2004). It was concluded that

the flanking chromosomal DNA and the context of a tagged

locus within the linear molecule of chromosomal DNA restrict

chromatin movement. A comparison of actual movement with

computer simulations shows that the MSD curve of the excised

particle fits exactly that of a simulated randomwalk, with a radius

the same as that of the nucleus—0.9 mm—whereas the inte-

grated locus exhibits additional constraint (Neumann et al.,

2012). Some constraint likely arises from natural chromosomal

anchorage sites, such as centromere tethering to the

membrane-associated spindle pole body, the interaction of telo-

meres to structural proteins of the nuclear envelope, or the asso-

ciation of stress-induced genes with pores (Cabal et al., 2006;

Gartenberg et al., 2004; Hediger et al., 2002; Taddei and Gasser,

2012; Taddei et al., 2006, 2010; Zimmer and Fabre, 2011). The

association of mammalian loci with the nucleolus can also

constrain locus movement (Chubb et al., 2002; Wiesmeijer

et al., 2008), and, in telomerase-deficient ALT cells (alternative

lengthening of telomeres ), telomeres appeared to be tethered

to promyelocytic leukemia (PML) bodies (Molenaar et al., 2003).

In conclusion, the mobility of a DNA locus in the interphase

nucleus can be considered as nondirected motion that fluctu-

ates with ATP levels and depends disproportionately on temper-

ature. The constraint on DNA stems from the chromatin fiber

itself, the nature of the nucleoplasm, and protein-protein interac-

tions that anchor loci to nuclear structures.

Cellular Mechanisms that Regulate ChromatinMovementDNA mobility changes during the cell cycle and during develop-

ment, which raises the possibility that it may be regulated. For

instance, culturedDrosophila spermatocytes display twomodes

of movement during premeiotic development. Whereas larger

changes in positions are observed early in differentiation, more

constrained motion is detected in mature spermatocytes (Vaz-

quez et al., 2001). For tagged loci in yeast, less movement is

observed in S phase than in G1 phase nuclei, a drop that corre-

lates inversely with the number of active replication forks and

possibly also with dNTP levels (Heun et al., 2001b). In mammals,

results obtained by visualizing chromosomal regions using

a photoactivatable histone fusion suggest that no change in

mobility occurs between cells in mid- and late G1, S, and G2

(Walter et al., 2003; Wiesmeijer et al., 2008). It appeared,

however, that there is significantly more mobility early in G1, as

compared to later stages of the cell cycle. Indeed, measuring

the distance between chromosome territories labeled with

dNTP analogs shows that chromosome territories can move

over distances ranging between 0.47 and 4.44 mm in early G1,

whereas at later cell-cycle stages, the distances observed are

only within 0.25 to 2.11 mm (Walter et al., 2003). Taken together,

these results argue that the mobility of a chromosomal locus is

under the control of biological, as well as physical, parameters.

In some instances, changes in transcriptional activity are

correlated with the nuclear position of a locus. For example,

yeast telomeres, which silence nearby genes, are found at the

nuclear periphery, where they are anchored through an interac-

tion of the silencing machinery with the nuclear envelope (Gar-

tenberg et al., 2004; Gotta et al., 1996; Taddei and Gasser,

2012; Taddei et al., 2004; Zimmer and Fabre, 2011), while

active genes can be tethered to nuclear pores (Cabal et al.,

2006; Casolari et al., 2004; Egecioglu and Brickner, 2011;

Taddei et al., 2006). It was hypothesized that an increase in

transcriptional output might enhance the mobility of a locus to

facilitate its relocalization to the appropriate nuclear compart-

ment. This agrees with experiments by Chuang et al. (2006),

who have shown that the activation of transcription by targeting

the viral transactivator VP16 to a heterochromatic transgene

array in mammalian cells leads to long-range directional move-

ment perpendicular to the nuclear membrane. This experiment

establishes a link between transcriptional activation, decompac-

tion, and the mobility of chromatin and provides a striking

example of non-Brownian motion. Similarly, the targeting of

a fusion of LexA-VP16 fusion to a nontelomeric locus in yeast

increases both transcriptional activity and movement, scored

as the radius of constraint and number of large steps (Neumann

et al., 2012). Moreover, the targeting of this same transcriptional

activator to an otherwise silent telomeric locus shifts it away from

the nuclear envelope (Hediger et al., 2006; Taddei et al., 2006).

Although these examples link transcriptional control with

movement, there are many examples in which transcription

and mobility can be uncoupled. For example, the highly tran-

scribed genes that associate with nuclear pores become

anchored and are therefore highly constrained (Cabal et al.,

2006; Taddei et al., 2006). In contrast, a transcriptionally silent

chromatin ring can diffuse freely throughout the nucleus if the

proteins necessary for its anchoring to the nuclear envelope

are missing (Gartenberg et al., 2004). Most significantly, the

directed binding of a LexA-Gal4 fusion protein to a promoter

can increase its transcriptional output without altering chromatin

movement (Neumann et al., 2012), and both genetic and chem-

ical inhibitors of transcriptional elongation failed to alter chro-

matin mobility in yeast (Neumann et al., 2012; A. Taddei, F.R.

Neumann, and S.M.G., unpublished data). Pliss et al. (2009)

demonstrated that transcription does not correlate with chro-

matin movement in cultured mammalian cells, and others have

shown that chromatin moves similarly whether or not it binds

CFP-SUV39H1, an enzyme that methylates histone H3 lysine 9

(H3K9) in heterochromatin (Wiesmeijer et al., 2008). In brief, tran-

scriptional activation and repression are not obligatorily linked to

either movement or tethering, even though transcription can

correlate with enhanced movement in specific cases.

If transcription can be uncoupled from locus mobility, then

what drives chromatin movement and why is it sensitive to

ATP levels? One ATP-dependent activity that correlates with

transcription in a context-dependent manner is nucleosome re-

modeling. For example, the activation of the yeast PHO5

promoter coincides with the removal of nucleosomes from the

promoter region by two nucleosome remodeling complexes,

Swi2/Snf2 and INO80 (Barbaric et al., 2007; Steger et al.,

2003). When PHO5 is tracked by the LacI-lacO system during

its activation, the open chromatin structure shows an increased

diffusion coefficient and a larger Rc (Neumann et al., 2012). In the

presence of phosphate, on the other hand, which represses

PHO5 transcription by preventing the removal of nucleosomes

in the promoter, the diffusion coefficient and Rc were smaller.

Furthermore, deleting ARP8, which encodes a subunit of the

INO80 nucleosome remodeler, severely reduces transcriptional

output, provokes a failure to respond to phosphate levels, and

leads to a nucleosomal structure in the promoter that is only

partially accessible (Barbaric et al., 2007; Steger et al., 2003). If

chromatin structure were responsible for the mobility of this

locus, one would predict that the locus would have an interme-

diate level of motion and would not respond to phosphate

levels in an arp8 mutant. This, indeed, was the case (Neumann

et al., 2012). These findings argue that nucleosome remodeling

at the endogenous PHO5 locus correlates tightly with induced

locus mobility.

Nucleosome remodelers are characterized by the presence

of a large ATPase subunit of the Snf2 family, which typically

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1357

associates with numerous accessory subunits and influences

virtually all DNA-based transactions. Not surprisingly, recent

work has begun to examine the impact of remodelers on DNA

mobility in contexts other than transcription (Clapier and Cairns,

2009; Dion et al., 2010; Lans et al., 2012). Specifically, the

recruitment of the INO80 remodeler, which helps remodel nucle-

osomes at double-strand breaks (Morrison et al., 2004; Tsukuda

et al., 2005; van Attikum et al., 2004), increases the Rc of an

undamaged locus to which it is targeted without increasing tran-

scription (Neumann et al., 2012). The effect is entirely dependent

on the Ino80 catalytic subunit, as the targeting of a mutant that

cannot bind ATP fails to increase chromatin mobility (Neumann

et al., 2012). Moreover, the targeting of another remodeler, the

Swi2/Snf2 ATPase complex, did not promote movement in

a similar manner. It is unclear why this is the case, but it may

reflect differences in biochemical activities of the two enzymes

or an absence of cofactors or histone modifications at the loci

tested. Given that there are 17 Snf2-type ATPases in yeast and

53 in human (Flaus et al., 2006), it is attractive to imagine that

different chromatin remodelers alter chromatin mobility in

different ways, regulating long-range chromatin movement while

they alter local nucleosomal organization.

Mobility of Damaged DNAThere are several ways of probing for the mobility of damaged

DNA. One is to introduce specific patterns of DNA damage, for

example, with a linear UV light or ionizing radiation (IR) tracks

and fixing the cells at several time points after damage induction

(e.g., Aten et al., 2004). Immunofluorescence against specific

DNA repair markers can then be applied to seewhether the linear

track has changed its shape (e.g., Jakob et al., 2011). Alterna-

tively, live-cell imaging can be used after damage induction to

watch the diffusion of a repair factor of choice fused to a fluores-

cent protein. This assay, when coupled to discrete patterns of

DNA damage tracks, tends to be qualitative because discrete

particles to track are not present. However, randomly induced

damage by IR or DNA-damaging drugs lead to discernible repair

foci (Haaf et al., 1995), which can be followed using the same

single-particle tracking described above for lacO-tagged chro-

mosomal loci. Finally, for site-specific damage, one can label

the genomic site to be damaged with bacterial operators (e.g.,

Nagai et al., 2008; Soutoglou et al., 2007). This is particularly

useful as the differences in mobility between the same damaged

and undamaged locus can be addressed. It is, however, limited

to double-strand breaks and for a specific protein-DNA adduct in

yeast (Nielsen et al., 2009).

Several recent studies have investigated whether repair foci—

and, by extension, DNA lesions—show long-range mobility in

mammalian cells. These studies have yielded mixed results.

For instance, Nelms et al. (1998) showed in human cells that irra-

diation-induced damage imaged by an incorporation of the

thymidine analog bromodeoxyuridine (BrdU) moves very little.

Similar results were obtained using live-cell imaging of a single

double-strand break induced by the endonuclease I-SceI (Sou-

toglou et al., 2007) and by tracking laser-damaged regions in

photosensitized cells (Kruhlak et al., 2006). Meanwhile, Jakob

et al. (2009a, 2009b) have used IR induced by heavy ion

sources to show that repair foci have similar kinetics as undam-

1358 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

aged loci, but do not appear to be constrained over several

hours, suggesting that damaged DNA could travel large

distances given enough time. Finally, damage induced by

a particle irradiation is highly mobile and moved over large

distances within minutes (Agarwal et al., 2011; Aten et al.,

2004; Krawczyk et al., 2012). This latter situation is reminiscent

of results obtained with uncapped telomeres in mouse embry-

onic cells (Dimitrova et al., 2008). To reconcile this wide range

of results, we propose that different types of damage, different

cell lines, variable growth conditions, the specific marker protein

tracked, and/or the method of visualizing movement all con-

tribute to different results.

Two recent studies in budding yeast, in which various param-

eters of damage and imaging could be better controlled, showed

that a single double-strand break is more mobile than the same

undamaged locus (Dion et al., 2012; Mine-Hattab and Rothstein,

2012). In these studies, MSD analyses show that the genomic

locus monitored is constrained to a Rc of about 0.4 mm, whereas

after DSB induction, the Rc increases to about 0.7 mm in haploid

cells and 0.9 mm in diploids. These values are similar even though

one group used a haploid strain and the other used a diploid

strain. The change in mobility ranged from 13% to 47% of the

nuclear volume in haploid cells or from 3% to 30% in diploid cells

(Dion et al., 2012; Mine-Hattab and Rothstein, 2012).

In yeast, specific genetic factors that affect the movement of

broken DNA have been identified. Mine-Hattab and Rothstein

(2012) defined that deletion of SAE2, which codes for an enzyme

important in DSB end resection, has no effect on mobility save

a delayed time between DSB induction and the increase in

movement. These data suggest that resection, which is delayed,

but not abolished, in a sae2mutant, is required for the enhanced

mobility of DSBs. Moreover, Rad51 and Rad54, two proteins

that work downstream of the resection step, are required for

full induction of DSB mobility but have no effect on the mobility

of an undamaged locus (Dion et al., 2012; Mine-Hattab and

Rothstein, 2012). Rad54 is a SNF2-type ATPase, like INO80,

that functions in assisting strand invasion during homologous

recombination (Ceballos and Heyer, 2011). As is the case for

INO80 targeting, the role of Rad54 requires its remodeling

activity; a point mutant that abolishes the Rad54 ATPase activity

has the same effects as a full deletion (Dion et al., 2012).

Impairing DSB repair is not the onlyway to decrease themove-

ment of damaged chromatin. Indeed, mutating upstream

components of the DNA damage response (DDR), Mec1 and

Rad9 (ATRand53BP1 inmammals), abolish the enhancedmove-

ment of DSBs (Dion et al., 2012). In contrast, deletion of the

downstream kinase, Rad53 (homologues of mammalian ATR

and 53BP1), does not, suggesting that downstream checkpoint

functions do not regulate DSB mobility (Dion et al., 2012). It is

possible that Mec1/ATR activation is required to modify another

protein that acts directly on chromatin to enhance movement.

We note that INO80 components are direct targets ofMec1 (Mor-

rison et al., 2007). Another possibility is that the DDR modifies

chromatin itself, for example, by phosphorylating H2A (H2AX in

mammals). This would in turn recruit remodelers, such as

INO80, and scaffold proteins, including Rad9. In this way, the

checkpoint kinase could change the properties of chromatin

and enhance its movement by triggering a cascade of events.

The DDR also seems to affect DSB mobility in mammalian

cells. Indeed, Dimitrova et al. (2008) and colleagues have shown

that, when telomeres are uncapped and therefore readily

confused with DSBs, their movement increases in a 53BP1-

dependent manner. ATM, a key DDR regulatory kinase, is also

involved in this, as ATM null cells have uncapped telomeres

that show lower mobility (Dimitrova et al., 2008), and chemical

inhibition of ATM results in a similarly reduced Rc in human cells

(Krawczyk et al., 2012). Taken together, these results provide

evidence that DSBmovement requires DNA damage checkpoint

kinases in yeast, mouse, and human.

Consistent with the idea that chromatin remodeling contrib-

utes to DNA mobility, the deletion of arp8, which impairs

INO80-dependent remodeling, also leads to decreased mobility

of a DSB (Neumann et al., 2012). Although the effect was

partial, other studies in cultured human cells find that inhibition

of either histone deacetylases (HDACs) or histone acetyl-

transferases (HAT) also reduces Rc values for damaged DNA

(Krawczyk et al., 2012). Although the exact enzymes responsible

for these effects are not known, the results suggest that

chromatin structure may be an important regulator of the

mobility of a damaged chromosomal locus, both in mammals

and in yeast.

Chromatin Mobility, Homology Search, and DSB RepairDSBs can be repaired by homologous recombination (HR) or

nonhomologous end joining (NHEJ). In yeast, the primary repair

pathway is HR, whereas, in mammals, NHEJ predominates.

During HR, a DSB needs to ‘‘search’’ for an identical template

for repair (Barzel and Kupiec, 2008; Gehlen et al., 2011). Often,

a DSB is repaired by exchange with its identical sister chromatid,

which is synthesized during S phase because the damaged site

and undamaged template are held together by cohesin. This

leads to largely error-free repair. In diploid cells, the homologous

chromosome can also be used as a template, although this is

riskier, as the cell can lose heterozygosity upon repair. In the

rare cases in which the sister chromatid is not available as

a template, a long-range search for a homologous sequence

may be needed, for example, when a DSB occurs before replica-

tion or if the sister is also broken. A well-studied example of this

is the repair of a regulated DSB at the budding yeastMAT locus,

which encodes mating type information. Gene conversion of the

cleavedMAT locus by one of two templates,HML orHMR, found

at the ends of the same chromosome, allows yeast to switch its

mating type as often as once per cell cycle (Haber, 2012).

Although the HM loci are preferentially used as donors, HR can

also occur with a template on another chromosome (Agmon

et al., 2009; Ira et al., 2003; Keogh et al., 2006).

The search for templates on other chromosomes occurs

slowly, but approaches 100% efficiency over extended periods

of time (Aylon and Kupiec, 2003; Dion et al., 2012). The question

in all cases, however, is how the DSB finds its homologous

partner in a vast excess of nonhomologous sequence. Although

the homology search has been established as a major rate-

limiting step in HR in yeast (Wilson et al., 1994), the process itself

remains poorly understood. It seems likely that chromatin

movement is involved, given the requirement for cut site and

template to meet (Gehlen et al., 2011).

Recent studies show that the kinetics and efficiency of repair

by recombination correlated positively with DNA mobility. For

instance, targeting INO80 subunits to ectopic recombination

substrate in yeast increases the rates of homologous recombi-

nation (Neumann et al., 2012). Conversely, in rad9 mutants,

which have more restricted DSB mobility, the appearance of

recombination intermediates is delayed (Dion et al., 2012). This

effect is not due to the role of Rad9 in arresting the cell cycle

(Weinert and Hartwell, 1988) or in repressing resection (Chen

et al., 2012; Lazzaro et al., 2008; Ngo and Lydall, 2010).

Moreover, the delayed kinetics of MAT recombination in rad9

mutants was seen only when recombination templates were

found on an unlinked chromosome—that is, not when repair

was effected by recombination with templates in cis—arguing

that the long-range search is specifically limited by DNAmobility

(Dion et al., 2012). In mouse embryonic stem cells, one of Rad9’s

orthologs, 53BP1, is required for both telomere fusion (i.e., repair

by NHEJ) and the mobility of uncapped telomeres (Dimitrova

et al., 2008). Even though these data are largely correlative, we

speculate that enhanced mobility facilitates DNA repair in both

yeast and higher eukaryotes.

The movement of DNA damage could also be harnessed for

other purposes. For example, in yeast cells, persistent DSBs

are recruited to the nuclear periphery for processing,

whereas DSBs that can be repaired by HR are predominantly

found in the center of the nucleus (Bystricky et al., 2009; Nagai

et al., 2008). The relocalization of DSBs to different compart-

ments of the nucleus requires that chromatin is mobile, although

it is unclear whether mobility is rate limiting for the accumulation

of DSBs at the nuclear periphery. Indeed, rad9-deficient cells

have little difficulty shifting DSBs to the nuclear periphery (Nagai

et al., 2008), even though their mobility is low (Dion et al., 2012). It

remains to be seen whether other factors involved in the periph-

eral recruitment of DSBs impact their mobility—for instance, the

histone variant Htz1 (H2A.Z) (Kalocsay et al., 2009) or the

conserved SUN domain protein Mps3 (Oza et al., 2009).

In 2007, a yeast study showed that DSBs that occur within the

ribosomal DNA (rDNA) accumulate outside of the nucleolus

(Torres-Rosell et al., 2007). The exclusion of DSBs from the

nucleolus depends on two cohesin-like factors, Smc5 and

Smc6. DNA mobility in this case could facilitate the change in

nuclear location. Importantly, a similar study using live-cell

imaging in Drosophila cells showed that DSBs induced by

ionizing radiation are eventually excluded from large heterochro-

matic domains (Chiolo et al., 2011). Here again, there is a require-

ment for Smc5 and Smc6, suggesting that a similar mechanism

functions in yeast and flies. Strikingly, many of the factors

involved in DSB mobility in yeast are also implicated in the

movement of damage away from a heterochromatic domain,

including the DDR and the HR machinery (Chiolo et al., 2011;

Dion et al., 2012).

In cultured human cells, ionizing radiation can be delivered in

linear tracks, and mobility can be inferred by fixing the cells at

different time points after damage induction and scoring defor-

mities in the track path. By marking the damage path with an

antibody against the phosphorylated form of H2AX, it was shown

that the track curves around heterochromatin domains, suggest-

ing that the DNA damage occurs within the domain but is then

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1359

Figure 2. Chromatin Movement Driven by Nucleosome RemodelingA model for how remodeling-based nucleosome eviction might drive chro-matin movement to impact both transcription and DSB repair, adapted from(Neumann et al., 2012). Chromatin can be thought of as a polymer chain of stiffsegments interspersed by flexible linkers. The stiffness of the overall fiber isdetermined by its persistence length, which is defined as the length of thepolymer over which there is no apparent change in direction (i.e., no bending).Thus, the larger the persistence length is, the stiffer the fiber. We propose thatthe remodeling that occurs during transcriptional activation or during theprocessing of DSBs can enhance movement by inserting a flexible linker intoa stiff chromatin domain. In other words, the persistence length of the chro-matin domain will be smaller due to nucleosome removal in the middle of thedomain. The extra flexibility will, in turn, increase the volume in which a locuscan move. This can be harnessed either to enhance HR with an ectopic donorsequence or to reach a nuclear compartment conducive for transcription,repair, splicing, or export.

excluded from heterochromatin during its repair (Jakob et al.,

2011). It should be noted that a haploid yeast nucleus has an

average diameter of 1.8 mm (Heun et al., 2001b), which is similar

to the size of a single heterochromatic domain in human cells.

Thus, chromatin movement on the scale seen in budding yeast

may be relevant to the exclusion of damaged DNA from densely

packaged heterochromatin, as described in higher eukaryotes.

Chromatin Mobility: A Double-Edged Sword for GenomeStabilityBased on the studies summarized here, we propose a model in

which chromatin remodeling activities that accompany DSB

repair can be harnessed to promote recombination with ectopic

sequences and/or to move away from nuclear compartments

that are refractory to repair (Figure 2). We propose that this

movement derives from chromatin-remodeling enzymes and is

regulated by the DNA repair machinery and the DNA damage

response. Long-range movement, in the order of 1 mm in yeast,

of a troublesomeDNA lesion would thus promote its repair by HR

and suppress the lethality provoked by an irreparable DSB (Ben-

nett et al., 1993). On the other hand, it could lead to loss of

heterozygosity and translocations if not properly regulated.

Damage movement is thus a double-edged sword that needs

careful regulation to avoid genomic rearrangements.We imagine

that this balance is kept by the DDR that will only modify specific

downstream targets when the damage is severe enough to

1360 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

require a genome-wide search of a template. This model

predicts that different types of DNA damage could lead to

different modes of movement, depending on how the lesion is

sensed and repaired. Indeed, in haploid yeast, spontaneous

repair centers marked by Rad52 are confined to �6% of the

nuclear volume as compared to �15% for a single protein-

DNA adduct and nearly 50% for a DSB (Dion et al., 2012).

This model conforms well to the data obtained in yeast, but

obvious problems exist in the case of mammalian cells. Clearly,

given that the size of an average mammalian nucleus is much

larger than a nucleus in yeast (on average, 200- to 400-fold in

volume), much more movement would be required to explore

nuclear space. Nonetheless, some aspects of the model may

hold true. Specifically, it was shown in mammalian cells that

different damaging agents lead to different degrees of repair

center mobility—that is, topoisomerase-II-dependent DNA

breaks move within a larger radius of constraint than IR-induced

damage (Krawczyk et al., 2012). This lesion-specific character

of repair focus mobility may account for the differences seen

in studies of mammalian chromatin movement; each study

analyzed a different type of DNA damage. Given that many

secondary tumors arise from chromosomal translocations in

cancers treated with DNA-damaging drugs, it may be valuable

to consider, in the design of therapeutic protocols, the diffusion

properties of the different types of DNA lesions induced.

Open QuestionsAs the mechanisms behind chromatin mobility start to be un-

tangled, a number of major questions remain.

Is There a Cause-Effect Relationship between

Chromatin Mobility and DSB Repair or Transcription?

The data obtained so far on the relationship between DSB repair

and transcription, on one hand, and chromatin mobility, on the

other, are largely correlative. One experiment that could estab-

lish causation would be to visualize the homology search step

live and, at the same time, to target a remodeler that enhances

movement to the template site and see whether the pairing itself

occurs faster in these conditions. A similar experiment could be

done in the case of uncapped telomeres in mammalian systems

to ask whether they encounter each other at higher frequencies

when there is more movement (for instance, in the absence of

53BP1). In the context of transcription, one can imagine

following the mobility of a locus and, at the same time, the tran-

scriptional output. The cell line to do this experiment is available

already (Janicki et al., 2004). In this assay, the locus is tagged

with a lacO array, and the RNA is tagged with a MS2 binding

consensus, which is bound by the bacteriophage protein MS2

fused to a fluorescent protein. The mobility of the DNA locus

can be visualized while obtaining real-time quantitative data

on transcriptional output. Such studies, although challenging,

will certainly yield interesting results.

If the Induction of Chromatin Movement Is Intrinsic to

a Subset of Nucleosome Remodelers, which of Their

Functions Actually Drives Locus Mobility?

Nucleosome remodeling appears to influence the movement of

both damaged and undamaged chromatin. However, the

changes that drive this movement remain unclear. We proposed

that the displacement of nucleosomes by means of a chromatin

remodeler leads to a more flexible chromatin by disrupting the

structure of the chromatin fiber. This may lead to a smaller

persistence length, given that an open linker would be intro-

duced in the midst of a higher-order structure (Neumann et al.,

2012) (Figure 2). Targeting different chromatin remodelers

with different biochemical activities and interrogating their

effects on nucleosome positioning near the target site may

help us decipher the mechanism through which remodelers

influence chromatin mobility. The generation or disruption of

higher-order chromatin structures may respectively restrain or

promote locus mobility.

Our model would vary in its details depending on which mode

of folding is adopted the nucleosome fiber (Grigoryev andWood-

cock, 2012). There is not enough information at the moment to

identify which specific changes to chromatin structure would

enhance or restrainmobility. The simplest biophysical parameter

that could account for movement of a polymer fiber, however, is

the flexibility of the fiber.

What Is the Role of Cohesin in the Movement of Repair

Foci?

Cohesin holds sister chromatids together (Nasmyth, 2011) and

accumulates at sites of DSBs (Unal et al., 2004). These charac-

teristics make it an appealing candidate to help control the

movement of DSBs, especially for those that occur spontane-

ously during replication or when forks encounter protein-DNA

adducts. In these cases, the template for HR-mediated repair

is readily available in the form of an undamaged sister chromatid

held in place by cohesin. Although this maywell restrain mobility,

cohesin is clearly not the only factor that restricts movement at

DSBs. In certain conditions that allow the visualization of

damaged DNA in G1 phase cells, the constraint on damage

mobility is nearly identical to that observed in S phase cells,

even though there is no sister-sister cohesion in G1 (Dion

et al., 2012). Determining what effect chromatid-chromatid

linkage through cohesin has on chromatin dynamics will go

a long way toward elucidating the regulation of chromatin move-

ment and its controlled release.

Is Enhanced Mobility Restricted to Sites of Damage in

Yeast? If Not, What Drives Genome-wide Changes in

Mobility, and What Is Its Purpose during Repair? Is This

Found in Other Organisms?

Mine-Hattab and Rothstein (2012) showed that, upon DSB

induction in budding yeast, there is also an increase in chromatin

mobility at unrelated, undamaged loci. This is likely to depend on

the dosage of damage incurred because a single DSB does not

cause a similar increase at an ectopic locus in haploid cells (Dion

et al., 2012). Although Mine-Hattab and Rothstein monitored the

increase in movement in diploid cells, it is difficult to imagine

mechanisms regulating chromatin mobility that are ploidy

specific. Reconciling divergent results, we propose that a

threshold of damage is necessary to provoke Mec1/ATR activa-

tion. Given that a DNA checkpoint response is necessary for the

increase in mobility at the break itself, its propagation to other

sites may be dose dependent, requiring activation of the DDR.

If, indeed, sufficient damage enhances chromatin movement

genome wide, then it is possible that the checkpoint kinase

Mec1/ATR and its downstream cascade are directly implicated

in this phenomenon. Further work with appropriate mutants is

needed to identify what signals a global increase in chromatin

movement in response to DNA damage. In addition to a check-

point signaling cascade, it is conceivable that ectopic movement

might also depend on chromatin remodelers or histonemodifica-

tions. Finally, whatever the mechanism may be, it will be impor-

tant to examine whether a global increase in chromatin mobility

has functional implications for DSB repair, such as promoting the

homology search required for HR (Mine-Hattab and Rothstein,

2012).

What Is the Contribution of DNA Mobility to the Genesis

of Translocations?

There have been two models put forth to account for the gener-

ation of recurrent carcinogenic translocations in humans, called

the ‘‘breakage first’’ and ‘‘contact first’’ models (Savage, 1996).

The first model posits that breaks must occur first and then will

roam throughout the nucleus until they find each other, leading

to translocation between distant sites. The contact first model,

on the other hand, predicts that the two breaks needed for

a translocation will occur preferentially on juxtaposed chromo-

somes. Quite naturally, after breakage, these two sites would re-

combine at higher frequencies. Both of these models are

extreme scenarios. Although it is unlikely that DSBs can explore

an entire mammalian nucleus, given its 200 to 400 times larger

volume than a typical yeast nucleus, it is also unlikely that

mobility has no influence whatsoever on which DNA ends are

ligated to each other.

Richardson and Jasin (2000) showed unambiguously that two

DSBs must occur before a translocation can be generated. It

seems obvious that, even if DSBs are extremely mobile, they

are still more likely to encounter break sites that occur close to

their starting point rather than those that are further away.

Indeed, recent large-scale studies have confirmed this, demon-

strating that translocations tend to occur between sites that are

spatially juxtaposed in the nucleus (Hakim et al., 2012; Zhang

et al., 2012). Arguing in favor of movement, on the other hand,

Spehalski et al. (2012) showed that Myc-Igh translocations in

mouse cells occur at the same frequency regardless of where

the Igh is placed in the genome. Understanding what regulates

DSBmovement and its impact on specific recombination events

is clearly important for understanding oncogenic translocations.

We note, however, that there may be other reasons that

damaged sites move. It may be important that a break moves

far enough to encounter a nuclear compartment that favors

repair or to move away from an environment rich in repetitive

elements. Moving too far, on the other hand, may generate dele-

terious recombination events. The mechanisms that regulate

chromatin mobility may thus influence genome stability. It is an

intriguing thought that one might harness these observations

on how chromatin movement impacts chromosomal transloca-

tions to design cancer therapies that minimize treatment-

induced chromosome exchange.

ACKNOWLEDGMENTS

We thank Fisun Hamaratoglu, Michael Hauer, and Andrew Seeber for critical

reading of the manuscript. Work in the Gasser laboratory is supported by

the Novartis Research Foundation, the Swiss National Science Foundation,

and various EU Marie Curie Networks.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1361

REFERENCES

Agarwal, S., van Cappellen, W.A., Guenole, A., Eppink, B., Linsen, S.E.,

Meijering, E., Houtsmuller, A., Kanaar, R., and Essers, J. (2011). ATP-depen-

dent and independent functions of Rad54 in genome maintenance. J. Cell

Biol. 192, 735–750.

Agmon, N., Pur, S., Liefshitz, B., and Kupiec, M. (2009). Analysis of repair

mechanism choice during homologous recombination. Nucleic Acids Res.

37, 5081–5092.

Aten, J.A., Stap, J., Krawczyk, P.M., van Oven, C.H., Hoebe, R.A., Essers, J.,

and Kanaar, R. (2004). Dynamics of DNA double-strand breaks revealed by

clustering of damaged chromosome domains. Science 303, 92–95.

Aylon, Y., and Kupiec, M. (2003). The checkpoint protein Rad24 of Saccharo-

myces cerevisiae is involved in processing double-strand break ends and in

recombination partner choice. Mol. Cell. Biol. 23, 6585–6596.

Barbaric, S., Luckenbach, T., Schmid, A., Blaschke, D., Horz, W., and Korber,

P. (2007). Redundancy of chromatin remodeling pathways for the induction of

the yeast PHO5 promoter in vivo. J. Biol. Chem. 282, 27610–27621.

Barzel, A., and Kupiec, M. (2008). Finding a match: how do homologous

sequences get together for recombination? Nat. Rev. Genet. 9, 27–37.

Bennett, C.B., Lewis, A.L., Baldwin, K.K., and Resnick, M.A. (1993). Lethality

induced by a single site-specific double-strand break in a dispensable yeast

plasmid. Proc. Natl. Acad. Sci. USA 90, 5613–5617.

Berg, H.C. (1993). Random Walks in Biology, Expanded Edition (Princeton,

New Jersey: Princeton University Press).

Bornfleth, H., Edelmann, P., Zink, D., Cremer, T., and Cremer, C. (1999).

Quantitative motion analysis of subchromosomal foci in living cells using

four-dimensional microscopy. Biophys. J. 77, 2871–2886.

Bystricky, K., Van Attikum, H., Montiel, M.D., Dion, V., Gehlen, L., and Gasser,

S.M. (2009). Regulation of nuclear positioning and dynamics of the silent

mating type loci by the yeast Ku70/Ku80 complex. Mol. Cell. Biol. 29, 835–848.

Cabal, G.G., Genovesio, A., Rodriguez-Navarro, S., Zimmer, C., Gadal, O.,

Lesne, A., Buc, H., Feuerbach-Fournier, F., Olivo-Marin, J.C., Hurt, E.C., and

Nehrbass, U. (2006). SAGA interacting factors confine sub-diffusion of

transcribed genes to the nuclear envelope. Nature 441, 770–773.

Campos, E.I., and Reinberg, D. (2009). Histones: annotating chromatin. Annu.

Rev. Genet. 43, 559–599.

Casolari, J.M., Brown, C.R., Komili, S., West, J., Hieronymus, H., and Silver,

P.A. (2004). Genome-wide localization of the nuclear transport machinery

couples transcriptional status and nuclear organization. Cell 117, 427–439.

Ceballos, S.J., and Heyer, W.D. (2011). Functions of the Snf2/Swi2 family

Rad54 motor protein in homologous recombination. Biochim. Biophys. Acta

1809, 509–523.

Chen, X., Cui, D., Papusha, A., Zhang, X., Chu, C.D., Tang, J., Chen, K., Pan,

X., and Ira, G. (2012). The Fun30 nucleosome remodeller promotes resection of

DNA double-strand break ends. Nature 489, 576–580.

Chiolo, I., Minoda, A., Colmenares, S.U., Polyzos, A., Costes, S.V., and

Karpen, G.H. (2011). Double-strand breaks in heterochromatin move outside

of a dynamic HP1a domain to complete recombinational repair. Cell 144,

732–744.

Chuang, C.H., Carpenter, A.E., Fuchsova, B., Johnson, T., de Lanerolle, P.,

and Belmont, A.S. (2006). Long-range directional movement of an interphase

chromosome site. Curr. Biol. 16, 825–831.

Chubb, J.R., Boyle, S., Perry, P., and Bickmore, W.A. (2002). Chromatin

motion is constrained by association with nuclear compartments in human

cells. Curr. Biol. 12, 439–445.

Clapier, C.R., and Cairns, B.R. (2009). The biology of chromatin remodeling

complexes. Annu. Rev. Biochem. 78, 273–304.

Dimitrova, N., Chen, Y.C., Spector, D.L., and de Lange, T. (2008). 53BP1

promotes non-homologous end joining of telomeres by increasing chromatin

mobility. Nature 456, 524–528.

1362 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Dion, V., Shimada, K., and Gasser, S.M. (2010). Actin-related proteins in the

nucleus: life beyond chromatin remodelers. Curr. Opin. Cell Biol. 22, 383–391.

Dion, V., Kalck, V., Horigome, C., Towbin, B.D., and Gasser, S.M. (2012).

Increased mobility of double-strand breaks requires Mec1, Rad9 and the

homologous recombination machinery. Nat. Cell Biol. 14, 502–509.

Egecioglu, D., and Brickner, J.H. (2011). Gene positioning and expression.

Curr. Opin. Cell Biol. 23, 338–345.

Flaus, A., Martin, D.M., Barton, G.J., and Owen-Hughes, T. (2006). Identifica-

tion of multiple distinct Snf2 subfamilies with conserved structural motifs.

Nucleic Acids Res. 34, 2887–2905.

Gartenberg, M.R., Neumann, F.R., Laroche, T., Blaszczyk, M., and Gasser,

S.M. (2004). Sir-mediated repression can occur independently of chromo-

somal and subnuclear contexts. Cell 119, 955–967.

Gehlen, L., Gasser, S.M., and Dion, V. (2011). How broken DNA finds

its template for repair: a computational approach. Prog. Theor. Phys.

191(Suppl. 191), 20–29.

Gotta, M., Laroche, T., Formenton, A., Maillet, L., Scherthan, H., and Gasser,

S.M. (1996). The clustering of telomeres and colocalization with Rap1, Sir3,

and Sir4 proteins in wild-type Saccharomyces cerevisiae. J. Cell Biol. 134,

1349–1363.

Grigoryev, S.A., and Woodcock, C.L. (2012). Chromatin organization - the

30 nm fiber. Exp. Cell Res. 318, 1448–1455.

Haaf, T., Golub, E.I., Reddy, G., Radding, C.M., andWard, D.C. (1995). Nuclear

foci of mammalian Rad51 recombination protein in somatic cells after DNA

damage and its localization in synaptonemal complexes. Proc. Natl. Acad.

Sci. USA 92, 2298–2302.

Haber, J.E. (2012). Mating-type genes and MAT switching in Saccharomyces

cerevisiae. Genetics 191, 33–64.

Hakim, O., Resch, W., Yamane, A., Klein, I., Kieffer-Kwon, K.R., Jankovic, M.,

Oliveira, T., Bothmer, A., Voss, T.C., Ansarah-Sobrinho, C., et al. (2012). DNA

damage defines sites of recurrent chromosomal translocations in B lympho-

cytes. Nature 484, 69–74.

Hediger, F., Neumann, F.R., Van Houwe, G., Dubrana, K., and Gasser, S.M.

(2002). Live imaging of telomeres: yKu and Sir proteins define redundant telo-

mere-anchoring pathways in yeast. Curr. Biol. 12, 2076–2089.

Hediger, F., Berthiau, A.S., van Houwe, G., Gilson, E., andGasser, S.M. (2006).

Subtelomeric factors antagonize telomere anchoring and Tel1-independent

telomere length regulation. EMBO J. 25, 857–867.

Heun, P., Laroche, T., Raghuraman, M.K., and Gasser, S.M. (2001a). The

positioning and dynamics of origins of replication in the budding yeast nucleus.

J. Cell Biol. 152, 385–400.

Heun, P., Laroche, T., Shimada, K., Furrer, P., and Gasser, S.M. (2001b). Chro-

mosome dynamics in the yeast interphase nucleus. Science 294, 2181–2186.

Ira, G., Malkova, A., Liberi, G., Foiani, M., and Haber, J.E. (2003). Srs2 and

Sgs1-Top3 suppress crossovers during double-strand break repair in yeast.

Cell 115, 401–411.

Jakob, B., Splinter, J., Durante, M., and Taucher-Scholz, G. (2009a). Live cell

microscopy analysis of radiation-induced DNA double-strand break motion.

Proc. Natl. Acad. Sci. USA 106, 3172–3177.

Jakob, B., Splinter, J., and Taucher-Scholz, G. (2009b). Positional stability

of damaged chromatin domains along radiation tracks in mammalian cells.

Radiat. Res. 171, 405–418.

Jakob, B., Splinter, J., Conrad, S., Voss, K.O., Zink, D., Durante, M., Lobrich,

M., and Taucher-Scholz, G. (2011). DNA double-strand breaks in heterochro-

matin elicit fast repair protein recruitment, histone H2AX phosphorylation and

relocation to euchromatin. Nucleic Acids Res. 39, 6489–6499.

Janicki, S.M., Tsukamoto, T., Salghetti, S.E., Tansey, W.P., Sachidanandam,

R., Prasanth, K.V., Ried, T., Shav-Tal, Y., Bertrand, E., Singer, R.H., and Spec-

tor, D.L. (2004). From silencing to gene expression: real-time analysis in single

cells. Cell 116, 683–698.

Kalocsay, M., Hiller, N.J., and Jentsch, S. (2009). Chromosome-wide Rad51

spreading and SUMO-H2A.Z-dependent chromosome fixation in response

to a persistent DNA double-strand break. Mol. Cell 33, 335–343.

Keogh, M.C., Kim, J.A., Downey, M., Fillingham, J., Chowdhury, D., Harrison,

J.C., Onishi, M., Datta, N., Galicia, S., Emili, A., et al. (2006). A phosphatase

complex that dephosphorylates gammaH2AX regulates DNA damage check-

point recovery. Nature 439, 497–501.

Krawczyk, P.M., Borovski, T., Stap, J., Cijsouw, T., ten Cate, R., Medema, J.P.,

Kanaar, R., Franken, N.A., and Aten, J.A. (2012). Chromatin mobility is

increased at sites of DNA double-strand breaks. J. Cell Sci. 125, 2127–2133.

Kruhlak, M.J., Celeste, A., Dellaire, G., Fernandez-Capetillo, O., Muller, W.G.,

McNally, J.G., Bazett-Jones, D.P., and Nussenzweig, A. (2006). Changes in

chromatin structure and mobility in living cells at sites of DNA double-strand

breaks. J. Cell Biol. 172, 823–834.

Lans, H., Marteijn, J.A., and Vermeulen, W. (2012). ATP-dependent chromatin

remodeling in the DNA-damage response. Epigenetics Chromatin 5, 4.

Lazzaro, F., Sapountzi, V., Granata, M., Pellicioli, A., Vaze, M., Haber, J.E.,

Plevani, P., Lydall, D., and Muzi-Falconi, M. (2008). Histone methyltransferase

Dot1 and Rad9 inhibit single-stranded DNA accumulation at DSBs and uncap-

ped telomeres. EMBO J. 27, 1502–1512.

Marshall, W.F., Straight, A., Marko, J.F., Swedlow, J., Dernburg, A., Belmont,

A., Murray, A.W., Agard, D.A., and Sedat, J.W. (1997). Interphase chromo-

somes undergo constrained diffusional motion in living cells. Curr. Biol. 7,

930–939.

Michaelis, C., Ciosk, R., and Nasmyth, K. (1997). Cohesins: chromosomal

proteins that prevent premature separation of sister chromatids. Cell 91,

35–45.

Mine-Hattab, J., and Rothstein, R. (2012). Increased chromosome mobility

facilitates homology search during recombination. Nat. Cell Biol. 14, 510–517.

Molenaar, C., Wiesmeijer, K., Verwoerd, N.P., Khazen, S., Eils, R., Tanke, H.J.,

and Dirks, R.W. (2003). Visualizing telomere dynamics in living mammalian

cells using PNA probes. EMBO J. 22, 6631–6641.

Morrison, A.J., Highland, J., Krogan, N.J., Arbel-Eden, A., Greenblatt, J.F.,

Haber, J.E., and Shen, X. (2004). INO80 and gamma-H2AX interaction links

ATP-dependent chromatin remodeling to DNA damage repair. Cell 119,

767–775.

Morrison, A.J., Kim, J.A., Person, M.D., Highland, J., Xiao, J., Wehr, T.S.,

Hensley, S., Bao, Y., Shen, J., Collins, S.R., et al. (2007). Mec1/Tel1 phosphor-

ylation of the INO80 chromatin remodeling complex influences DNA damage

checkpoint responses. Cell 130, 499–511.

Nagai, S., Dubrana, K., Tsai-Pflugfelder, M., Davidson, M.B., Roberts, T.M.,

Brown, G.W., Varela, E., Hediger, F., Gasser, S.M., and Krogan, N.J. (2008).

Functional targeting of DNA damage to a nuclear pore-associated SUMO-

dependent ubiquitin ligase. Science 322, 597–602.

Nasmyth, K. (2011). Cohesin: a catenase with separate entry and exit gates?

Nat. Cell Biol. 13, 1170–1177.

Nelms, B.E., Maser, R.S., MacKay, J.F., Lagally, M.G., and Petrini, J.H. (1998).

In situ visualization of DNA double-strand break repair in human fibroblasts.

Science 280, 590–592.

Neumann, F.R., Dion, V., Gehlen, L.R., Tsai-Pflugfelder, M., Schmid, R.,

Taddei, A., and Gasser, S.M. (2012). Targeted INO80 enhances subnuclear

chromatin movement and ectopic homologous recombination. Genes Dev.

26, 369–383.

Ngo, H.P., and Lydall, D. (2010). Survival and growth of yeast without telomere

capping by Cdc13 in the absence of Sgs1, Exo1, and Rad9. PLoS Genet. 6,

e1001072.

Nielsen, I., Bentsen, I.B., Lisby,M., Hansen, S., Mundbjerg, K., Andersen, A.H.,

and Bjergbaek, L. (2009). A Flp-nick system to study repair of a single protein-

bound nick in vivo. Nat. Methods 6, 753–757.

Oza, P., Jaspersen, S.L., Miele, A., Dekker, J., and Peterson, C.L. (2009).

Mechanisms that regulate localization of a DNA double-strand break to the

nuclear periphery. Genes Dev. 23, 912–927.

Pliss, A., Malyavantham, K., Bhattacharya, S., Zeitz, M., and Berezney, R.

(2009). Chromatin dynamics is correlatedwith replication timing. Chromosoma

118, 459–470.

Richardson, C., and Jasin, M. (2000). Frequent chromosomal translocations

induced by DNA double-strand breaks. Nature 405, 697–700.

Robinett, C.C., Straight, A., Li, G., Willhelm, C., Sudlow, G., Murray, A., and

Belmont, A.S. (1996). In vivo localization of DNA sequences and visualization

of large-scale chromatin organization using lac operator/repressor recogni-

tion. J. Cell Biol. 135, 1685–1700.

Savage, J.R. (1996). Insight into sites. Mutat. Res. 366, 81–95.

Segal, E., and Widom, J. (2009). What controls nucleosome positions? Trends

Genet. 25, 335–343.

Soutoglou, E., Dorn, J.F., Sengupta, K., Jasin, M., Nussenzweig, A., Ried, T.,

Danuser, G., and Misteli, T. (2007). Positional stability of single double-strand

breaks in mammalian cells. Nat. Cell Biol. 9, 675–682.

Spehalski, E., Kovalchuk, A.L., Collins, J.T., Liang, G., Dubois, W., Morse,

H.C., 3rd, Ferguson, D.O., Casellas, R., and Dunnick, W.A. (2012). Oncogenic

Myc translocations are independent of chromosomal location and orientation

of the immunoglobulin heavy chain locus. Proc. Natl. Acad. Sci. USA 109,

13728–13732.

Steger, D.J., Haswell, E.S., Miller, A.L., Wente, S.R., and O’Shea, E.K. (2003).

Regulation of chromatin remodeling by inositol polyphosphates. Science 299,

114–116.

Taddei, A., and Gasser, S.M. (2012). Structure and function in the budding

yeast nucleus. Genetics 192, 107–129.

Taddei, A., Hediger, F., Neumann, F.R., Bauer, C., and Gasser, S.M. (2004).

Separation of silencing from perinuclear anchoring functions in yeast Ku80,

Sir4 and Esc1 proteins. EMBO J. 23, 1301–1312.

Taddei, A., Van Houwe, G., Hediger, F., Kalck, V., Cubizolles, F., Schober, H.,

and Gasser, S.M. (2006). Nuclear pore association confers optimal expression

levels for an inducible yeast gene. Nature 441, 774–778.

Taddei, A., Schober, H., and Gasser, S.M. (2010). The budding yeast nucleus.

Cold Spring Harb. Perspect. Biol. 2, a000612.

Torres-Rosell, J., Sunjevaric, I., De Piccoli, G., Sacher, M., Eckert-Boulet, N.,

Reid, R., Jentsch, S., Rothstein, R., Aragon, L., and Lisby, M. (2007). The

Smc5-Smc6 complex and SUMOmodification of Rad52 regulates recombina-

tional repair at the ribosomal gene locus. Nat. Cell Biol. 9, 923–931.

Tsukuda, T., Fleming, A.B., Nickoloff, J.A., and Osley, M.A. (2005). Chromatin

remodelling at a DNA double-strand break site in Saccharomyces cerevisiae.

Nature 438, 379–383.

Unal, E., Arbel-Eden, A., Sattler, U., Shroff, R., Lichten, M., Haber, J.E., and

Koshland, D. (2004). DNA damage response pathway uses histone modifica-

tion to assemble a double-strand break-specific cohesin domain. Mol. Cell

16, 991–1002.

van Attikum, H., Fritsch, O., Hohn, B., and Gasser, S.M. (2004). Recruitment of

the INO80 complex by H2A phosphorylation links ATP-dependent chromatin

remodeling with DNA double-strand break repair. Cell 119, 777–788.

Vazquez, J., Belmont, A.S., and Sedat, J.W. (2001). Multiple regimes of con-

strained chromosome motion are regulated in the interphase Drosophila

nucleus. Curr. Biol. 11, 1227–1239.

Walter, J., Schermelleh, L., Cremer, M., Tashiro, S., and Cremer, T. (2003).

Chromosome order in HeLa cells changes during mitosis and early G1, but

is stably maintained during subsequent interphase stages. J. Cell Biol. 160,

685–697.

Weber, S.C., Spakowitz, A.J., and Theriot, J.A. (2010). Bacterial chromosomal

loci move subdiffusively through a viscoelastic cytoplasm. Phys. Rev. Lett.

104, 238102.

Weber, S.C., Spakowitz, A.J., and Theriot, J.A. (2012). Nonthermal

ATP-dependent fluctuations contribute to the in vivo motion of chromosomal

loci. Proc. Natl. Acad. Sci. USA 109, 7338–7343.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1363

Weinert, T.A., and Hartwell, L.H. (1988). The RAD9 gene controls the cell

cycle response to DNA damage in Saccharomyces cerevisiae. Science 241,

317–322.

Wiesmeijer, K., Krouwels, I.M., Tanke, H.J., and Dirks, R.W. (2008). Chromatin

movement visualized with photoactivable GFP-labeled histone H4. Differenti-

ation 76, 83–90.

Wilson, J.H., Leung, W.Y., Bosco, G., Dieu, D., and Haber, J.E. (1994). The

frequency of gene targeting in yeast depends on the number of target copies.

Proc. Natl. Acad. Sci. USA 91, 177–181.

1364 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Zhang, Y., McCord, R.P., Ho, Y.J., Lajoie, B.R., Hildebrand, D.G., Simon, A.C.,

Becker, M.S., Alt, F.W., and Dekker, J. (2012). Spatial organization of the

mouse genome and its role in recurrent chromosomal translocations. Cell

148, 908–921.

Zimmer, C., and Fabre, E. (2011). Principles of chromosomal organization:

lessons from yeast. J. Cell Biol. 192, 723–733.

Zink, D., Cremer, T., Saffrich, R., Fischer, R., Trendelenburg, M.F., Ansorge,

W., and Stelzer, E.H. (1998). Structure and dynamics of human interphase

chromosome territories in vivo. Hum. Genet. 102, 241–251.

Leading Edge

Review

When Lamins Go Bad:Nuclear Structure and Disease

Katherine H. Schreiber1 and Brian K. Kennedy1,2,*1Buck Institute for Research on Aging, 8001 Redwood Boulevard, Novato, CA 94945, USA2Aging Research Institute, Guangdong Medical College, Dongguan 523808, Guangdong, China*Correspondence: bkennedy@buckinstitute.org

http://dx.doi.org/10.1016/j.cell.2013.02.015

Mutations in nuclear lamins or other proteins of the nuclear envelope are the root cause of a group ofphenotypically diverse genetic disorders known as laminopathies, which have symptoms thatrange from muscular dystrophy to neuropathy to premature aging syndromes. Although precisedisease mechanisms remain unclear, there has been substantial progress in our understandingof not only laminopathies, but also the biological roles of nuclear structure. Nuclear envelopedysfunction is associated with altered nuclear activity, impaired structural dynamics, and aberrantcell signaling. Building on these findings, small molecules are being discovered that may becomeeffective therapeutic agents.

IntroductionSince their discovery more than 35 years ago as constituents of

the nuclear lamina (Gerace et al., 1978), the nuclear lamins have

been the subject of intense speculation regarding their possible

roles in almost everything that happens in the nucleus. Early

studies focused on biochemistry and cell biology, with the goal

of achieving a basic understanding of the principles governing

nuclear organization. The nuclear envelope entered the medical

realm in the mid-1990s, when mutations in emerin were identi-

fied in patients with Emery-Dreifuss muscular dystrophy

(EDMD) (Bione et al., 1994). The LMNA gene, encoding all

A-type nuclear lamins, was linked to EDMD a few years later

(Bonne et al., 1999), and links between nuclear structure and

human disease have been studied extensively since then in

labs throughout the world.

With around 15 diseases, including a range of dystrophic and

progeroid syndromes, attributed to LMNA mutations and muta-

tions in genes encoding associated nuclear envelope proteins

causing an overlapping set of diseases, the questions and exper-

imental approaches have evolved. Why do alterations in nuclear

envelope proteins confer disease? What are the mechanisms

underlying disease pathology? Do A-type lamins have a role in

normal aging? Can effective therapies be developed for these

debilitating diseases? Though a range of exciting discoveries

have been made in the last decade, many unknowns remain.

Here, we seek to frame the current questions, propose possible

paths toward mechanistic understanding, and briefly evaluate

the therapeutic possibilities that are starting to emerge. Given

the amount of interest and momentum in the lamin field, it is

feasible that therapies to rescue the pathogenic consequences

of misbehaved nuclear structural components will be developed

in the not-too-distant future.

Nuclear LaminsThe nuclear envelope is comprised of two membranes: the

outer nuclear membrane, which is continuous with the endo-

plasmic reticulum, and the inner nuclear membrane, which

associates with the nuclear lamina. Nuclear pore complexes

perforate the nuclear envelope to allow transport between the

cytoplasm and nucleus. The nuclear lamina is primarily

composed of nuclear lamins, which were originally identified

as lamins A, B, and C (Gerace et al., 1978). These proteins

constitute the only class of intermediate filament proteins in

the nucleus and form associated filamentous structures that

underlie the nuclear envelope and interact with neighboring

proteins (Gerace and Huber, 2012). Lamins A and C, as well

as two other variants (C2 and AD10), are classed as A-type

lamins and are encoded by the LMNA gene through alternative

splicing. Three different lamin B family members (B-type

lamins) are encoded by two genes (lamin B1 by LMNB1 and

lamins B2 and B3 by LMNB2).

A- and B-type lamins have fundamentally different properties,

perhaps most importantly by virtue of their different isoelectric

points, which dictate that B-type lamins stay associated with

the nuclear envelope during mitosis while A-type lamins become

soluble. Expression patterns differ as well, with B-type lamins

expressed in most or all cell types and A-type lamins expressed

during cell differentiation in many different developmental

lineages (Rober et al., 1989). At the cellular level, both classes

of proteins have been ascribed structural roles in the nucleus

as well as a range of other activities, including coordination of

transcription and replication. While specific functions of A-type

lamins remain somewhat elusive, a number of recent discoveries

point to key interactions between lamins and cell proliferation,

differentiation, and stress response pathways.

Both A- and B-type lamins undergo posttranslational pro-

cessing based on a C-terminal CaaX motif that dictates a series

of modifications (Weber et al., 1989); only lamin C avoids this by

virtue of alternative splicing of the LMNA transcript that lacks the

C terminus. As a first step, the cysteine residue is farnesylated.

Next, proteolytic processing leads to cleavage after the cysteine

residue, followed by a carboxymethylation of the new C-terminal

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1365

Table 1. Diseases Caused by Mutations in Genes Encoding

Lamins and Lamin-Associated Proteins

Gene Mutated

Striated Muscle Diseases

Emery-Dreifuss muscular dystrophy LMNA, EDMD, SYNE1,

SYNE2, TMEM43, TMPO

Limb-girdle muscular dystrophy LMNA

Dilated cardiomyopathy LMNA, EDMD, SYNE1,

SYNE2, TMEM43, TMPO

Congenital muscular dystrophy LMNA

Heart-hand syndrome LMNA

Lipodystrophy Syndromes

Dunnigan-type familial partial

lipodystrophy

LMNA

Mandibuloacral dysplasia LMNA, ZMPSTE24

Lipoatrophy LMNA

Partial lipodystrophy LMNB2

Accelerated Aging Disorders

Atypical Werner syndrome LMNA

Hutchinson-Gilford progeria syndrome LMNA

Restrictive dermopathy LMNA, ZMPSTE24

Atypical progeria syndrome BANF1

Peripheral Nerve Disorders

Charcot-Marie-Tooth disease LMNA

Adult-onset leukodystrophy LMNB1

Spinocerebellar ataxia type 8 SYNE1

Bone Diseases

Buschke-Ollendorff syndrome LEMD3

Melorheostosis LEMD3

Osteopoikilosis LEMD3

Greenberg skeletal dysplasia LBR

Other

Pelger-Huet anomaly LBR

Arthrogryposis SYNE2

residue. Many membrane-associated proteins, including Ras,

undergo this processing event. However, in the case of lamin

A, isoprenylation is a transient event, as a second proteolytic

event mediated by the zinc metalloproteinase Zmpste24 leads

to excision of another 15 amino acids. Due to this cleavage,

mature lamin A lacks the modified cysteine. This process is

clearly important to pathologic states, as laminopathies are

linked to altered processing of lamin A, as well as loss-of-func-

tion mutations in ZMPSTE24.

The reasons for farnesylation of lamin A remain to be eluci-

dated despite extensive efforts. Until recently, the thinking has

been that the transient farnesylation event was needed, through

association of the hydrophobic farnesyl group with the nuclear

envelope, to provide initial recruitment of lamin A to the nuclear

periphery (Hennekes and Nigg, 1994). After assembly into fila-

ments, farnesylation of lamin A may no longer be required.

Consistent with this hypothesis, the nucleus has been shown

to be the site of both lamin A carboxymethylation and proteolytic

cleavage by ZMPSTE24. However, several recent studies using

mice and/or cells engineered to express mutant forms of lamin A

indicate that farnesylation is not required for recruitment (Davies

et al., 2011). For instance, when only a nonfarnesylated version

of lamin A is expressed, normal localization of the lamin A variant

to the nuclear periphery was observed (Davies et al., 2010; Lee

et al., 2010), although mice generated in this manner develop

cardiomyopathy (see below). In addition, mice expressing only

lamin C (not farnesylated) or a mature (preprocessed) lamin A

are (surprisingly) normal and have apparently correct localization

of the respective protein to the nuclear periphery. Though these

studies do not preclude amore subtle role for lamin A processing

in filament assembly or envelope association, they raise serious

questions about the importance of these events in the mouse

and provide an interesting puzzle to be pieced together by future

studies.

Diseases Linked to Mutations in Nuclear StructureProteinsThe number of different diseases linked tomutations in LMNA, at

least 15 by now, surpasses that of any other human gene. It is

hard to establish absolute numbers because many of the

associated syndromes have overlapping pathologies. Neverthe-

less, the range of tissues and functions that can be adversely

affected by mutation in LMNA is striking (Table 1). Diseases

include the aforementioned Emery-Dreifuss muscular dystrophy

(EDMD2/3) (Bonne et al., 1999) and a second muscular

dystrophy (Limb-girdle, LGMD1B) (Muchir et al., 2000) that

affects different skeletal muscle groups. Patients with both forms

of muscular dystrophy also present with dilated cardiomyop-

athy, which is often the cause of mortality. Other LMNA muta-

tions lead to dilated cardiomyopathy (CDM1A) without skeletal

muscle involvement (Fatkin et al., 1999). Finally, a form of

congenital muscular dystrophy has more recently been linked

to mutations in LMNA (Quijano-Roy et al., 2008), as well as

Heart-hand syndrome, which couples a range of cardiac defects

to brachydactyly (Renou et al., 2008).

Pathology associated with LMNAmutations is not restricted to

striated muscle tissue, as other diseases confer loss of adipose

tissue, including Dunnigan-type familial partial lipodystrophy

1366 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

(FPLD2) (Shackleton et al., 2000), Mandibuloacral dysplasia

(MAD) (Novelli et al., 2002), generalized lipoatrophy (Caux

et al., 2003), restrictive dermopathy (RD) (Navarro et al., 2004),

and other overlapping disorders. Highlighting the importance

of lamin A processing, mutations resulting in loss of ZMPSTE24

function, which result in partially processed lamin A, lead to both

MAD and RD (Agarwal et al., 2003; Navarro et al., 2005).

However, links between lamin A processing and pathology

extend beyond mutations in ZMPSTE24 and connect with

another set of disorders termed progeroid, which give rise to

the appearance of premature aging. The most noted of these

is Hutchinson-Gilford progeria syndrome (HGPS), a severe

disorder for which symptoms, including cachexia, alopecia,

and atherosclerosis, become apparent shortly after birth. Death

results from heart attack or stroke usually before the patient

reaches the age of 20. The most common LMNA mutation

leading to HGPS, G608G, does not affect coding sequence but

instead creates a cryptic splice site leading to removal of

50 amino acids in the C terminus of lamin A (De Sandre-

Giovannoli et al., 2003; Eriksson et al., 2003), resulting in a protein

named progerin. A similar splicing mutant has been identified

that leads to removal of an extra 40 amino acids (90 total) in

a patient diagnosed with RD (Navarro et al., 2004), leading to

speculation that RD is a more severe version of HGPS, though

the two diseases do not overlap entirely. This splicing event re-

moves the cleavage site for ZMPSTE24, creating a permanently

farnesylated protein that likely causes a dominant gain-of-func-

tion toxicity. Other mutations in LMNA that do not obviously

affect C-terminal splicing lead to HGPS aswell as other generally

less severe progeroid pathologies (Cao and Hegele, 2003; Chen

et al., 2003; Verstraeten et al., 2006). Finally, with regard to

LMNA mutations, homozygous loss of lamin A function leads

to Charcot-Marie-Tooth syndrome, characterized by loss of

peripheral nerve myelination (De Sandre-Giovannoli et al., 2002).

Before leaving A-type lamins, it is worth noting the interesting

connections that have arisen with cancer progression (Butin-

Israeli et al., 2012). Most laminopathies are not associated with

cancer, but an increasing range of tumors are characterized by

downregulation of A-type lamin expression (Broers et al., 1993;

Kaufmann, 1992), though results differ in tumor types. Recalling

that this family of lamins is expressed in differentiated cells, but

not stem cells, speculation has developed that A-type lamins

may act as tumor suppressors, perhaps by blocking dedifferen-

tiation into a more stem-cell-like state. A-type lamins have also

been ascribed roles in regulating cell proliferation and the DNA

damage response, either of which could be linked to cancer

progression (Redwood et al., 2011). Among these activities,

A-type lamins are required to stabilize the retinoblastoma tumor

suppressor protein (Johnson et al., 2004). This may be relevant

because the one tumor described in HGPS patients (the sample

size is quite small) is an early onset osteosarcoma (Shalev et al.,

2007), one of the most common tumors linked to homozygous

mutation of the Rb locus (Friend et al., 1986). Although progerin

can stabilize pRb levels (Nitta et al., 2006), the HGPS patient with

osteosarcoma had a rare T623S LMNA mutation that has not

been tested with regard to pRb stability (Shalev et al., 2007).

Mutations in genes encoding other nuclear envelope proteins

are also associated with disease (described in greater detail

in Mendez-Lopez and Worman, 2012). In addition to emerin

and LMNA, mutations in SYNE1 and SYNE2 (encoding

nesprin-1 and nesprin-2), TMEM43 (encoding LUMA), and

TMPO (encoding LAP2-a) are all associated with dilated cardio-

myopathy and muscular dystrophy (Liang et al., 2011; Taylor

et al., 2005; Zhang et al., 2007). These genes encode proteins

that all interact as part of the linker of nucleoskeleton and cyto-

skeleton (LINC) complex, suggesting that altered LINC function

may underlie striated muscle pathology (Puckelwartz and

McNally, 2011). Unrelated SYNE1 and SYNE2 mutations are

also linked to autosomal-recessive spinocerebellar ataxia type

8 and autosomal recessive arthrogryposis, respectively (Attali

et al., 2009; Gros-Louis et al., 2007). LEMD3, encoding MAN1,

an LEM-domain-containing protein, is also associated with

disease, withmutations linked to a series of disorders associated

with increased bone density (Hellemans et al., 2004). In addition,

mutations in BANF1, encoding the nuclear envelope protein BAF

that binds DNA and is involved in chromatin organization and

nuclear envelope assembly, are associated with Atypical proge-

ria (Puente et al., 2011).

Not to be left out, LMNB1 and LMNB2 mutations are both

linked to rare diseases. Autosomal-dominant mutations in

LMNB1 lead to adult-onset leukodystrophy, which is character-

ized by central nervous system demyelination (Padiath et al.,

2006). In the case of LMNB2, individuals with heterozygous

mutations are susceptible to acquired partial lipodystrophy,

likely triggered by one of several autoimmune diseases (Hegele

et al., 2006). Finally, the lamin B receptor (LBR), which interacts

with B-type lamins andmay serve to help link them to the nuclear

envelope and chromatin, is also a target for mutation in two

syndromes: homozygous mutations in LBR cause Greenberg

skeletal dysplasia (Waterham et al., 2003), whereas heterozy-

gous mutations are associated with Pelger-Huet anomaly,

a benign condition characterized by altered chromatin organiza-

tion in granulocytes (Best et al., 2003; Hoffmann et al., 2002).

Given the rate of new discoveries of disease association with

nuclear structural factors, it is fair to speculate that new associ-

ations between disease and nuclear proteins will continue to

emerge.

Disease Mechanisms: Mouse Models Lead the WayHow could altered function of nuclear structural components

lead to such awide range of diseases? A decade or so ago, there

were few connections between lamins and known disease

mechanisms; however, lamins were known to be important for

a wide range of nuclear functions, including replication and tran-

scription. Many of the initial ideas were based on changes

observed at the level of cell biology. For instance, the shape of

the nucleus was found to be disrupted in fibroblasts lacking

A-type lamins, with enhanced nuclear deformation and sensi-

tivity to mechanical stress (Lammerding et al., 2004). Emerin-

deficit cells have similar properties, and reduced mechanical

stress could explain part of the pathology associated with dis-

eases such as dilated cardiomyopathy and muscular dystrophy,

where affected tissues are under regular strain (Lammerding

et al., 2005). However, cells isolated from human and mouse

tissue from the various laminopathies, all display abnormal

nuclear structure. These phenotypes range from abnormal

nuclear shape to nuclear blebbing and even dispersal of DNA

into the cytoplasm. While these observations may relate to

disease, they do not clearly differentiate one laminopathy from

another, and researchers have turned to more detailed assess-

ments of cellular function to generate more recent hypotheses.

Theories to explain the pathology associated with nuclear

structure defects have emerged largely from two areas: an

extensive set of mouse models and, more recently, studies of

stem cells expressing a range of mutant forms of A-type lamins.

An informative starting point for the former was the generation of

mice lacking A-type lamins (Sullivan et al., 1999). In addition to

being cachexic, these mice present with a subset of the pathol-

ogies associated with LMNA mutation, including muscular

dystrophy, dilated cardiomyopathy, and Charcot-Marie-Tooth

syndrome and succumb to the cardiac phenotype at about

6 weeks of age. Lmna+/� heterozygous mice also develop the

cardiac pathology, although at a slower rate, and mice express-

ing two different LMNA alleles associated with striated muscle

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1367

disease recapitulate at least some of the human phenotypes

(Arimura et al., 2005; Mounkes et al., 2005). One assertion arising

from these findings is that the muscle and peripheral myelination

diseases result from reduced A-type lamin function. This is not

surprising for Charcot-Marie-Tooth syndrome, which is a reces-

sive disorder in humans (De Sandre-Giovannoli et al., 2002).

However, both dominant and recessive mutations have been

identified in the muscle pathologies, and one possibility is that

autosomal-dominant alleles have a dominant-negative effect,

interfering with intermediate filament assembly or some other

property of A-type lamins. Haploinsufficiency also likely explains

the onset of disease in many cases.

It should be noted that the Lmna�/� mouse described origi-

nally may in fact not be a null allele of the gene (Sullivan et al.,

1999). Recent evidence suggests that this mouse expresses

a still incompletely characterized, truncated 54 kDa protein

derived from a splicing event that bypasses the removed exons

(Jahn et al., 2012). While the dust has not settled from this

finding, most data suggest that the lamin A variant expressed

in this mouse is hypomorphic. Interestingly, another Lmna�/�

model has been derived through disruption with a reporter

gene, and this mouse presents with defective development of

heart liver and somites, leading to death beforeweaning (Kubben

et al., 2011). This latter mouse is more consistent with a homozy-

gous LMNA nonsense mutation that resulted in the complete

absence of A-type lamins and was associated with the death

of a newborn patient (van Engelen et al., 2005). Clearly, these

findings call for some re-evaluation of studies performed in the

Lmna�/�mouse despite its past and continued value to the field.

One recent highly informativemousemodel was engineered to

homozygously express a nonfarnesylated version of lamin A in

the absence of lamin C (Davies et al., 2010). Thesemice were ex-

pected to resemble the phenotype of mice lacking ZMPSTE24

(see below) but instead present with cardiomyopathy. The inves-

tigators sought to determine whether the cardiac pathology was

attributable to gain-of-function toxicity or a reduced hypomor-

phic function of the lamin A variant. To distinguish, they gener-

ated a mouse expressing a nonfarnesylated allele over a null,

finding that this mouse has a more severe phenotype, consistent

with further reduced lamin A function. If the pathology was

a result of toxicity, the heterozygous mouse would have had

a less severe cardiac phenotype. These findings are consistent

with the data that cardiomyopathy derives from reduced lamin

A function.

While striated muscle pathology represents one cluster of

mouse LMNA models, progeria characterizes the other. In this

case, the data are generally supportive of amodel whereby lamin

A variants with processing defects show dominant onset of

a subset of features associated with HGPS. These models are

covered in greater detail in a recent review (Zhang et al., 2013).

Recall that the primary human lesion associated with HGPS is

a heterozygous G608Gmutation, which creates a splicing defect

and leads to permanently farnesylated lamin A. Much debate

centers around which of the many different HGPS models are

the best to develop mechanistic explanations and therapies for

human patients. Most of the models, including Lmna mutants

and Zmpste24�/�, present with a subset of phenotypes that

are characteristic of progeroid mice, including cachexia,

1368 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

reduced bone density and rib fractures, loss of subcutaneous

fat, kyphosis, alopecia, and premature death. However, a model

generated to express human progerin from a BAC clone does

not exhibit these phenotypes, instead displaying arterial smooth

muscle defects (Varga et al., 2006). While the differences are

unknown, both types of models may have advantages. For

instance, the BAC progerin model mimics atherosclerosis, which

by leading to heart attacks and strokes results in mortality in

most patients. Therefore, studies in this mouse explore effects

on what may be the most important pathology in children with

disease. However, the rapid presentation and wider array of

phenotypes in the other mice offer clear advantages as well. Of

note, some of the progeria model mice display cardiac defects

that are more consistent with dilated cardiomyopathy (Davies

et al., 2010; Yang et al., 2011). One point worth considering is

that a LMNA mutation could lead to gain-of-function toxicity

for some phenotypes and loss-of-function for others.

In the next two sections, we focus on the two classes of

LMNA-associated disease about which we understand the

most: striated muscle disease and progeroid disorders. The

exciting progress in these two areas has led to possible thera-

peutic approaches.

Disease Mechanisms and Possible Therapies forLMNA-Associated Striated Muscle DiseasesInteresting findings have emerged on several fronts with respect

to LMNA-associated dilated cardiomyopathy with conduction

defects and muscular dystrophies. While these findings do not

yet come together in a neat package, continued studies may

begin to generate such a composite understanding. The fact

that LMNA mutants leading to EDMD2/3 so closely resemble

X-linked EDMD that is caused by Emerin mutations is a critical

consideration for any mechanistic disease model. Unlike

A-type lamins, emerins reside in the inner and outer nuclear

membranes, interacting with lamins in the former case and

with microtubules in the latter. Lamin A/C binding to emerin is

required for its localization to the nuclear envelope (Vaughan

et al., 2001). This raises the possibility that emerins might be

a conduit by which the nuclear lamina communicates with the

cytoskeleton. However, no clear understanding has emerged

as to how and why the lamin A/C-emerin interaction is important

in skeletal and cardiac muscle.

The linker of nucleoskeleton and cytoskeleton (LINC) protein

complex, consisting of SUN1 and -2 as well as Nesprin 1 and

-2, also connects A-type lamins to the cytoskeleton with Sun

proteins directly interacting with lamin A/C at the inner nuclear

membrane and Nesprins in the lumen (Mejat and Misteli, 2010).

Nesprins cross the outer nuclear membranes and connect to

the cytoskeleton in the cytoplasm. In addition to linking the

nucleo- and cytoskeleton, LINC complexes have a wide range

of cellular functions, including in cell division, in centrosome-

nucleus association, in nuclear migration, and in positioning.

Disruption of any of these activities could contribute to disease

progression. A recent study has implicated SUN1 in disease

progression, albeit through an unexpected mechanism. In

Lmna�/� mice, SUN1 is dramatically overexpressed and

directed to the Golgi, presumably after nuclear occupancy sites

are saturated (Chen et al., 2012a). RNAi-mediated knockdown of

Figure 1. Signaling Pathways Disrupted by LMNA MutationsRecent years have seen several discoveries of signal transduction pathwaysthat are altered in LMNA mutant backgrounds associated with gain-of-function toxicity (those involved in Progeria), loss-of-function toxicity (i.e.,hypomorphic), or both. A list of pathways is provided, as described in detail inthe text.

SUN1 rescued nuclear defects in cell culture, and knockout of

SUN1 significantly extended the survival of Lmna�/�mice. While

many questions remain unresolved, this report suggests that one

significant problem associated with reduced A-type lamin func-

tion is SUN1-mediated toxicity in the Golgi.

Another intermediate filament factor, desmin, serves as a link-

ing factor between lamins and many cytoplasmic structures in

striated muscle cells. Desmin mutations can result in desmin-

related myopathies (DRM), which are characterized by cardiac

and skeletal muscle weakness with a highly variability of presen-

tation. Inherent in DRMat the cellular level is disruption of desmin

filaments and accumulation of desmin-containing protein aggre-

gates. Interestingly, cardiomyoctes from Lmna�/� mice display

disrupted desmin networks and elevated protein levels (Nikolova

et al., 2004). This may not be the case for skeletal muscle, as

electron micrographs of muscle biopsies from human patients

failed to detect abnormal desmin localization (Frock et al.,

2012; Piercy et al., 2007). The authors of this study also looked

at murine embryonic stem cells transfected with a human

EDMD mutation and differentiated into cardiomyocytes, finding

no defects in desmin localization. These latter findings appear

to differ from the in vivo studies described above and may

suggest that knockout of A-type lamins, as opposed to expres-

sion of an EDMD missense mutation, is required to induce

abnormal desmin localization. Alternatively, the cell culture

model may not recapitulate events regarding desmin.

Myoblasts generated from Lmna�/� mice are reported to have

differentiation defects, suggesting that reduced regenerative

potential of adult stem cells may combine with increased

damage to myofibers from mechanical stress sensitivity to

explain the rapid onset of dystrophic pathology (Frock et al.,

2006). Interestingly, a small percentage of Lmna�/� myoblasts

responds normally to differentiation signals, whereas a majority

fails to induce the differentiation program. The majority of prolif-

erating Lmna�/� myoblasts also display reduced levels of both

MyoD and desmin. Stable transfection of desmin rescues the

differentiation defects of these cells, implying that reduced des-

min levels during the proliferation phase may, in part, be respon-

sible for the inability of cells to respond to differentiation cues.

Stable expression of MyoD also rescues differentiation defects.

With respect to EDMD mutations, MyoD-transformed human

patient fibroblasts were reported to differentiate normally (Piercy

et al., 2007). Again, the differences may be attributable to the

relative severity of the LMNA mutation, or they may have been

suppressed in the latter case due to artificially high MyoD levels

(Frock et al., 2006; Piercy et al., 2007).

In recent years, it has become apparent that Lmna mutations

can lead to altered activation of major signal transduction path-

ways in the cell (Figure 1). While the mechanisms connecting the

nuclear envelope to cell signaling have not been fully elucidated,

the findings are important because (1) altered signaling can be

linked to pathological progression, and (2) in some cases, small

molecules are available as therapeutic options to correct

signaling defects. In cardiac tissue, three different branches of

the MAP-kinase-signaling pathways have been found to be

aberrantly activated in a mouse model homozygously express-

ing the human LMNA H222P mutant associated with dilated

cardiomyopathy (Muchir et al., 2007b, 2012). One of these, the

extracellular signal-regulated kinase 1/2 (ERK1/2) pathway,

was also upregulated in Emerin-deficient mice, while the Jun

N-terminal kinase (JNK) pathway was not elevated and the

p38a pathway remains to be tested (Muchir et al., 2007a).

Elevated ERK1/2 phosphorylation has also been detected in

human cancer cells lines in which A-type lamin or Emerin expres-

sion was inhibited by an siRNA approach and in cardiac tissue

from Lmna�/� mice (Frock et al., 2012; Muchir et al., 2009b). In

Lmna�/� hearts, aberrant phosphorylation could be corrected

by restoration of lamin A expression specifically in cardiomyo-

cytes, indicating that the defects are cell autonomous (Frock

et al., 2012). Finally, elevated p38a phosphorylation has been

detected in heart tissue from human dilated cardiomyopathy

patients (Muchir et al., 2012).

A variety of MAP kinase inhibitors have been generated, and

many have been tested in the clinic for other disease indications.

Worman and colleagues have tested several of these in

LmnaH222p/H222p mice, finding that inhibition of each branch of

the Map kinase pathway delays either onset or progression of

cardiac symptoms (Muchir et al., 2009a, 2012; Wu et al.,

2010). Given that some of these inhibitors appear to be relatively

well tolerated in humans, these findings lead to a potential

therapeutic route for dilated cardiomyopathies associated with

LMNA mutation. Potential benefits for muscular dystrophy

have not been assessed.

Howdoes LMNAmutation lead to activation of theMAP kinase

pathways? While the answer to this question remains to be

determined, ideas have emerged. For instance, MAP kinases

are known to be activated by mechanical stress, and reduced

A-type lamin function is associated with impaired activation

of mechanosensitive genes in cardiomyocytes. A second more

direct model has potentially emerged that involves direct

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1369

Figure 2. Potential Therapeutic Approaches to LaminopathiesSeveral small molecules have been proposed as treatments for laminopathies.The major ones are listed with arrows indicating the diseases to which theymay have efficacy. Question marks indicate that animal data have yet to bepresented. Notably, FTIs have been tested in human children with HGPS, withpromising initial results (Gordon et al., 2012).

interaction between ERK1/2 and A-type lamins in the nucleus.

ERK1/2 is reported to interact with lamin A and the retinoblas-

toma protein (pRb) at the nuclear periphery. Stabilization of

pRb by A-type lamins is important to maintain normal cell-cycle

control (Nitta et al., 2006). Upon serum stimulation of quiescent

cells, ERK1/2 phosphorylates c-Fos, releasing it to stimulate

Ap-1 activation, and also dislodges pRb from A-type lamins,

leading to pRb phosphorylation and E2F activation (Gonzalez

et al., 2008; Ivorra et al., 2006; Rodrıguez et al., 2010). It is

unclear presently how disruption of the ERK1/2-A-type lamin

interaction by LMNA mutation affects ERK1/2 activation, but

this question needs to be investigated.

Equally unclear are the pathways downstream of MAP kinases

that mediate cardiac pathology. Two possibilities have emerged.

The first involves an observation that connexins are mislocalized

in mice expressing a different mutant associated with DCM

(N195K). Here, connexin 43 was found to be mislocalized and

not associated with gap junctions, a finding that could explain

conduction defects associated with altered A-type lamin func-

tion (Mounkes et al., 2005). Expression of another DCM mutant

(E82K) was found to lead to downregulation and mislocalization

of connexin 43 in neonatal myocytes (Sun et al., 2010). Finally,

a recent study has demonstrated mislocalization of connexin

43 in heart cardiomyocytes of Lmna�/� mice (Frock et al.,

2012). Re-expression of lamin A rescued aberrant ERK1/2 phos-

phorylation and restored connexin 43 localization. Given that

connexins are known substrates of ERK1/2, the possibility exists

that aberrant activity of this pathway disrupts normal connexin

43 localization and interferes with cardiac conduction (Chen

et al., 2012b).

Two recent studies point to the involvement of another major

signal transduction pathway in LMNA-related cardiac and

skeletal muscle disease. In Lmna�/� mice, the mTORC1

pathway was found to be upregulated in cardiac and skeletal

muscle, leading at least in the heart to impaired autophagy

(Ramos et al., 2012). ReducedmTORC1 signaling by the specific

kinase inhibitor rapamycin led to enhanced cardiac function and

survival, with indications of improved skeletal muscle function;

the latter possibility needs to be more fully explored. A similar

study conducted in the LmnaH222P/H222P mouse led to highly

overlapping findings, suggesting that aberrant mTORC1

signalingmay be a common feature of this class of laminopathies

(Choi et al., 2012). Among several upstream activators of

mTORC1 are ERK1/2 MAP kinases, and one possibility is that

increased mTORC1 signaling occurs by this mechanism. How-

ever, there are numerous upstream activators of mTORC1 that

need to bemore fully explored. The possibility of testing rapamy-

cin as a treatment for LMNA-associated DCM is intriguing

because the drug has been tested in a wide range of clinical trials

and is approved for multiple disease indications. However, there

are side effects such as dyslipidemia and impaired insulin

signaling that, while generally manageable, must be considered

for treatment of cardiac disease. Elevated mTORC1 signaling,

which is classically associated with increased protein translation

and cell growth, is already linked to forms of cardiac hyper-

trophy. However, general levels of translation do not appear to

be elevated in the Lmna�/� heart (Ramos et al., 2012), suggest-

ing that other pathways are offsetting the translational effects of

1370 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

mTORC1 in this scenario. Interestingly, rapamycin has been

reported to improve autophagic flux and suppress nuclear

blebbing in fibroblasts expressing progerin, indicating that

suppression of the mTOR pathway may be efficacious in

LMNA-associated progeria models as well (Cao et al., 2011).

Given the remarkable progress in this cluster of LMNA-associ-

ated diseases, it has been possible to move from identification of

LMNA mutations in EDMD and DCM to possible therapeutic

approaches in less than two decades (Figure 2). Whether the

current drugs will prove efficacious in humans remains to be

seen. Even if this is not the case, new candidate therapeutic

approaches will surely continue to emerge.

Disease Mechanisms and Possible Therapies forLMNA-Associated ProgeriasAlthough very rare, progeria syndromes have long been of great

interest, based in part on the hypothesis that, by learning the

mechanisms underlying their pathology, insights will be made

into the normal aging process. This assumption is yet to be vali-

dated, and researchers in the aging field have a wide range of

viewpoints. One thing is clear. The studies into LMNA-associ-

ated progerias have yielded major biological insights and have

provided hope that therapeutic approaches can be developed

to slow the impact of these very severe syndromes. In this

section, the latest findings in progeria and lamin A processing

will be discussed.

A large body of work suggests that HGPS mutants in LMNA

at least in part confer toxicity by virtue of being permanently

farnesylated. Several deformations of the nucleus were found

in cells expressing progerin or other nonfarnesylated versions

of lamin A, and several studies indicated that these phenotypes

could be rescued by a class of drugs that inhibit farnesyltrans-

ferases (Young et al., 2006). These drugs were initially gener-

ated based on their ability to block Ras farnesylation and the

promise that that would inhibit tumor progression. Though

cancer studies continue, their development has been fortuitous

to the study of HGPS. Not only do they rescue cellular defects,

but they have beneficial properties when delivered to HGPS

mouse models, extending survival and improving other physio-

logical readouts, including bone and cardiovascular defects

(Capell et al., 2008; Yang et al., 2008b). These findings,

together with the fact that FTIs have good safety profiles in

the clinic, were cause for great optimism, leading to the first

clinical trial in human patients with HGPS. Initial findings were

recently reported showing variable rates of improvement in

vascular function, enhanced bone rigidity, and improved senso-

rineural hearing in 25 patients treated with Ionafarnib for at least

2 years (Gordon et al., 2012).

One reason FTIs may have limited potency is that lamin

A variants can become geranylgeranylated, especially when

farnesyltransferase activity is blocked (Varela et al., 2008). This

has led to the assumption that blocking the HMG-CoA reductase

pathway upstream in a manner that inhibits both lamin A modifi-

cations might have enhanced efficacy. Consistently, combined

treatment of Zmpste24�/� mice with two such agents, statins

and aminobisphosphonates, enhances survival and improves

several pathologies. Another potential approach has emerged

in a mouse that is genetically engineered to have the exact

G608G mutation (G609G in mice) (Osorio et al., 2011). As in

the human case, alternative splicing leads to progerin production

and progeroid phenotypes. Interestingly, treatment of the mice

with a morpholino-based therapy that prevents pathogenic

splicing delays pathology and extends survival, suggesting an

alternative therapeutic approach.

Genetic studies support the toxicity of farnesylated lamin A in

progerias. For instance, mice lacking Zmpste24 develop pro-

geroid features linked to the toxicity of an unprocessed lamin

A, as deletion of one copy of LMNA in this background improves

the range of phenotypes (Fong et al., 2004). Extensive studies by

Young and colleagues have further elucidated the role of

farnesylation in vivo. Mice engineered to express a nonfarnesy-

lated version of progerin still develop progeroid features, albeit

at a slower rate (Yang et al., 2008a). However, mice expressing

a nonfarnesylated version of prelamin A do not develop proge-

roid features, as described earlier (Davies et al., 2010). One

possible interpretation of these studies in that farnesylation

may be required for toxicity in the case of prelamin A but that

the 50 amino acid deletion in progerin also contributes to disease

progression.

Several lines of evidence implicate enhanced DNA damage

and/or an impaired DNA damage response pathway in the

etiology of HGPS. HGPS cells have higher levels of reactive

oxygen species and greater rates of basal DNA damage (Viteri

et al., 2010). These findings are likely connected, as a reduction

in ROS by exposure to n-acetylcysteine reduces double-strand

break formation. These alterations lead, in part, to enhanced

activation of DNA response pathways, including enhanced

ATM and RAD3-associated foci, which may adversely affect

cell-cycle proliferation. An interesting and unusual feature of

HGPS cells is persistent basal levels of phosphorylated gH2AX

foci marking double-strand breaks that also stain positive

for Xeroderma pigmentosum group A protein (XPA) (Liu et al.,

2008), a component of nucleotide excision repair. No other

related factors are upregulated, suggesting that the foci have

an abnormal set of repair proteins and the type of DNA damage

in HGPS cells may have unique features.

Cells from mice lacking Zmpste24 also exhibit a significant

delay in recruitment of 53BP1 to sites of DNA repair after induc-

tion of double-strand breaks (Liu et al., 2005). p53 targets such

as GADD45, p21, and ATF3 were also elevated, and deletion of

p53 was sufficient to rescue some of the progeroid phenotypes

of the Zmpste24�/� mouse (Varela et al., 2005). Though p53

targets were not elevated in HGPS fibroblasts, inactivation of

the transcription factor was sufficient to suppress premature

senescence (Kudlow et al., 2008). More recent data indicate

that ATM and NEMO pathways become activated and promote

NF-kB-dependent inflammation in both Zmpste24�/� and

LmnaG609G/G609Gmice (Osorio et al., 2012). Genetic and pharma-

cological interventions of these pathways slow progeroid

pathology and enhance survival. These findings are particularly

interesting because (1) they suggest that NF-kB inhibitors may

be effective therapeutic agents and (2) enhanced inflammation

may be a major driver of normal aging processes. Furthermore,

the tissues affected by altered lamin A processing have remained

unresolved. Progeria involves systemic pathology, and one

possibility is that defects in every tissue cause cell autonomous

phenotypes.More likely, defects in a smaller set of tissues lead to

systemic responses that impact the whole organism. Enhanced

NF-kB signaling could mediate such a systemic effect.

A more straightforward approach to understanding the role of

A-type lamins in DNA damage responses may involve loss-of-

function studies. In contrast to progeroid models, loss of

A-type lamins leads to 53BP1 degradation by the proteasome

(Gonzalez-Suarez et al., 2009). In its absence, repair of double-

strand breaks proceeds more slowly, hindering effective nonho-

mologous end joining (Redwood et al., 2011). Homologous

recombination is also compromised through a transcriptional

mechanism by which enhanced proteasome-dependent degra-

dation of pRb and p107 leads to repression of RAD51 and

BRCA1 (Redwood et al., 2011). It remains unclear why enhanced

protein turnover of pRb and 53BP1 occur in the absence of

A-type lamins, but the hypothesis has been put forward that

A-type lamins may have a general role in promoting the

stability of several nuclear regulatory factors through keeping

proteasome-dependent degradation in check (Parnaik et al.,

2011). It should also be noted that many of these properties

may explain why loss of A-type lamin expression could have

tumor-promoting properties.

In addition to impaired DNA damage response pathways,

telomere dysregulation may also contribute to progeroid

pathology. In culture, HGPS fibroblasts experience faster telo-

mere shortening, and normal fibroblasts expressing progerin

recapitulate this phenotype and also enhance formation of

signal-free ends (Decker et al., 2009). Enhanced telomere attri-

tion may contribute to proliferation defects and early senes-

cence, as telomerase expression restores both properties in

fibroblasts (Benson et al., 2010; Kudlow et al., 2008). One role

of telomerase may be to enhance resolution of DNA damage

foci, which were found to localize in regions near telomeres

(Benson et al., 2010). Themechanisms bywhich this might occur

and the extent to which altered telomere dynamics promotes

progeroid pathology remains to be determined.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1371

Given that HGPS (and other laminopathies) primarily affect

tissues of mesenchymal origin, altered mesenchymal stem cell

function may be a major site of progerin-induced dysfunction.

Gene expression profiling in fibroblasts expressing progerin

provides support for this assertion, as Notch signaling was found

to be highly enhanced (Scaffidi and Misteli, 2008). Elevated

Notch activity associated with progerin expression was found

to promote expression of a range of differentiation markers in

human mesenchymal stem cells. As a possible mechanism,

progerin was found to disrupt nuclear matrix association of

SKIP, a coactivator of Notch genes, leading to its release into

the nucleoplasm and activation of targets. Reduced mesen-

chymal stem cell function could promote a subset of progeroid

phenotypes in vivo, but this remains to be tested.

The Wnt/b-catenin pathway is altered in a variety of laminopa-

thies as well. In both Zmpste24�/� and HGPS mice, reduced

b-catenin levels were detected, and cell proliferation defects

could be rescued by inhibition of Gsk-3b, leading to b-catenin

stabilization (Espada et al., 2008; Hernandez et al., 2010). Of

note, the Wnt pathway may be disrupted in mice lacking emerin

(Markiewicz et al., 2006; Tilgner et al., 2009). Given that the Wnt

pathway may have critical roles in maintaining adult stem cell

function with age, the role of this pathway in laminopathies

requires further interrogation.

Adult stem cells may also be impaired in progeroid laminopa-

thies due to impaired SIRT1 function. A recent study has demon-

strated that A-type lamins interact with the protein deacetylase

and that preprocessed lamin A disrupts this association in

Zmpste24�/� cells, leading to reduced deacetylase activity and

rapid in vivo stem cell depletion (Liu et al., 2012). Treatment of

mice with resveratrol restores SIRT1 activity, reduces the

pathology, and extends survival, indicating that enhancing

SIRT1 activity may be another therapeutic approach in progeroid

disorders associated with LMNA mutation.

ConclusionsSince the identification, in 1999, of diseases caused by muta-

tions in genes encoding for nuclear lamina proteins, research

has been dedicated toward understanding the molecular mech-

anisms leading to these specific phenotypes. Understanding

how the nuclear lamina interacts with structural proteins, chro-

matin, transcription factors, and other signaling partners will

likely give us an understanding of mechanistic links to disease.

At this moment, the puzzle is starting to come together, but the

overall picture of how lamins regulate all of these pathways

and how this regulation leads to disease is still developing.

Understanding the mechanisms by which mutations in lamins

cause these rare diseases will provide molecular insight into

other common conditions that laminopathies model, such as

muscle diseases and cardiomyopathy. Additionally, because

mutations in the nuclear lamina result in rapid aging-like disease,

defining the role of the nuclear lamina in regulating normal human

longevity will be of great importance.

ACKNOWLEDGMENTS

The authors would like to apologize to those scientists whose studies were

not referenced due to space limitations and also acknowledge the editorial

1372 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

contributions of Juniper Pennypacker. Lamin-related research in the lab of

B.K.K. is supported by a grant from the National Institute of Aging (R01

AG024287). K.H.S. is supported by a Ruth L. Kirschstein NRSA Postdoctoral

Fellowship. B.K.K. is an Ellison Medical Foundation Senior Scholar in Aging.

REFERENCES

Agarwal, A.K., Fryns, J.P., Auchus, R.J., and Garg, A. (2003). Zinc metallo-

proteinase, ZMPSTE24, is mutated in mandibuloacral dysplasia. Hum. Mol.

Genet. 12, 1995–2001.

Arimura, T., Helbling-Leclerc, A., Massart, C., Varnous, S., Niel, F., Lacene, E.,

Fromes, Y., Toussaint, M., Mura, A.M., Keller, D.I., et al. (2005). Mouse model

carrying H222P-Lmna mutation develops muscular dystrophy and dilated

cardiomyopathy similar to human striated muscle laminopathies. Hum. Mol.

Genet. 14, 155–169.

Attali, R., Warwar, N., Israel, A., Gurt, I., McNally, E., Puckelwartz, M., Glick, B.,

Nevo, Y., Ben-Neriah, Z., and Melki, J. (2009). Mutation of SYNE-1, encoding

an essential component of the nuclear lamina, is responsible for autosomal

recessive arthrogryposis. Hum. Mol. Genet. 18, 3462–3469.

Benson, E.K., Lee, S.W., and Aaronson, S.A. (2010). Role of progerin-induced

telomere dysfunction in HGPS premature cellular senescence. J. Cell Sci. 123,

2605–2612.

Best, S., Salvati, F., Kallo, J., Garner, C., Height, S., Thein, S.L., and Rees, D.C.

(2003). Lamin B-receptor mutations in Pelger-Huet anomaly. Br. J. Haematol.

123, 542–544.

Bione, S., Maestrini, E., Rivella, S., Mancini, M., Regis, S., Romeo, G., and

Toniolo, D. (1994). Identification of a novel X-linked gene responsible for

Emery-Dreifuss muscular dystrophy. Nat. Genet. 8, 323–327.

Bonne, G., Di Barletta, M.R., Varnous, S., Becane, H.M., Hammouda, E.H.,

Merlini, L., Muntoni, F., Greenberg, C.R., Gary, F., Urtizberea, J.A., et al.

(1999). Mutations in the gene encoding lamin A/C cause autosomal dominant

Emery-Dreifuss muscular dystrophy. Nat. Genet. 21, 285–288.

Broers, J.L., Raymond, Y., Rot, M.K., Kuijpers, H., Wagenaar, S.S., and

Ramaekers, F.C. (1993). Nuclear A-type lamins are differentially expressed

in human lung cancer subtypes. Am. J. Pathol. 143, 211–220.

Butin-Israeli, V., Adam, S.A., Goldman, A.E., and Goldman, R.D. (2012).

Nuclear lamin functions and disease. Trends Genet. 28, 464–471.

Cao, H., and Hegele, R.A. (2003). LMNA is mutated in Hutchinson-Gilford pro-

geria (MIM 176670) but not inWiedemann-Rautenstrauch progeroid syndrome

(MIM 264090). J. Hum. Genet. 48, 271–274.

Cao, K., Graziotto, J.J., Blair, C.D., Mazzulli, J.R., Erdos, M.R., Krainc, D., and

Collins, F.S. (2011). Rapamycin reverses cellular phenotypes and enhances

mutant protein clearance in Hutchinson-Gilford progeria syndrome cells. Sci.

Transl. Med. 3, 89ra58.

Capell, B.C., Olive, M., Erdos, M.R., Cao, K., Faddah, D.A., Tavarez, U.L.,

Conneely, K.N., Qu, X., San, H., Ganesh, S.K., et al. (2008). A farnesyltransfer-

ase inhibitor prevents both the onset and late progression of cardiovascular

disease in a progeria mouse model. Proc. Natl. Acad. Sci. USA 105, 15902–

15907.

Caux, F., Dubosclard, E., Lascols, O., Buendia, B., Chazouilleres, O., Cohen,

A., Courvalin, J.-C., Laroche, L., Capeau, J., Vigouroux, C., and Christin-

Maitre, S. (2003). A new clinical condition linked to a novel mutation in lamins

A and C with generalized lipoatrophy, insulin-resistant diabetes, disseminated

leukomelanodermic papules, liver steatosis, and cardiomyopathy. J. Clin.

Endocrinol. Metab. 88, 1006–1013.

Chen, L., Lee, L., Kudlow, B.A., Dos Santos, H.G., Sletvold, O., Shafeghati, Y.,

Botha, E.G., Garg, A., Hanson, N.B., Martin, G.M., et al. (2003). LMNA muta-

tions in atypical Werner’s syndrome. Lancet 362, 440–445.

Chen, C.Y., Chi, Y.H., Mutalif, R.A., Starost, M.F., Myers, T.G., Anderson, S.A.,

Stewart, C.L., and Jeang, K.T. (2012a). Accumulation of the inner nuclear

envelope protein Sun1 is pathogenic in progeric and dystrophic laminopathies.

Cell 149, 565–577.

Chen, S.C., Kennedy, B.K., and Lampe, P.D. (2012b). Phosphorylation of con-

nexin43 on S279/282 may contribute to laminopathy-associated conduction

defects. Exp. Cell Res. Published online December 21, 2012. http://dx.doi.

org/10.1016/j.yexcr.2012.12.014.

Choi, J.C., Muchir, A., Wu, W., Iwata, S., Homma, S., Morrow, J.P., and

Worman, H.J. (2012). Temsirolimus activates autophagy and ameliorates

cardiomyopathy caused by lamin A/C gene mutation. Sci. Transl. Med. 4,

144ra102.

Davies, B.S., Barnes, R.H., 2nd, Tu, Y., Ren, S., Andres, D.A., Spielmann, H.P.,

Lammerding, J., Wang, Y., Young, S.G., and Fong, L.G. (2010). An accumula-

tion of non-farnesylated prelamin A causes cardiomyopathy but not progeria.

Hum. Mol. Genet. 19, 2682–2694.

Davies, B.S., Coffinier, C., Yang, S.H., Barnes, R.H., 2nd, Jung, H.J., Young,

S.G., and Fong, L.G. (2011). Investigating the purpose of prelamin A process-

ing. Nucleus 2, 4–9.

De Sandre-Giovannoli, A., Chaouch, M., Kozlov, S., Vallat, J.M., Tazir, M.,

Kassouri, N., Szepetowski, P., Hammadouche, T., Vandenberghe, A., Stewart,

C.L., et al. (2002). Homozygous defects in LMNA, encoding lamin A/C nuclear-

envelope proteins, cause autosomal recessive axonal neuropathy in human

(Charcot-Marie-Tooth disorder type 2) and mouse. Am. J. Hum. Genet. 70,

726–736.

De Sandre-Giovannoli, A., Bernard, R., Cau, P., Navarro, C., Amiel, J.,

Boccaccio, I., Lyonnet, S., Stewart, C.L., Munnich, A., Le Merrer, M., and

Levy, N. (2003). Lamin a truncation in Hutchinson-Gilford progeria. Science

300, 2055.

Decker, M.L., Chavez, E., Vulto, I., and Lansdorp, P.M. (2009). Telomere length

in Hutchinson-Gilford progeria syndrome. Mech. Ageing Dev. 130, 377–383.

Eriksson, M., Brown, W.T., Gordon, L.B., Glynn, M.W., Singer, J., Scott, L.,

Erdos, M.R., Robbins, C.M., Moses, T.Y., Berglund, P., et al. (2003). Recurrent

de novo point mutations in lamin A cause Hutchinson-Gilford progeria

syndrome. Nature 423, 293–298.

Espada, J., Varela, I., Flores, I., Ugalde, A.P., Cadinanos, J., Pendas, A.M.,

Stewart, C.L., Tryggvason, K., Blasco, M.A., Freije, J.M., and Lopez-Otın, C.

(2008). Nuclear envelope defects cause stem cell dysfunction in premature-

aging mice. J. Cell Biol. 181, 27–35.

Fatkin, D., MacRae, C., Sasaki, T., Wolff, M.R., Porcu, M., Frenneaux, M.,

Atherton, J., Vidaillet, H.J.J., Jr., Spudich, S., De Girolami, U., et al. (1999).

Missense mutations in the rod domain of the lamin A/C gene as causes of

dilated cardiomyopathy and conduction-system disease. N. Engl. J. Med.

341, 1715–1724.

Fong, L.G., Ng, J.K., Meta, M., Cote, N., Yang, S.H., Stewart, C.L., Sullivan, T.,

Burghardt, A., Majumdar, S., Reue, K., et al. (2004). Heterozygosity for Lmna

deficiency eliminates the progeria-like phenotypes in Zmpste24-deficient

mice. Proc. Natl. Acad. Sci. USA 101, 18111–18116.

Friend, S.H., Bernards, R., Rogelj, S., Weinberg, R.A., Rapaport, J.M., Albert,

D.M., and Dryja, T.P. (1986). A human DNA segment with properties of the

gene that predisposes to retinoblastoma and osteosarcoma. Nature 323,

643–646.

Frock, R.L., Kudlow, B.A., Evans, A.M., Jameson, S.A., Hauschka, S.D., and

Kennedy, B.K. (2006). Lamin A/C and emerin are critical for skeletal muscle

satellite cell differentiation. Genes Dev. 20, 486–500.

Frock, R.L., Chen, S.C., Da, D.F., Frett, E., Lau, C., Brown, C., Pak, D.N.,

Wang, Y., Muchir, A., Worman, H.J., et al. (2012). Cardiomyocyte-specific

expression of lamin a improves cardiac function in Lmna-/- mice. PLoS ONE

7, e42918.

Gerace, L., and Huber, M.D. (2012). Nuclear lamina at the crossroads of the

cytoplasm and nucleus. J. Struct. Biol. 177, 24–31.

Gerace, L., Blum, A., and Blobel, G. (1978). Immunocytochemical localization

of the major polypeptides of the nuclear pore complex-lamina fraction. Inter-

phase and mitotic distribution. J. Cell Biol. 79, 546–566.

Gonzalez, J.M., Navarro-Puche, A., Casar, B., Crespo, P., and Andres, V.

(2008). Fast regulation of AP-1 activity through interaction of lamin A/C,

ERK1/2, and c-Fos at the nuclear envelope. J. Cell Biol. 183, 653–666.

Gonzalez-Suarez, I., Redwood, A.B., Perkins, S.M., Vermolen, B., Lichtenszte-

jin, D., Grotsky, D.A., Morgado-Palacin, L., Gapud, E.J., Sleckman, B.P.,

Sullivan, T., et al. (2009). Novel roles for A-type lamins in telomere biology

and the DNA damage response pathway. EMBO J. 28, 2414–2427.

Gordon, L.B., Kleinman, M.E., Miller, D.T., Neuberg, D.S., Giobbie-Hurder, A.,

Gerhard-Herman, M., Smoot, L.B., Gordon, C.M., Cleveland, R., Snyder, B.D.,

et al. (2012). Clinical trial of a farnesyltransferase inhibitor in children with

Hutchinson-Gilford progeria syndrome. Proc. Natl. Acad. Sci. USA 109,

16666–16671.

Gros-Louis, F., Dupre, N., Dion, P., Fox, M.A., Laurent, S., Verreault, S., Sanes,

J.R., Bouchard, J.P., and Rouleau, G.A. (2007). Mutations in SYNE1 lead to

a newly discovered form of autosomal recessive cerebellar ataxia. Nat. Genet.

39, 80–85.

Hegele, R.A., Cao, H., Liu, D.M., Costain, G.A., Charlton-Menys, V., Rodger,

N.W., and Durrington, P.N. (2006). Sequencing of the reannotated LMNB2

gene reveals novel mutations in patients with acquired partial lipodystrophy.

Am. J. Hum. Genet. 79, 383–389.

Hellemans, J., Preobrazhenska, O., Willaert, A., Debeer, P., Verdonk, P.C.,

Costa, T., Janssens, K., Menten, B., Van Roy, N., Vermeulen, S.J., et al.

(2004). Loss-of-function mutations in LEMD3 result in osteopoikilosis,

Buschke-Ollendorff syndrome and melorheostosis. Nat. Genet. 36, 1213–

1218.

Hennekes, H., and Nigg, E.A. (1994). The role of isoprenylation in membrane

attachment of nuclear lamins. A single point mutation prevents proteolytic

cleavage of the lamin A precursor and confers membrane binding properties.

J. Cell Sci. 107, 1019–1029.

Hernandez, L., Roux, K.J., Wong, E.S., Mounkes, L.C., Mutalif, R.,

Navasankari, R., Rai, B., Cool, S., Jeong, J.W., Wang, H., et al. (2010).

Functional coupling between the extracellular matrix and nuclear lamina by

Wnt signaling in progeria. Dev. Cell 19, 413–425.

Hoffmann, K., Dreger, C.K., Olins, A.L., Olins, D.E., Shultz, L.D., Lucke, B.,

Karl, H., Kaps, R., Muller, D., Vaya, A., et al. (2002). Mutations in the gene

encoding the lamin B receptor produce an altered nuclear morphology in

granulocytes (Pelger-Huet anomaly). Nat. Genet. 31, 410–414.

Ivorra, C., Kubicek, M., Gonzalez, J.M., Sanz-Gonzalez, S.M., Alvarez-

Barrientos, A., O’Connor, J.E., Burke, B., and Andres, V. (2006). A mecha-

nism of AP-1 suppression through interaction of c-Fos with lamin A/C. Genes

Dev. 20, 307–320.

Jahn, D., Schramm, S., Schnolzer, M., Heilmann, C.J., de Koster, C.G.,

Schutz, W., Benavente, R., and Alsheimer, M. (2012). A truncated lamin A in

the Lmna -/- mouse line: implications for the understanding of laminopathies.

Nucleus 3, 463–474.

Johnson, B.R., Nitta, R.T., Frock, R.L., Mounkes, L., Barbie, D.A., Stewart,

C.L., Harlow, E., and Kennedy, B.K. (2004). A-type lamins regulate retinoblas-

toma protein function by promoting subnuclear localization and preventing

proteasomal degradation. Proc. Natl. Acad. Sci. USA 101, 9677–9682.

Kaufmann, S.H. (1992). Expression of nuclear envelope lamins A and C in

human myeloid leukemias. Cancer Res. 52, 2847–2853.

Kubben, N., Voncken, J.W., Konings, G., van Weeghel, M., van den

Hoogenhof, M.M., Gijbels, M., van Erk, A., Schoonderwoerd, K., van den

Bosch, B., Dahlmans, V., et al. (2011). Post-natal myogenic and adipogenic

developmental: defects and metabolic impairment upon loss of A-type lamins.

Nucleus 2, 195–207.

Kudlow, B.A., Stanfel, M.N., Burtner, C.R., Johnston, E.D., and Kennedy, B.K.

(2008). Suppression of proliferative defects associated with processing-

defective lamin A mutants by hTERT or inactivation of p53. Mol. Biol. Cell

19, 5238–5248.

Lammerding, J., Schulze, P.C., Takahashi, T., Kozlov, S., Sullivan, T., Kamm,

R.D., Stewart, C.L., and Lee, R.T. (2004). Lamin A/C deficiency causes

defective nuclear mechanics and mechanotransduction. J. Clin. Invest.

113, 370–378.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1373

Lammerding, J., Hsiao, J., Schulze, P.C., Kozlov, S., Stewart, C.L., and Lee,

R.T. (2005). Abnormal nuclear shape and impaired mechanotransduction in

emerin-deficient cells. J. Cell Biol. 170, 781–791.

Lee, R., Chang, S.Y., Trinh, H., Tu, Y., White, A.C., Davies, B.S., Bergo, M.O.,

Fong, L.G., Lowry, W.E., and Young, S.G. (2010). Genetic studies on the func-

tional relevance of the protein prenyltransferases in skin keratinocytes. Hum.

Mol. Genet. 19, 1603–1617.

Liang, W.C., Mitsuhashi, H., Keduka, E., Nonaka, I., Noguchi, S., Nishino, I.,

and Hayashi, Y.K. (2011). TMEM43 mutations in Emery-Dreifuss muscular

dystrophy-related myopathy. Ann. Neurol. 69, 1005–1013.

Liu, B., Wang, J., Chan, K.M., Tjia, W.M., Deng, W., Guan, X., Huang, J.D., Li,

K.M., Chau, P.Y., Chen, D.J., et al. (2005). Genomic instability in laminopathy-

based premature aging. Nat. Med. 11, 780–785.

Liu, Y., Wang, Y., Rusinol, A.E., Sinensky, M.S., Liu, J., Shell, S.M., and Zou, Y.

(2008). Involvement of xeroderma pigmentosum group A (XPA) in progeria

arising from defective maturation of prelamin A. FASEB J. 22, 603–611.

Liu, B., Ghosh, S., Yang, X., Zheng, H., Liu, X., Wang, Z., Jin, G., Zheng, B.,

Kennedy, B.K., Suh, Y., et al. (2012). Resveratrol rescues SIRT1-dependent

adult stem cell decline and alleviates progeroid features in laminopathy-based

progeria. Cell Metab. 16, 738–750.

Markiewicz, E., Tilgner, K., Barker, N., van de Wetering, M., Clevers, H.,

Dorobek, M., Hausmanowa-Petrusewicz, I., Ramaekers, F.C., Broers, J.L.,

Blankesteijn, W.M., et al. (2006). The inner nuclear membrane protein emerin

regulates beta-catenin activity by restricting its accumulation in the nucleus.

EMBO J. 25, 3275–3285.

Mejat, A., and Misteli, T. (2010). LINC complexes in health and disease.

Nucleus 1, 40–52.

Mendez-Lopez, I., andWorman, H.J. (2012). Inner nuclearmembrane proteins:

impact on human disease. Chromosoma 121, 153–167.

Mounkes, L.C., Kozlov, S.V., Rottman, J.N., and Stewart, C.L. (2005). Expres-

sion of an LMNA-N195K variant of A-type lamins results in cardiac conduction

defects and death in mice. Hum. Mol. Genet. 14, 2167–2180.

Muchir, A., Bonne, G., van der Kooi, A.J., van Meegen, M., Baas, F., Bolhuis,

P.A., de Visser, M., and Schwartz, K. (2000). Identification of mutations in the

gene encoding lamins A/C in autosomal dominant limb girdle muscular

dystrophy with atrioventricular conduction disturbances (LGMD1B). Hum.

Mol. Genet. 9, 1453–1459.

Muchir, A., Pavlidis, P., Bonne, G., Hayashi, Y.K., and Worman, H.J. (2007a).

Activation of MAPK in hearts of EMD null mice: similarities between mouse

models of X-linked and autosomal dominant Emery Dreifuss muscular

dystrophy. Hum. Mol. Genet. 16, 1884–1895.

Muchir, A., Pavlidis, P., Decostre, V., Herron, A.J., Arimura, T., Bonne, G., and

Worman, H.J. (2007b). Activation of MAPK pathways links LMNAmutations to

cardiomyopathy in Emery-Dreifuss muscular dystrophy. J. Clin. Invest. 117,

1282–1293.

Muchir, A., Shan, J., Bonne, G., Lehnart, S.E., and Worman, H.J. (2009a). Inhi-

bition of extracellular signal-regulated kinase signaling to prevent cardiomyop-

athy caused by mutation in the gene encoding A-type lamins. Hum. Mol.

Genet. 18, 241–247.

Muchir, A., Wu, W., and Worman, H.J. (2009b). Reduced expression of A-type

lamins and emerin activates extracellular signal-regulated kinase in cultured

cells. Biochim. Biophys. Acta 1792, 75–81.

Muchir, A., Wu,W., Choi, J.C., Iwata, S., Morrow, J., Homma, S., andWorman,

H.J. (2012). Abnormal p38a mitogen-activated protein kinase signaling in

dilated cardiomyopathy caused by lamin A/C gene mutation. Hum. Mol.

Genet. 21, 4325–4333.

Navarro, C.L., De Sandre-Giovannoli, A., Bernard, R., Boccaccio, I., Boyer, A.,

Genevieve, D., Hadj-Rabia, S., Gaudy-Marqueste, C., Smitt, H.S., Vabres, P.,

et al. (2004). Lamin A and ZMPSTE24 (FACE-1) defects cause nuclear disorga-

nization and identify restrictive dermopathy as a lethal neonatal laminopathy.

Hum. Mol. Genet. 13, 2493–2503.

Navarro, C.L., Cadinanos, J., De Sandre-Giovannoli, A., Bernard, R., Courrier,

S., Boccaccio, I., Boyer, A., Kleijer, W.J., Wagner, A., Giuliano, F., et al. (2005).

1374 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Loss of ZMPSTE24 (FACE-1) causes autosomal recessive restrictive dermop-

athy and accumulation of Lamin A precursors. Hum. Mol. Genet. 14, 1503–

1513.

Nikolova, V., Leimena, C., McMahon, A.C., Tan, J.C., Chandar, S., Jogia, D.,

Kesteven, S.H., Michalicek, J., Otway, R., Verheyen, F., et al. (2004). Defects

in nuclear structure and function promote dilated cardiomyopathy in lamin

A/C-deficient mice. J. Clin. Invest. 113, 357–369.

Nitta, R.T., Jameson, S.A., Kudlow, B.A., Conlan, L.A., and Kennedy, B.K.

(2006). Stabilization of the retinoblastoma protein by A-type nuclear lamins

is required for INK4A-mediated cell cycle arrest. Mol. Cell. Biol. 26, 5360–5372.

Novelli, G., Muchir, A., Sangiuolo, F., Helbling-Leclerc, A., D’Apice, M.R.,

Massart, C., Capon, F., Sbraccia, P., Federici, M., Lauro, R., et al. (2002). Man-

dibuloacral dysplasia is caused by a mutation in LMNA-encoding lamin A/C.

Am. J. Hum. Genet. 71, 426–431.

Osorio, F.G., Navarro, C.L., Cadinanos, J., Lopez-Mejia, I.C., Quiros, P.M.,

Bartoli, C., Rivera, J., Tazi, J., Guzman, G., Varela, I., et al. (2011). Splicing-

directed therapy in a new mouse model of human accelerated aging. Sci.

Transl. Med. 3, 106ra107.

Osorio, F.G., Barcena, C., Soria-Valles, C., Ramsay, A.J., de Carlos, F., Cobo,

J., Fueyo, A., Freije, J.M., and Lopez-Otın, C. (2012). Nuclear lamina defects

cause ATM-dependent NF-kB activation and link accelerated aging to

a systemic inflammatory response. Genes Dev. 26, 2311–2324.

Padiath, Q.S., Saigoh, K., Schiffmann, R., Asahara, H., Yamada, T., Koeppen,

A., Hogan, K., Ptacek, L.J., and Fu, Y.H. (2006). Lamin B1 duplications cause

autosomal dominant leukodystrophy. Nat. Genet. 38, 1114–1123.

Parnaik, V.K., Chaturvedi, P., and Muralikrishna, B. (2011). Lamins, laminopa-

thies and disease mechanisms: possible role for proteasomal degradation of

key regulatory proteins. J. Biosci. 36, 471–479.

Piercy, R.J., Zhou, H., Feng, L., Pombo, A., Muntoni, F., and Brown, S.C.

(2007). Desmin immunolocalisation in autosomal dominant Emery-Dreifuss

muscular dystrophy. Neuromuscul. Disord. 17, 297–305.

Puckelwartz, M., and McNally, E.M. (2011). Emery-Dreifuss muscular

dystrophy. Handb. Clin. Neurol. 101, 155–166.

Puente, X.S., Quesada, V., Osorio, F.G., Cabanillas, R., Cadinanos, J., Fraile,

J.M., Ordonez, G.R., Puente, D.A., Gutierrez-Fernandez, A., Fanjul-Fernandez,

M., et al. (2011). Exome sequencing and functional analysis identifies BANF1

mutation as the cause of a hereditary progeroid syndrome. Am. J. Hum. Genet.

88, 650–656.

Quijano-Roy, S., Mbieleu, B., Bonnemann, C.G., Jeannet, P.Y., Colomer, J.,

Clarke, N.F., Cuisset, J.M., Roper, H., De Meirleir, L., D’Amico, A., et al.

(2008). De novo LMNA mutations cause a new form of congenital muscular

dystrophy. Ann. Neurol. 64, 177–186.

Ramos, F.J., Chen, S.C., Garelick, M.G., Dai, D.F., Liao, C.Y., Schreiber, K.H.,

Mackay, V.L., An, E.H., Strong, R., Ladiges, W.C., et al. (2012). Rapamycin

reverses elevated mTORC1 signaling in lamin A/C-deficient mice, rescues

cardiac and skeletal muscle function, and extends survival. Sci. Transl. Med.

4, 144ra103.

Redwood, A.B., Perkins, S.M., Vanderwaal, R.P., Feng, Z., Biehl, K.J.,

Gonzalez-Suarez, I., Morgado-Palacin, L., Shi, W., Sage, J., Roti-Roti, J.L.,

et al. (2011). A dual role for A-type lamins in DNA double-strand break repair.

Cell Cycle 10, 2549–2560.

Renou, L., Stora, S., Yaou, R.B., Volk, M., Sinkovec, M., Demay, L., Richard,

P., Peterlin, B., and Bonne, G. (2008). Heart-hand syndrome of Slovenian

type: a new kind of laminopathy. J. Med. Genet. 45, 666–671.

Rober, R.A., Weber, K., and Osborn, M. (1989). Differential timing of nuclear

lamin A/C expression in the various organs of the mouse embryo and the

young animal: a developmental study. Development 105, 365–378.

Rodrıguez, J., Calvo, F., Gonzalez, J.M., Casar, B., Andres, V., and Crespo, P.

(2010). ERK1/2 MAP kinases promote cell cycle entry by rapid, kinase-

independent disruption of retinoblastoma-lamin A complexes. J. Cell Biol.

191, 967–979.

Scaffidi, P., and Misteli, T. (2008). Lamin A-dependent misregulation of adult

stem cells associated with accelerated ageing. Nat. Cell Biol. 10, 452–459.

Shackleton, S., Lloyd, D.J., Jackson, S.N., Evans, R., Niermeijer, M.F., Singh,

B.M., Schmidt, H., Brabant, G., Kumar, S., Durrington, P.N., et al. (2000).

LMNA, encoding lamin A/C, is mutated in partial lipodystrophy. Nat. Genet.

24, 153–156.

Shalev, S.A., De Sandre-Giovannoli, A., Shani, A.A., and Levy, N. (2007). An

association of Hutchinson-Gilford progeria and malignancy. Am. J. Med.

Genet. A. 143A, 1821–1826.

Sullivan, T., Escalante-Alcalde, D., Bhatt, H., Anver, M., Bhat, N., Nagashima,

K., Stewart, C.L., and Burke, B. (1999). Loss of A-type lamin expression

compromises nuclear envelope integrity leading to muscular dystrophy. J.

Cell Biol. 147, 913–920.

Sun, L.P., Wang, L., Wang, H., Zhang, Y.H., and Pu, J.L. (2010). Connexin

43 remodeling induced by LMNA gene mutation Glu82Lys in familial dilated

cardiomyopathy with atrial ventricular block. Chin. Med. J. (Engl.) 123, 1058–

1062.

Taylor, M.R., Slavov, D., Gajewski, A., Vlcek, S., Ku, L., Fain, P.R., Carniel, E.,

Di Lenarda, A., Sinagra, G., Boucek, M.M., et al.; Familial Cardiomyopathy

Registry Research Group. (2005). Thymopoietin (lamina-associated polypep-

tide 2) gene mutation associated with dilated cardiomyopathy. Hum. Mutat.

26, 566–574.

Tilgner, K., Wojciechowicz, K., Jahoda, C., Hutchison, C., and Markiewicz, E.

(2009). Dynamic complexes of A-type lamins and emerin influence adipogenic

capacity of the cell via nucleocytoplasmic distribution of beta-catenin. J. Cell

Sci. 122, 401–413.

van Engelen, B.G., Muchir, A., Hutchison, C.J., van der Kooi, A.J., Bonne, G.,

and Lammens, M. (2005). The lethal phenotype of a homozygous nonsense

mutation in the lamin A/C gene. Neurology 64, 374–376.

Varela, I., Cadinanos, J., Pendas, A.M., Gutierrez-Fernandez, A., Folgueras,

A.R., Sanchez, L.M., Zhou, Z., Rodrıguez, F.J., Stewart, C.L., Vega, J.A.,

et al. (2005). Accelerated ageing in mice deficient in Zmpste24 protease is

linked to p53 signalling activation. Nature 437, 564–568.

Varela, I., Pereira, S., Ugalde, A.P., Navarro, C.L., Suarez, M.F., Cau, P.,

Cadinanos, J., Osorio, F.G., Foray, N., Cobo, J., et al. (2008). Combined treat-

ment with statins and aminobisphosphonates extends longevity in a mouse

model of human premature aging. Nat. Med. 14, 767–772.

Varga, R., Eriksson, M., Erdos, M.R., Olive, M., Harten, I., Kolodgie, F., Capell,

B.C., Cheng, J., Faddah, D., Perkins, S., et al. (2006). Progressive vascular

smooth muscle cell defects in a mouse model of Hutchinson-Gilford progeria

syndrome. Proc. Natl. Acad. Sci. USA 103, 3250–3255.

Vaughan, A., Alvarez-Reyes, M., Bridger, J.M., Broers, J.L., Ramaekers, F.C.,

Wehnert, M., Morris, G.E., Hutchison, C.J., and Hutchison, C.J.; Whitfield

WGF. (2001). Both emerin and lamin C depend on lamin A for localization at

the nuclear envelope. J. Cell Sci. 114, 2577–2590.

Verstraeten, V.L., Broers, J.L., van Steensel, M.A., Zinn-Justin, S., Ram-

aekers, F.C., Steijlen, P.M., Kamps, M., Kuijpers, H.J., Merckx, D., Smeets,

H.J., et al. (2006). Compound heterozygosity for mutations in LMNA

causes a progeria syndrome without prelamin A accumulation. Hum. Mol.

Genet. 15, 2509–2522.

Viteri, G., Chung, Y.W., and Stadtman, E.R. (2010). Effect of progerin on the

accumulation of oxidized proteins in fibroblasts from Hutchinson Gilford

progeria patients. Mech. Ageing Dev. 131, 2–8.

Waterham, H.R., Koster, J., Mooyer, P., Noort Gv, Gv., Kelley, R.I., Wilcox,

W.R., Wanders, R.J., Hennekam, R.C., and Oosterwijk, J.C. (2003). Autosomal

recessive HEM/Greenberg skeletal dysplasia is caused by 3 beta-hydrox-

ysterol delta 14-reductase deficiency due to mutations in the lamin B receptor

gene. Am. J. Hum. Genet. 72, 1013–1017.

Weber, K., Plessmann, U., and Traub, P. (1989). Maturation of nuclear lamin A

involves a specific carboxy-terminal trimming, which removes the polyisopre-

nylation site from the precursor; implications for the structure of the nuclear

lamina. FEBS Lett. 257, 411–414.

Wu, W., Shan, J., Bonne, G., Worman, H.J., and Muchir, A. (2010). Pharma-

cological inhibition of c-Jun N-terminal kinase signaling prevents cardiomy-

opathy caused by mutation in LMNA gene. Biochim. Biophys. Acta 1802,

632–638.

Yang, S.H., Andres, D.A., Spielmann, H.P., Young, S.G., and Fong, L.G.

(2008a). Progerin elicits disease phenotypes of progeria in mice whether or

not it is farnesylated. J. Clin. Invest. 118, 3291–3300.

Yang, S.H., Qiao, X., Fong, L.G., andYoung, S.G. (2008b). Treatmentwith a far-

nesyltransferase inhibitor improves survival in mice with a Hutchinson-Gilford

progeria syndrome mutation. Biochim. Biophys. Acta 1781, 36–39.

Yang, S.H., Chang, S.Y., Ren, S., Wang, Y., Andres, D.A., Spielmann, H.P.,

Fong, L.G., and Young, S.G. (2011). Absence of progeria-like disease pheno-

types in knock-in mice expressing a non-farnesylated version of progerin.

Hum. Mol. Genet. 20, 436–444.

Young, S.G., Meta, M., Yang, S.H., and Fong, L.G. (2006). Prelamin A farnesy-

lation and progeroid syndromes. J. Biol. Chem. 281, 39741–39745.

Zhang, Q., Bethmann, C., Worth, N.F., Davies, J.D., Wasner, C., Feuer, A.,

Ragnauth, C.D., Yi, Q., Mellad, J.A., Warren, D.T., et al. (2007). Nesprin-1

and -2 are involved in the pathogenesis of Emery Dreifuss muscular

dystrophy and are critical for nuclear envelope integrity. Hum. Mol. Genet.

16, 2816–2833.

Zhang, H., Kieckhaefer, J.E., and Cao, K. (2013). Mouse models of laminopa-

thies. Aging Cell 12, 2–10. Published online November 26, 2012. http://dx.doi.

org/10.1111/acel.12021.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1375

Leading Edge

Review

Nuclear Positioning

Gregg G. Gundersen1,* and Howard J. Worman1,2,*1Department of Pathology and Cell Biology2Department of Medicine

College of Physicians and Surgeons, Columbia University, 630 West 168th Street, New York, NY 10032, USA*Correspondence: ggg1@columbia.edu (G.G.G.), hjw14@columbia.edu (H.J.W.)

http://dx.doi.org/10.1016/j.cell.2013.02.031

The nucleus is the largest organelle and is commonly depicted in the center of the cell. Yet duringcell division, migration, and differentiation, it frequently moves to an asymmetric position alignedwith cell function. We consider the toolbox of proteins that move and anchor the nucleus withinthe cell and how forces generated by the cytoskeleton are coupled to the nucleus to move it. Thesignificance of proper nuclear positioning is underscored by numerous diseases resulting fromgenetic alterations in the toolbox proteins. Finally, we discuss how nuclear position may influencecellular organization and signaling pathways.

IntroductionDiagrams in biology textbooks usually depict the nucleus as

a spheroid in the center of the cell. However, the position of

nuclei varies dramatically from this simple view. Nuclei are

frequently positioned asymmetrically depending on cell type,

stage of the cell cycle, migratory state, and differentiation status.

For example, during cell division in budding yeast, nuclei are

moved into the bud neck so that each daughter cell receives

one (Figure 1A). Nuclei are actively positioned in the middle of

the fission yeast S. pombe, ensuring that the division plane

produces two equal daughter cells. In fertilized mammalian

and invertebrate eggs, male and female pronuclei move toward

each other and fuse near the middle of the zygote, ensuring that

the ensuing cell division creates two equal daughter blasto-

meres. Asymmetric divisions—typical of early embryos and

stem cells—frequently reflect a prepositioning of the nucleus.

Though nuclear positioning to affect the cell division plane

makes intuitive sense, asymmetric positioning occurs in nondi-

viding cells, where the purpose is less obvious. For example, in

the developing optic epithelium in Drosophila, nuclei move

basally and then apically to establish the characteristic arrange-

ment of cells in the ommatidium (Figure 1A). An analogousmove-

ment of nuclei occurs over the cell cycle in the developing

vertebrate neuroepithelium. In most migrating cells, the nucleus

is positioned in the rear, well removed from the protruding front

(Figure 1B). Nuclei in numerous differentiated animal tissues,

such as skeletal muscle, many epithelia, and neurons, are also

asymmetrically positioned (Figure 1C and Table1). These exam-

ples suggest that nuclei are positioned for specialized cellular

functions and that abnormal positioning could lead to dysfunc-

tion and disease.

Position of nuclei can be modified secondarily to changes in

cytoplasmic organization. For example, when macrovesicular

fat accumulates in hepatocytes in alcoholic or nonalcoholic stea-

tosis, nuclei are forced to the cell’s periphery. Similar changes in

nuclear position may occur in cells with abundant secretory

granules. However, recent research has discovered regulated,

1376 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

cytoplasmic mechanical systems that function primarily to exert

forces on the nucleus via connections to the nuclear envelope.

These systems maintain the position of the nucleus or move it

during processes such as cell migration and differentiation.

Though their role in homeostatic nuclear positioning is poorly

understood, mechanistic details are being deciphered in cases

where nuclei move.

We review systems in which progress is being made in under-

standing nuclear movement and positioning, and we identify

the molecular toolbox that cells use for these processes. This

toolbox includes specific nuclear envelope connections to

cytoskeletal force-generating systems. We then evaluate how

this toolbox is employed and identify conserved mechanisms

that use microtubules (MTs) and actin filaments as force gener-

ators. Genes encoding toolbox proteins are targets of mutations

that cause disease, raising the possibility that inappropriate

nuclear positioning contributes to pathogenesis. As active

nuclear movement suggests that its relative position may influ-

ence other cellular systems, we consider the significance of

nuclear positioning for cytoskeletal organization, signaling, and

transcriptional control.

The Nuclear Positioning ToolboxThe molecular toolbox for nuclear positioning contains: (1)

elements of the cytoskeleton and (2) protein complexes of the

nuclear envelope. The cytoskeletal elements generate forces

to move the nucleus. The protein complexes spanning the

nuclear membranes mediate attachment of cytoskeletal ele-

ments to the nucleoskeleton (Figure 2).

Cytoskeletal Elements

Actin filaments, MTs, and associatedmotor proteins are the prin-

cipal cytoskeletal elements of the nuclear positioning toolbox.

Cytoplasmic intermediate filaments may also play a role, but

this is currently poorly defined. In some cases, a single cytoskel-

etal element drives nuclear movement, as in MT-dependent

movement of male and female pronuclei after fertilization and

actin-dependent rearward movement of nuclei in fibroblasts

polarizing for migration. In other cases, MTs and actin filaments

collaborate to move nuclei, as in migrating neuronal cells. The

role of these cytoplasmic elements in different systems is dis-

cussed in detail below.

Protein Complexes in the Nuclear Envelope

An exciting advance in the past few years has been the identifi-

cation of the linker of nucleoskeleton and cytoskeleton (LINC)

complex in the nuclear envelope that mediates connections to

both MTs and actin filaments (Crisp et al., 2006). LINC

complexes are composed of outer nuclear membrane KASH

(klarsicht, Anc1, and Syne homology) proteins and inner nuclear

membrane SUN (Sad1 and Unc-83) proteins, both of which are

type II membrane proteins with a single transmembrane seg-

ment (Starr and Fridolfsson, 2010) (Figure 2A). KASH and SUN

proteins have been described in metazoan, fungi, and recently

plants (Razafsky and Hodzic, 2009; Zhou et al., 2012a). KASH

proteins are characterized by a conserved �60 residue KASH

domain at their C terminus, which includes a transmembrane

segment and up to 30 residues that project into the perinuclear

space between inner and outer nuclear membranes. KASH

domains in fungi and plants are less conserved than those in

metazoans. SUN proteins contain a conserved SUN domain

located within the perinuclear space. Five genes encode SUN

proteins in mammals, although only two of these (SUN1 and

SUN2) are widely expressed; lower eukaryotes have one or

two SUN proteins (Starr and Fridolfsson, 2010).

The crystal structure of SUN2 reveals an interesting mush-

room-like trimer with a ‘‘cap’’ composed of SUN domains and

a triple coiled-coil stalk, which is required for trimer formation

(Figure 2B) (Sosa et al., 2012; Zhou et al., 2012b). Predictions

of the length of this stalk suggest that the SUN protein could

span the nearly 50 nm between inner and outer nuclear

membranes (Sosa et al., 2012). Each SUN protein binds three

KASH peptides in deep grooves between adjacent SUN

domains in the trimer (Figure 2B). A KASH-SUN disulfide bond

may further stabilize the complex.

The trimeric SUN-KASH structure raises intriguing questions

about higher-ordered KASH-SUN protein assemblies, particu-

larly if KASH proteins are indeed dimericmolecules as predicted.

The binding pocket between SUN2 subunits suggests that it

will accommodate related KASH domains and that SUN1 and

SUN2 bind KASH proteins promiscuously (Starr and Fridolfsson,

2010). Yet there is an example in cells in which a specific KASH-

SUN pair assembles to move the nucleus (Luxton et al., 2011).

The apparent tight packing within the SUN-KASH complex

also raises questions about its assembly and regulation. KASH

and SUN proteins have diffusional mobilities similar to other

nuclear membrane proteins, indicating that they are likely in

dynamic complexes (Ostlund et al., 2009). TorsinA is a potential

regulator of the LINC complex, as it localizes to the ER lumen and

perinuclear space and shows affinity for KASH domains (Nery

et al., 2008; Tanabe et al., 2009). TorsinA’s homology to AAA

ATPases suggests that it may chaperone assembly or disas-

sembly of LINC complexes (Tanabe et al., 2009).

Specificity of LINC complexes is determined by the N termini

of KASH proteins, which are variable in size and ability to bind

cytoskeletal elements (Figure 2C). In mammals, KASH proteins

(termed nesprins) are encoded by five genes, some of which

generate multiple isoforms by alternative RNA splicing. The

‘‘giant’’ isoforms nesprin-1G and nesprin-2G (>800 kDa) en-

coded by SYNE1 and SYNE2, respectively, bind actin through

calponin homology (CH) domains near their N termini (Luxton

et al., 2011; Zhang et al., 2001). Much of their large cytoplas-

mic region is predicted to be composed of spectrin repeats, sug-

gesting a structure reminiscent of dystrophin with an extended

but flexible core and the potential for dimerization. Nesprin-1

and nesprin-2 isoforms also interact with the MT motors kine-

sin-1 and dynein, although whether binding is direct is unknown

(Yu et al., 2011; Zhang et al., 2009). In C. elegans, the KASH

protein Unc-83 interacts directly with kinesin-1, dynein, and

dynein regulators, including BicaudalD and NudE homologs (Fri-

dolfsson et al., 2010; Fridolfsson and Starr, 2010). Nesprin-3a,

an isoform encoded by SYNE3, binds the crosslinking protein

plectin, which binds cytoplasmic intermediate filaments (Wil-

helmsen et al., 2005). Nesprin-4 encoded by SYNE4 has a short

N terminus that associates with MTs through kinesin-1 and is

restricted in expression to highly secretory cells and hair cells

of the cochlea (Horn et al., 2013; Roux et al., 2009). Aside from

spectrin repeats, there are no other recognizable domains in

nesprins 1–4. A meiosis-specific ‘‘nesprin’’ termed KASH5

binds the dynein regulator dynactin (Morimoto et al., 2012).

Lower eukaryotes express actin- and MT motor-binding KASH

proteins, although there is less genetic complexity in these

organisms. For example, there are two KASH proteins in

Drosophila and four in C. elegans (Figure 2C) (Starr and Fridolfs-

son, 2010).

At the intranuclear side of the LINC complex, SUN proteins

bind to nuclear lamins (Crisp et al., 2006; Haque et al., 2006).

Lamins are intermediate filament proteins that polymerize to

form the nuclear lamina, a meshwork underlying the inner

nuclear membrane. Lamins A and C (A-type lamins), which are

alternative splice isoforms of the same gene, and lamins B1

and B2 are the predominant lamins expressed in differentiated

mammalian somatic cells. N termini of SUN1 and SUN2 bind

to lamin A, mediating their interaction with the lamina. Hence,

the LINC complex, via KASH protein interactions with cytoskel-

etal proteins and SUN protein interactions with lamins, connects

the nucleoskeleton to the cytoskeleton.

In mammalian cells lacking A-type lamins, SUN proteins still

localize to the nucleus (Crisp et al., 2006; Haque et al., 2006),

although they and their nesprin partners have increased mem-

brane diffusional mobility (Ostlund et al., 2009). This suggests

that other factors contribute to LINC complex anchoring. Indeed,

yeast lack lamins but still employ KASH and SUN proteins to

attach the nucleus to the cytoskeleton. In S. pombe, the hetero-

chromatin-binding protein Ima1 anchors the SUN protein Sad1,

a component of the spindle pole body (King et al., 2008). SAMP1,

the mammalian lma1 ortholog, localizes to LINC complex

assemblies that attach actin to the nucleus (Borrego-Pinto

et al., 2012). Emerin, which is an integral protein predominantly

localized to the inner nuclear membrane, binds to lamins and

nesprins (Mislow et al., 2002; Zhang et al., 2005). Additionally,

SUN1 associates with nuclear pore complexes (Liu et al., 2007).

LINC complex components constitute the major tools for con-

necting the nucleus to the cytoskeleton, yet they may not be the

only ones. Dynein interacts with Bicaudal2, which in turn binds to

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1377

Figure 1. Diversity of Nuclear Positioning(A) Schematics of nuclear positioning in dividing cells and developing epithelium. Arrows indicatemovements of nuclei (blue). The nucleus is positioned relative tothe plane of division in yeast and fertilized eggs. The diagram of insect optic epithelium (adapted from Patterson et al., 2004; Tomlinson and Ready, 1986)represents a longitudinal section of a larval eye disc; two nuclei are shown. Nuclei that are anterior (A) to the morphogenetic furrow (mf), which moves anteriorly,move basally. Nuclei that are posterior (P) to the furrow move apically as cells are recruited into clusters comprising ommatidium (white cells, cones; gray cells,R cells). The diagram of vertebrate neuroepithelium represents a longitudinal section of the developing cerebral cortex. Nuclei move basally during G1 andapically during G2. Mitosis (M) occurs near the apical surface. Adapted from Buchman and Tsai (2008) with permission.(B) Rearward nuclear position is typical of migrating cells. (Left) Schematic of a migrating cell with protruding leading edge and contracting tail. (Red)Actin filaments. (Right) Montage of migrating cells with front-back dimensions normalized. Dotted line represents the midpoint between the front andback. Nuclei are positioned along the front-back axis but always rearward of the cell center. Images reproduced with permission from: fibroblast (Gomes

(legend continued on next page)

1378 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Table 1. Nuclear Positions in Mammalian Cells and Tissues

Cell Tissue Nuclear Position Axis Alignmenta Comments

Proliferating Cells

Somatic cells central NA

Stem cells usually asymmetric various; niche related

Germ cells, oocytes asymmetric NA moves centrally after fertilization

Migrating Cells

1D (cultured fibroblast) asymmetric front-rear

2D (cultured; many types) asymmetric front-rear see Figure 1

3D (cultured fibroblast) asymmetric front-rear

3D (dermal sarcoma cells) asymmetric front-rear

3D (neurons in cortex) asymmetric front-rear

Macrophages, neutrophils asymmetric front-rear

Tissues

Muscle, skeletal asymmetric, complex peripheral-central clustered at neuromuscular junction

Muscle, cardiac central NA

Muscle, smooth central NA

Epithelia, squamous central NA

Epithelia, cuboidal central NA

Epithelia, columnar asymmetric apical-basal

Epithelia, pseudostratified asymmetric apical-basal cell-cycle dependent

Epithelia, secretory asymmetric apical-basal aligned with secretory axis

Neurons asymmetric proximal-distal

Astrocytes/oligodendricytes central NA

Connective Tissue

Osteoblasts/osteocytes central NA

Osteocytes, actively secreting asymmetric front-rear relative to secretory axis

Osteoclasts asymmetric front-rear

Chondroblasts/chondrocytes central NA

Chondrocytes, actively secreting asymmetric front-rear relative to secretory axis

Fibrocytes, resting central NA

Adipocytes asymmetric NA

Hematopoetic

Macrophages asymmetric front-rear

T cells, migrating or contacting target cell asymmetric front-rear

B cells, plasma cells asymmetric front-rearaPosition in italics.

RANBP2 at the cytoplasmic face of the nuclear pore complex

(Splinter et al., 2010). This association targets dynein to the

nucleus during G2 and may contribute to nuclear envelope

breakdown. However, it could be an alternative means to target

dynein for nuclear movement. Certain muscle-specific nuclear

membrane proteins accumulate along MTs, suggesting that

the nuclear positioning toolbox may also contain tissue-specific

tools (Wilkie et al., 2011).

et al., 2005), breast carcinoma (McNiven, 2013), keratocyte (Barnhart et al., 2010neuron (Godin et al., 2012).(C) Nuclear positioning in mammalian tissues. Cross-sections of kidney cortex acentrally in the distal (D) convoluted tubules and basally in proximal (P) convolutedbut are found centrally in dystrophic tissue.

Initiation of Nuclear MovementSpecific sets of tools become activated to move the nucleus in

response to stimuli. In pronuclear migrations in fertilized eggs,

formation of MTs by the sperm centrosome initiates movement

of bothmale and female pronuclei. Activation of the Rho GTPase

Cdc42 by the serum factor lysophosphatidic acid (LPA) initiates

nuclear movement in migrating fibroblasts by activating actin

retrograde flow (Gomes et al., 2005; Palazzo et al., 2001).

), endothelial cell (Tsai and Meyer, 2012), astrocyte (Osmani et al., 2006), and

nd skeletal muscle stained with hematoxylin and eosin. Nuclei are positionedtubules. Nuclei are positioned at the periphery of normal skeletal muscle fibers

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1379

Figure 2. Molecular Toolbox for Nuclear Movement/Positioning(A) Schematic of an idealized LINC complex in nuclear envelope. The innernuclear membrane (INM) SUNs bind within the perinuclear space to outernuclear membrane (ONM) KASH proteins. KASH proteins bind directly orindirectly to cytoskeletal filaments, including MTs, actin microfilaments, andcytoplasmic intermediate filaments. In metazoans, SUNs bind to the nuclearlamina; in yeast and plants, other intranuclear proteins bind SUNs. A nuclearpore complex (NPC) is shown for reference.(B) Side view of the structure of the SUN2-nesprin2 KASH complex. TrimericSUN2 domains are represented by different shades of blue, and the KASHpeptide is in orange. The structure illustrates the orientation of the KASH

1380 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Cdc42 is also essential for nuclear movements in neuronal

migration (Solecki et al., 2004) and neuronal precursors in the

neuroepithelium (Cappello et al., 2006). Nuclear movement in

the neuroepithelium is under cell-cycle control, and interference

with cell-cycle progression prevents it (Taverna and Huttner,

2010). These examples indicate that initiating nuclear move-

ments involves the de novo assembly of cytoskeletal compo-

nents of the toolbox. However, this is a fledgling area of inquiry,

and other processes such as activation of motors or relaxation of

nuclear anchoring may contribute to initiating nuclear move-

ment. Almost nothing is known about factors terminating nuclear

movement.

Characteristics of Nuclear MovementsNuclear movements occur in different cellular contexts and are

powered by different cytoskeletal elements. It is therefore not

surprising that they have different characteristics (Table 2).

Velocities vary between 0.1 and 1.0 mm/min, although peak rates

can be >10 mm/min for sperm pronuclei in Xenopus eggs.

Distances transversed during single episodes are generally one

nuclear diameter (5–10 mm) or less, although they are longer in

fertilized eggs and in the neuroepithelium. Nuclear movements

are usually continuous and unidirectional. However, high-

temporal-resolution imaging of nuclei in C. elegans hypodermal

cells revealed short pauses and bidirectional movements, sug-

gesting additional complexity (Fridolfsson and Starr, 2010).

During basal movement in the rat neuroepithelium, nuclei pause

for hour-long intervals before continuing in the same direction,

suggesting complex regulation. This diversity of nuclear move-

ments provided an early clue that there is more than one mech-

anism responsible.

MT-Mediated Nuclear MovementPioneering studies on invertebrate and vertebrate eggs revealed

that there are distinct mechanisms by which MTs connect to the

nucleus to move it (reviewed in Reinsch and Gonczy, 1998). The

male pronucleus, which forms after sperm entry into the egg,

nucleates MTs from its centrosome and moves toward the

middle of the cell. The female pronucleus laterally engages

MTs emanating from the male pronuclear-centrosome complex

and moves along them to join the male nucleus near the cell

center. Male pronuclear movement is generated by MT growth

and pushing along cortical sites and/or sites within the cyto-

plasm (Reinsch and Gonczy, 1998). Force is transmitted to the

nucleus through its intimate association with the centrosome

and centrosomal MTs. Female pronuclear movement is gener-

ated by attached cytoplasmic dynein motors that move it toward

MT minus ends at the sperm centrosome. Research on nuclear

peptide between adjacent SUN domains. Modified from Sosa et al. (2012) withpermission.(C) Schematic diagrams of KASH proteins from representative organisms andthe cytoskeletal filaments to which they bind. Binding to actin filaments ismediated by CH domains and binding to cytoplasmic intermediate filamentsby plectin. Binding toMTs ismediated by dynein and kinesins; direct binding toMTs has not been reported. The specific splice variants of nesprin-1 andnesprin-2 that interact with MT motors are unknown; for simplicity, a shortvariant of each is depicted. H.s., Homo sapiens; M.m., Mus musculus; D.m.,Drosophila melanogaster; C.e., Caenorhabditis elegans; S.p., Schizosacchar-omyces pombe.

Table 2. Physical Characteristics of Typical Nuclear Movements

System Rate (mm min�1) Distance (mm) Mode Dependence References

Fertilized Egg

Male pronucleus, Xenopus 16 100–300 ? MT polymerization Reinsch and Gonczy, 1998

Female pronucleus, Xenopus 0.2–1.5 100–300 ? dynein Reinsch and Gonczy, 1998

Migrating Neurons

Cortical brain slice 0.33 1–5 saltatory MT and myosin II Tsai et al., 2007

SVZ explants, matrigel 1.2–5 2–5 saltatory Schaar and McConnell, 2005

Granular neurons on radial glia 1.0 1.3 saltatory MT and myosin II Solecki et al., 2004;

Solecki et al., 2009

Radial Glia INM, Cortical Brain Slice

Basal directed 0.14 30–50 intermittent with

long pauses

kinesin3 Tsai et al., 2010

Apical directed 0.06 30–50 continuous dynein Tsai et al., 2007

Other Systems

Fibroblasts polarizing for migration 0.28–0.35 5–10 continuous actomyosin flow Gomes et al., 2005;

Luxton et al., 2010

Astrocytes polarizing for migration 0.05 �10 continuous actomyosin flow Dupin et al., 2011

Oocyte (D.m.) 0.07 5–10 continuous MT polymerization Zhao et al., 2012

Hypodermal cell (C.e.) 0.23 3.3 continuous kinesin1 Fridolfsson and Starr, 2010

Budding yeast 1.18 1–2 continuous dynein (and MT

depolymerization)

Adames and Cooper, 2000

D.m., Drosophila melanogaster; C.e., Caenorhabditis elegans.

movement has progressed from fertilized eggs to more molecu-

larly tractable systems, yet the idea that distinct MT-dependent

processes move the nucleus has persisted and has been

strengthened by newer studies.

Nuclear Movement by MT Pushing and Pulling Forces

In the male pronuclear form of nuclear movement, an MT orga-

nizing center (MTOC) connects the nucleus to MTs, and MT

dynamics power movement (Figure 3A). This form of nuclear

movement occurs before cell division in the budding yeast

S. cerevisiae (Adames and Cooper, 2000), the fission yeast

S. pombe (Tran et al., 2001), early C. elegans embryos (Gonczy

et al., 1999), Drosophila oocytes (Zhao et al., 2012), and

cultured mammalian cells (Levy and Holzbaur, 2008). The

MTOC is either embedded in the nuclear envelope (yeast

spindle pole body) or is tightly associated with it (other

systems). In C. elegans, the centrosome connects to the nuclear

envelope through the LINC complex proteins Zyg-12, a KASH

protein, and SUN1 (Malone et al., 2003). Outer nuclear

membrane Zyg-12 binds to dynein, moving the centrosome

close to the nucleus and promoting association between Zyg-

12 and a centrosomal splice variant lacking the transmembrane

domain. Zyg-12 is not conserved, so whether a similar mecha-

nism is present in other organisms is unclear. Defects in A-type

lamins and emerin increase spacing between the nucleus and

centrosome in mammalian cells (Lee et al., 2007; Salpingidou

et al., 2007); however, it is not clear that these proteins directly

link them.

For male pronuclear type of nuclear movement, forces are

generated by MTs interacting with cortical or cytoplasmic sites

(Figure 3A). The interaction can be simply physical or mediated

by anchored dynein. In S. pombe, interaction of growing MTs

with the periphery generates pushing forces that maintain the

nucleus in themiddle of the cell (Tran et al., 2001). Pushing forces

are restricted to systems in which relatively short distances

(�10 mm) are involved because longer MTs cannot withstand

compressive forces. Thus, in larger cells, MT pulling forces

contribute to centrosome movements. In most cases, pulling

forces are generated by cortically anchored dynein (Grill et al.,

2003; Schmoranzer et al., 2009), as originally described in

budding yeast, where dynein immobilized in the bud pulls on

spindle-pole-body-associated MTs, moving the nucleus toward

the bud neck (Adames and Cooper, 2000).

In syncytial cells with multiple nuclei, a more complex MT pull-

ing mechanism exists. In the filamentous fungus Aspergillus, in

which genetic screens revealed roles for dynein and its regula-

tors in nuclear positioning (Morris et al., 1998), MT anchoring at

cortical sites appears to evenly space nuclei in the syncytial

hyphae (Gladfelter and Berman, 2009). In differentiating insect

and mammalian muscle cells, which lack active centrosomes,

MT minus ends associate directly with the nuclear envelope

through uncharacterized factors. These cells also use dynein

pulling and MT sliding by kinesin-1 and MT-associated proteins

to cluster nuclei near the center of syncytial myotubes (Folker

et al., 2012; Metzger et al., 2012).

Nuclear Movement by Attached MT Motor Forces

In the female pronuclear form of nuclear movement, nuclei

associate laterally with MTs and move along them, powered

by nuclear-envelope-associated motors (Figure 3B). This is

typical of nuclear movements that occur during developmental

events. Genetic screens that identified KASH (Unc83) and SUN

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1381

Figure 3. Mechanisms of Nuclear Movement(A) Schematic of male pronuclear-type nuclear movement mediated by MTs(green). Forces (arrows) can be generated by MT polymerization, depoly-merization, or dynein motors (red) anchored in the cortex or cytoplasmic sites.(B) Schematic of female pronuclear-type nuclear movement mediated by MTdynein (red) and kinesin (orange) motors. Forces (arrows) are generated bymotors that laterally connect nuclei to MTs.(C) Schematic of actomyosin-type nuclear movement. Force (arrows) isgenerated by the actomyosin-dependent flow of dorsal actin cables (red).

(Unc84) proteins in C. elegans revealed that these proteins were

required for nuclear movement in various cell types (Starr and

Fridolfsson, 2010). Unc83 recruits both dynein and kinesin-1

motors to the nuclear envelope, where kinesin-1 is responsible

for moving the nucleus while dynein contributes to

directionality (Fridolfsson and Starr, 2010).

Female pronuclear-type nuclear movements are pronounced

in the developing nervous system. Early genetic screens in

Drosophila identified Klarsicht, or Klar, as a founding member

of the KASH protein family, and it is required for apical move-

ments of nuclei that establish the proper arrangement of cells

in the ommatidium (Mosley-Bishop et al., 1999). Klar function

has been linked to kinesin and dynein (Welte, 2004), suggesting

that it may recruit these MT motors to the nuclear envelope.

1382 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Mutants in the dynein regulators dynactin and DLis1 have similar

nuclear migration defects asKlarmutants (Fan and Ready, 1997;

Swan et al., 1999). Mutants in lamin Dm(0) and the SUN protein

klaroid disrupt Klar localization to the nuclear envelope and

apical movement of the nucleus, generating the same Klar

phenotype (Kracklauer et al., 2007; Patterson et al., 2004). This

result was the first to suggest that the nuclear lamina anchored

the LINC complex.

Female pronuclear-type nuclear movements occur during two

stages of vertebrate central nervous system development. In

neuroepithelial radial glial cells, which serve as neuronal precur-

sors, nuclear movement occurs along the apical-basal axis in

a cell-cycle-dependent fashion. This has been termed interki-

netic nuclear migration (INM). During INM, the nucleus moves

basally during G1 and returns during G2 to an apical location

where mitosis occurs (Taverna and Huttner, 2010). As the

centrosome remains apical, basal and apical movements occur

in MT plus and minus end directions, respectively. MT motors

have been implicated in these movements. The kinesin-3 family

member, Kif1a, has been implicated for plus-end-directed

movement and dynein for minus-end-directed movement (Tsai

et al., 2007; Tsai et al., 2010). Nesprin-1 and nesprin-2may serve

as recruitment factors for MT motors in vertebrate INM.

Knockout of their genes in mice and zebrafish leads to defective

INM in the neocortex and retina, and mouse nesprin-2 coimmu-

noprecipitates with dynein and kinesin-1 (Tsujikawa et al., 2007;

Yu et al., 2011; Zhang et al., 2009). Interfering with the dynactin

and Lis1 gives similar phenotypes (Tsai et al., 2005; Tsujikawa

et al., 2007). Nuclear movements in vertebrate INM may be

more complex than in the Drosophila eye, as myosin II and acto-

myosin contractility may also play a role (Norden et al., 2009;

Schenk et al., 2009).

The second stage of vertebrate central nervous system

development involving female pronuclear-type movements is

neuron migration. After their ‘‘birth’’ in the neuroepithelium,

neurons migrate significant distances to their final locations.

Most migrating neurons exhibit a characteristic two-stroke

form of migration in which the narrow leading process extends;

the centrosome thenmoves forward into a swelling in the leading

process followed by the nucleus and the rest of the soma (Tsai

and Gleeson, 2005). Nuclear movement toward MT minus ends

at the centrosome is dependent on dynein and its regulators

Lis1 and NudE (Shu et al., 2004; Tsai et al., 2007). The centro-

some also moves in a dynein- and Lis1-dependent fashion.

Lis1 binds to a specific nucleotide state of dynein and enhances

force generation, which may be necessary for moving the

nucleus (McKenney et al., 2010). Nesprin-2 and SUN1/SUN2,

which are also required for the forwardmovement of the nucleus,

may recruit dynein to the nucleus (Zhang et al., 2009). Doublecor-

tin, a MT-associated protein, is also required for nuclear move-

ment during neuron migration (Koizumi et al., 2006). Importantly,

nuclear movement during neuronal migration is also dependent

on actomyosin contraction (see below), so this is not a pure

form of female pronuclear-type movement.

The two-stroke mode of migration with a large separation

(5–18 mm) between the centrosome and nucleus is thought to

be a particular feature of neurons and is not typically observed

in other migrating cells. Nonetheless, the same anterior

Table 3. Genes Encoding Proteins Functioning in Nuclear Positioning Linked to Human Disease

Human Gene Protein Function Human Disease(s) Disease Phenotypes

DCX doublecortin stabilizes microtubules lissencephaly mislocalization of cortical neurons,

‘‘smooth brain’’

LIS1 Lis1 dynein regulation lissencephaly mislocalization of cortical neurons,

‘‘smooth brain’’

TUBA3 a-tubulin MT component lissencephaly mislocalization of cortical neurons,

‘‘smooth brain’’

LMNB1 lamin B1 lamina component adult onset leukodystrophy results

from gene duplication

demyelination

LMNB2 lamin B2 lamina component susceptibility to acquired partial

lipodystrophy

regional fat loss

SUN1 Sun1 LINC complex none to date

SUN2 Sun2 LINC complex none to date

SYNE1 nesprin-1 LINC complex (1) cerebellar ataxia; (2) myopathies;

(3) arthrogryposis

(1) coordination defects; 2) cardiomyopathy

and muscular dystrophy; 3) congenital joint

contractures and muscle weakness

SYNE2 nesprin-2 LINC complex myopathies cardiomyopathy and skeletal muscular

dystrophy

SYNE4 nesprin-4 LINC complex high-frequency hearing loss progressive high-frequency hearing loss

LMNA A-type lamins lamina components (1) myopathy; (2) partial lipodystrophy;

(3) peripheral neuropathy; (4) progeria

(1) cardiomyopathy with variable skeletal

muscular dystrophy; (2) fat loss from

extremities; (3) peripheral nerve defects;

(4) accelerated aging phenotypes

localization of the centrosome relative to the nucleus, albeit in

closer proximity, occurs inmanymigrating cell types, and dynein

has been implicated in nuclear movements in migrating non-

neuronal cells (Luxton and Gundersen, 2011).

Actin-Mediated Nuclear MovementA groundbreaking study in C. elegans identified an outer nuclear

membrane protein, termed Anc-1, which bound to actin and was

essential for anchoring nuclei in the syncytial hypodermal and

intestinal cells (Starr andHan, 2002). Anc-1 is one of the founding

members of the KASH protein family and requires the SUN

protein Unc84 for its outer nuclear membrane localization. While

the discovery of Anc-1 showed that nuclear connections to the

actin cytoskeleton anchor nuclei, we now know that nuclei are

also actively moved through actin-dependent processes, typi-

cally in cells polarizing for migration.

Nuclear Movement by Tethering to Moving Actin Cables

The rearward positioning of the nucleus in migrating cells

(Figure 1B) may result, at least in part, from an extension of the

leading edge. Yet, studies in a number of cultured cell types

have revealed that rearward nuclear positioning is an active

process independent of cell protrusion (Desai et al., 2009; Dupin

et al., 2011; Gomes et al., 2005; Luxton et al., 2010).

A direct mechanism for moving the nucleus has been estab-

lished in experiments utilizing wounded monolayers of serum-

starved fibroblasts treated with LPA, which stimulates cell polar-

ization, but not protrusion ormigration (Luxton et al., 2010). In this

system, rearward-moving dorsal actin cables induced by Cdc42

provide force to move the nucleus (Figure 3C). Movement of

dorsal actin cables is likely powered by myosin II, as its inhibition

prevents actin flow and nuclear movement (Gomes et al., 2005).

These cables are directly coupled to the nucleus by nesprin-2G

and SUN2, which accumulate along them to form linear assem-

blies termed transmembrane actin-associated nuclear (TAN)

lines. Actin-binding CH domains of nesprin-2G are required for

TAN line formation and nuclear movement. A-type lamins anchor

TAN lines to the nucleoskeleton, and in their absence, TAN

lines slip over an immobile nucleus (Folker et al., 2011). This

anchorage is presumably mediated through SUN2 binding to

A-type lamins. Additional anchorage may be mediated by

SUN2 binding to SAMP1, which also localizes to TAN lines and

is necessary for nuclear movement (Borrego-Pinto et al., 2012).

Nuclear Movement by Actomyosin Contraction

Nuclear movement appears to be rate limiting for cells migrating

through narrow extracellular spaces in which nuclei become

deformed (Friedl et al., 2011). In at least some of these cases,

passage through a narrow opening specifically requires myosin

II (Beadle et al., 2008), suggesting that actomyosin-mediated

nuclear movement is necessary. Myosin II is also necessary for

the forward movement of the nucleus in migrating neurons (Sol-

ecki et al., 2009; Tsai et al., 2007), localizing behind it where it

may provide contractile forces that help to move it into the

leading process. This may reflect the difficulty of moving the

nucleus into the narrow leading process, which requires nuclei

to become elongated.

Nuclear Positioning and DiseaseWe have provided several examples of nuclear positioning

events that are required for specific cellular processes. Given

this requirement, one could imagine that defects in themolecular

toolbox for nuclear positioning could lead to cellular dysfunction.

Indeed, results from human subjects with inherited diseases and

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1383

mousemodels have shown that alterations in proteins involved in

nuclear positioning are associated with pathology. Mutations in

genes encoding proteins involved in MT function, LINC complex

components, and the nuclear lamina all cause human diseases

(Table 3).

Lissencephaly is characterized by mislocalization of cortical

neurons, resulting in decreasedcortical complexity anda smooth

brain surface. Affected children have severe psychomotor retar-

dation, seizures, muscle spasticity, and failure to thrive. At the

cellular level, neuronal migration required for brain development

is blocked. Most cases of ‘‘classic’’ lissencephaly are caused by

deletion or truncating mutations in LIS1 (Reiner et al., 1993). The

Lis1 protein is required for INM and nuclear and centrosomal

movement during two-stroke neuronal migration (Shu et al.,

2004; Tsai et al., 2007). Similarly, mutations in DCX encoding

doublecortin cause X-linked lissencephaly and defective nuclear

movement in neurons (Gleeson et al., 1998; Koizumi et al., 2006).

De novomutations in TUBA3 encoding a-1 tubulin also cause lis-

sencephaly and defective nuclear movement in neurons (Keays

et al., 2007).

Intriguingly, depletion of lamin B1, lamin B2, or both in mice

causes lissencephaly-like phenotypes (Coffinier et al., 2010,

2011). These phenotypes result from neuronal migration defects,

which likely have accompanying abnormalities in nuclear move-

ment, although this has not been assessed directly. Nuclei spin

in mouse fibroblasts lacking lamin B1, suggesting that B-type

lamins function in nuclear anchoring (Ji et al., 2007). B-type lam-

ins may therefore anchor LINC complexes. Mutations in genes

encoding B-type lamins have not yet been linked to human

developmental brain disorders, but duplications in LMNB1

cause overexpression of lamin B1 and an adult-onset demyelin-

ating disease (Padiath et al., 2006).

Experiments in knockout mice implicate SUN1, SUN2,

nesprin-1, and nesprin-2 in nuclear migration during neurogene-

sis and migration (Zhang et al., 2009). However, mutations in

genes encoding nesprins have been linked to diseases other

than lissencephaly. Mutations in SYNE1 encoding nesprin-1

cause adult-onset autosomal-recessive cerebellar ataxia char-

acterized by diffuse cerebellar atrophy and impaired walking,

dysarthria, and poor coordination (Gros-Louis et al., 2007).

This could potentially result from neuronal migration defects in

a specific region of the brain. Mutations in SYNE1 have also

been reported to cause an autosomal-recessive form of arthrog-

ryposis multiplex congenita characterized by congenital joint

contractures, muscle weakness, and progressive motor decline

(Attali et al., 2009). Mutations in SYNE1 and SYNE2 have further

been reported to cause Emery-Dreifuss muscular dystrophy

(EDMD)-like phenotypes (Zhang et al., 2007a). Mutations in the

gene encoding the LINC-complex-associated protein emerin

were first reported to cause X-linked EDMD (Bione et al.,

1994), and mutations in LMNA-encoding A-type lamins are

responsible for most autosomally inherited cases (Bonne et al.,

1999). This suggests an association between LINC complex

function and EDMD-like phenotypes, which generally share a

dilated cardiomyopathy with variable skeletal muscle involve-

ment. More recently, mutation in SYNE4 encoding nesprin-4

has been shown to cause autosomal-recessive, progressive

high-frequency hearing loss (Horn et al., 2013).

1384 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Nuclear positioning defects caused by SYNE1 and SYNE2

mutations have been described. One patient with a SYNE1

mutation and cerebellar ataxia was reported to have fewer

muscle nuclei under neuromuscular junctions (Gros-Louis

et al., 2007). Similarly, deletion of the KASH domain from

nesprin-1 in mice abolishes synaptic nuclei clustering and

disrupts spacing of nonsynaptic nuclei in skeletal muscle;

deletion of the nesprin-2 KASH domain has no effect but exac-

erbates the defect in mice lacking nesprin-1 (Zhang et al.,

2007b). Nesprin-2 deletion in mice disrupts nuclear movement

in cells of the neocortex and retina, causing reduced thickness

of the cortex and the outer nuclear layer into which newly

formed photoreceptor cells migrate (Yu et al., 2011; Zhang

et al., 2009). Mice lacking nesprin-4 suffer from deafness,

mimicking the human mutation phenotype, and have abnormal

positioning of nuclei in cochlear outer hair cells (Horn et al.,

2013). Although no disease-causing mutations in SUN1 or

SUN2 have been described in humans, depletion of both

proteins from mice cause nuclear positioning defects in muscle,

retina, and developing brain, similar to those in mice lacking

nesprin-1 and nesprin-2 (Lei et al., 2009; Yu et al., 2011; Zhang

et al., 2009). Mice without SUN1 also have hearing loss and

abnormal nuclear positioning in cochlear outer hair cells (Horn

et al., 2013).

The tissue-selective human diseases and pathology in mice

that occur in response to alterations in different SUNs and

nesprins may result because only certain isoforms are necessary

in different tissues. Data frommice demonstrate tissue-selective

differences in the expression of nesprins and SUNs, yet there is

no comprehensive analysis of the expression patterns and

tissue-type functionality of all of the different nesprins and

SUNs. Results from knockout mice also suggest redundancy in

the function of SUN1 and SUN2 and different tissue effects of

nesprin-1 and nesprin-2.

Mutations in LMNA encoding the A-type lamins cause

a broad range of human diseases often referred to as ‘‘lamino-

pathies’’ (Dauer and Worman, 2009). LMNA mutations that

cause EDMD and related myopathies are mostly missense or

small in-frame deletions, which lead to expression of variant

proteins, splice site truncations, or promoter mutations.

Depletion of A-type lamins from mice leads primarily to cardiac

and skeletal muscle phenotypes, suggesting that LMNA muta-

tions, even dominant ones leading to variant protein expres-

sion, somehow cause loss of function (Sullivan et al., 1999).

Skeletal muscles from humans with autosomal dominant

EDMD and Lmna null mice both have nuclei in the center of

myofibers rather than at their normal peripheral localization.

However, this also occurs in other myopathies not associated

with defects in proteins directly implicated in nuclear posi-

tioning. For more on laminopathies, please see the Review by

Schreiber and Kennedy on page 1365 of this issue (Schreiber

and Kennedy, 2013).

In migrating fibroblasts depleted of A-type lamins or express-

ing variants associated with myopathy, actin-dependent rear-

ward nuclear movement fails to occur (Folker et al., 2011). In

these cells, nesprin-2G assembles into TAN lines that slip over

the nucleus rather than moving with it, indicating an anchorage

defect. Amino acid substitutions within an immunoglobulin-like

motif in the tail of A-type lamins cause partial lipodystrophy,

which is characterized by fat loss from the extremities. In

contrast to those causing myopathy, expression of lamin A

variants that cause lipodystrophy inhibit MT-dependent centro-

some positioning, but not actin-dependent nuclear movement

in migrating fibroblasts (Folker et al., 2011).

Except for cases in which nuclear positioning defects asso-

ciate with abnormal neuronal migration, the relationship of the

positioning defects observed in model systems to pathogenic

mechanisms remains uncertain. It is not known why alterations

in the nuclear positioning proteins affect only cells in certain

tissues when the proteins are widely expressed. In some

instances, observed nuclear positioning defects may not directly

connect to the disease, such as mispositioning of nuclei at the

neuromuscular junction in cerebellar ataxia. Overall, alterations

in the nuclear positioning toolbox most often affect tissues,

such as the nervous system and striated muscle, in which cell

migration plays an important role in organ development or

homeostasis. Abnormal force transmission between the nucleus

and cytoplasm may also render cells more susceptible to

damage by mechanical stress, leading to activation of stress

response or apoptotic pathways, resulting, respectively, in cell

dysfunction or death.

Cellular Significance of Nuclear Positioning:Hypotheses and PerspectivesOur understanding of why cells move and position their nuclei

is still rudimentary. Yet, interfering with proteins involved in

nuclear movement inhibits many cell functions. Defects in

the nuclear positioning toolbox also cause disease. Thus,

nuclear positioning itself may influence other cellular activities.

Here, we put into perspective evidence supporting the hypoth-

eses that nuclear positioning influences the organization and

mechanical properties of the surrounding cytoplasm, cyto-

plasmic signaling, and accessibility of the nucleus to signaling

pathways.

The Nuclear Envelope as a Cytoskeletal Integrator

Identification of the LINC complex and other proteins mediating

nucleocytoskeletal connections raises the possibility that the

nucleus not only attaches to the cytoskeleton, but also organizes

it. Even before the identification of specific nucleocytoskeletal

connectors, a classical experiment by Ingber and colleagues re-

vealed that the nucleus was physically connected to integrins in

the plasma membrane (Maniotis et al., 1997). These investiga-

tors showed that applying force to fibronectin beads attached

to integrins moved the nucleus tens of microns away. Although

the nature of the connection was not identified, this observation

clearly reflects linkages that exist between the nucleus and the

plasma membrane.

The nucleus influences the MT cytoskeleton through its asso-

ciation with MTOCs, which determine where MT minus ends

are anchored. A more direct influence of the nucleus on MT

distribution occurs in cells with noncentrosomal MTs. In multi-

nucleated myotubes, which lack functional centrosomes, MTs

minus ends are attached to nuclei by unidentified linkers,

contributing to an overall bipolar array of MTs with mixed

polarity (Tassin et al., 1985). The nucleus may also affect orga-

nization of the actin cytoskeleton. CH-domain-containing

nesprins tether the nucleus to actin filaments, but whether

they organize actin arrays around it is less certain. In fibroblasts

polarizing for migration, depleting nesprin-2G or A-type lamins

does not alter the overall distribution of actin filaments or the

formation and movement of dorsal actin cables (Folker et al.,

2011; Luxton et al., 2010). However, alterations in actin fila-

ments and focal adhesions have been reported when LINC

complex components are perturbed (Hale et al., 2008; Khatau

et al., 2009). This may reflect lack of direct connection of the

actin arrays to the nuclear envelope or indirect effects. These

findings suggest that, at least under some circumstances,

the nucleus actively participates in organizing certain actin

structures.

Additional evidence that the nucleus organizes the cytoplasm

comes from biophysical measurements. Cytoplasmic stiffness

adjacent and distal to the nucleus is altered in cells depleted of

A-type lamins (Broers et al., 2004; Lammerding et al., 2004).

Whether this result solely reflects direct physical links between

the nucleus and cytoskeleton or indirect effects of signaling

pathways that are also modified by alterations in the nuclear

envelope (see below) is presently unclear.

The Nuclear Envelope as a Regulator of Signaling

Pathways

As the largest and most compression-resistant membrane-

bound organelle in the cell, the nucleus has been likened to

a ‘‘molecular shock absorber’’ (Dahl et al., 2004). Theoretically,

movement of such a large, non-deformable organelle through

the cytoplasm will result in tensile and/or compressive forces.

Mediated by nuclear connections to the cytoskeleton, these

forces could be transmitted to distal sites that are mechanical

transducers, such as integrin-based focal adhesions or cad-

herin-based cell-cell adhesions (Leckband et al., 2011; Parsons

et al., 2010). In a sense, the nucleus would act like the bead

in Ingber’s experiment, except that force would originate

inside rather than outside of the cell. Given that adhesions

respond to mechanical stimuli by regulating Rho GTPase and

mitogen-activated protein (MAP) kinase signaling, the predic-

tion is that nuclear movement may affect the activity of these

pathways.

The idea that nuclear movement may regulate cellular

signaling pathways has not been directly tested. Yet there is

evidence that alterations in the nuclear movement toolbox alter

signaling pathways. Lamin A variants that cause myopathies

increase MAP kinase signaling, as does knockdown of A-type

lamins or emerin (Muchir et al., 2007, 2009). Similar results

have been obtained for Rho signaling (Hale et al., 2008). Given

that alterations in A-type lamins interfere with actin-dependent

nuclear movement (Folker et al., 2011), it is possible that

changes in signaling result from altered nuclear positioning.

A-type lamins may also affect signaling by interacting with

proteins in the pathway, for example, by binding the MAP kinase

ERK1/2 (Gonzalez et al., 2008). KASH proteins may recruit

signaling molecules to the nuclear envelope and regulate their

activities, as nesprin-2 binds active ERK1/2, and its knockdown

results in prolonged ERK1/2 activity (Warren et al., 2010). As

other actin-dependentmembrane structures such as focal adhe-

sions regulate signaling, TAN lines assembled on the surface of

the nuclear envelope may also.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1385

Nuclear Position as a Response Regulator of Signaling

Pathways

The position of the nucleus may also alter its responsiveness to

pathways that regulate transcription and mRNA transport and

localization. It is generally assumed that latent cytoplasmic tran-

scription factors and second messengers activated by plasma

membrane receptors reach the nucleus in an unabated fashion.

However, the distance that they travel may depend on encoun-

ters with costimulatory and inhibitory factors in the cytoplasm

(Calvo et al., 2010). Thus, the nucleus’s position relative to the

origin of an external signal may modulate its response. This

could be particularly important for asymmetrically encountered

signals, for example, on the apical or basal aspects of epithelia

or in gradients of external factors during development. The

spatial relationship between the nucleus and the primary cilium

changes in many developing epithelia, such as the neuroepithe-

lium, and may affect the output of signaling pathways, such as

the Sonic hedgehog pathway that requires the cilium (Goetz

and Anderson, 2010). Signaling from intracellular sites, such as

the signaling endosome, may enhance responsiveness by

bringing the signal in close proximity to the nucleus.

Only one study has directly examined the relationship

between nuclear position and asymmetrical signaling (Del

Bene et al., 2008). A gradient of Notch signaling, highest at

the apical surface, exists in the retinal neuroepithelium, as in

other epithelia (Murciano et al., 2002). INM moves the nucleus

basally during G1, exposing it to lower Notch activity. A muta-

tion in the zebrafish mok gene encoding the dynactin p150glued

subunit causes longer and faster basal nuclear excursions,

resulting in increased basal mitoses and the formation of early

differentiating neurons at the expense of later ones (Del Bene

et al., 2008). Notch overexpression rescues themok phenotype,

showing that it results from inadequate exposure of the nucleus

to Notch due to defective nuclear movement. Alterations in

Syne-2 lead to similar changes in INM and cell fate in zebrafish

retina (Tsujikawa et al., 2007). Deficiencies in Cep120 and

TACC, proteins that affect the centrosome-MT connection, or

in nesprin-2 or SUN1/2 also affect INM in developing mouse

cerebral cortex and lead to early depletion of neural progenitors

(Xie et al., 2007; Zhang et al., 2009). Although altered cell

fate has not yet been demonstrated in these studies, they

are consistent with altered response to Notch or other apical

signals.

ConclusionsRather than being a passive or random phenomenon, active

mechanisms exist to position nuclei in cells. We have reviewed

the molecular tools and mechanisms that move and position

nuclei, most of which are conserved among eukaryotes. Human

diseases result from genetic abnormalities in nuclear movement

toolbox proteins and, in some cases, are linked to altered nuclear

movement. We have highlighted potential mechanisms by which

nuclear position may influence cellular processes and disease

pathogenesis. Additional investigation is needed to understand

how the nucleus affects these processes and to separate direct

from indirect effects of its positioning. Future basic research on

nuclear positioning and how it affects cellular processes is likely

to significantly impact public health.

1386 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

ACKNOWLEDGMENTS

We thank Susumu Antoku, Wakam Chang, Edgar Gomes, Gant Luxton, and

Alex Palazzo for their comments and Wakam Chang for Figures 1B and 2A.

The authors are supported by NIH grants R01GM099481, R01NS059352,

R01HD070713, and R01AR048997.

REFERENCES

Adames, N.R., and Cooper, J.A. (2000). Microtubule interactions with the cell

cortex causing nuclear movements in Saccharomyces cerevisiae. J. Cell Biol.

149, 863–874.

Attali, R., Warwar, N., Israel, A., Gurt, I., McNally, E., Puckelwartz, M., Glick, B.,

Nevo, Y., Ben-Neriah, Z., and Melki, J. (2009). Mutation of SYNE-1,

encoding an essential component of the nuclear lamina, is responsible for

autosomal recessive arthrogryposis. Hum. Mol. Genet. 18, 3462–3469.

Barnhart, E.L., Allen, G.M., Julicher, F., and Theriot, J.A. (2010). Bipedal loco-

motion in crawling cells. Biophys. J. 98, 933–942.

Beadle, C., Assanah, M.C., Monzo, P., Vallee, R., Rosenfeld, S.S., and Canoll,

P. (2008). The role of myosin II in glioma invasion of the brain. Mol. Biol. Cell 19,

3357–3368.

Bione, S., Maestrini, E., Rivella, S., Mancini, M., Regis, S., Romeo, G., and

Toniolo, D. (1994). Identification of a novel X-linked gene responsible for

Emery-Dreifuss muscular dystrophy. Nat. Genet. 8, 323–327.

Bonne, G., Di Barletta, M.R., Varnous, S., Becane, H.M., Hammouda, E.H.,

Merlini, L., Muntoni, F., Greenberg, C.R., Gary, F., Urtizberea, J.A., et al.

(1999). Mutations in the gene encoding lamin A/C cause autosomal dominant

Emery-Dreifuss muscular dystrophy. Nat. Genet. 21, 285–288.

Borrego-Pinto, J., Jegou, T., Osorio, D.S., Aurade, F., Gorjanacz, M., Koch, B.,

Mattaj, I.W., and Gomes, E.R. (2012). Samp1 is a component of TAN lines and

is required for nuclear movement. J. Cell Sci. 125, 1099–1105.

Broers, J.L., Peeters, E.A., Kuijpers, H.J., Endert, J., Bouten, C.V., Oomens,

C.W., Baaijens, F.P., and Ramaekers, F.C. (2004). Decreasedmechanical stiff-

ness in LMNA-/- cells is caused by defective nucleo-cytoskeletal integrity:

implications for the development of laminopathies. Hum. Mol. Genet. 13,

2567–2580.

Buchman, J.J., and Tsai, L.H. (2008). Putting a notch in our understanding of

nuclear migration. Cell 134, 912–914.

Calvo, F., Agudo-Ibanez, L., and Crespo, P. (2010). The Ras-ERK pathway:

understanding site-specific signaling provides hope of new anti-tumor thera-

pies. Bioessays 32, 412–421.

Cappello, S., Attardo, A., Wu, X., Iwasato, T., Itohara, S., Wilsch-Brauninger,

M., Eilken, H.M., Rieger, M.A., Schroeder, T.T., Huttner, W.B., et al. (2006).

The Rho-GTPase cdc42 regulates neural progenitor fate at the apical surface.

Nat. Neurosci. 9, 1099–1107.

Coffinier, C., Chang, S.Y., Nobumori, C., Tu, Y., Farber, E.A., Toth, J.I., Fong,

L.G., and Young, S.G. (2010). Abnormal development of the cerebral cortex

and cerebellum in the setting of lamin B2 deficiency. Proc. Natl. Acad. Sci.

USA 107, 5076–5081.

Coffinier, C., Jung, H.J., Nobumori, C., Chang, S., Tu, Y., Barnes, R.H., 2nd,

Yoshinaga, Y., de Jong, P.J., Vergnes, L., Reue, K., et al. (2011).

Deficiencies in lamin B1 and lamin B2 cause neurodevelopmental defects

and distinct nuclear shape abnormalities in neurons. Mol. Biol. Cell 22,

4683–4693.

Crisp, M., Liu, Q., Roux, K., Rattner, J.B., Shanahan, C., Burke, B., Stahl, P.D.,

and Hodzic, D. (2006). Coupling of the nucleus and cytoplasm: role of the LINC

complex. J. Cell Biol. 172, 41–53.

Dahl, K.N., Kahn, S.M., Wilson, K.L., and Discher, D.E. (2004). The nuclear

envelope lamina network has elasticity and a compressibility limit suggestive

of a molecular shock absorber. J. Cell Sci. 117, 4779–4786.

Dauer, W.T., and Worman, H.J. (2009). The nuclear envelope as a signaling

node in development and disease. Dev. Cell 17, 626–638.

Del Bene, F., Wehman, A.M., Link, B.A., and Baier, H. (2008). Regulation of

neurogenesis by interkinetic nuclear migration through an apical-basal notch

gradient. Cell 134, 1055–1065.

Desai, R.A., Gao, L., Raghavan, S., Liu, W.F., and Chen, C.S. (2009). Cell

polarity triggered by cell-cell adhesion via E-cadherin. J. Cell Sci. 122,

905–911.

Dupin, I., Sakamoto, Y., and Etienne-Manneville, S. (2011). Cytoplasmic inter-

mediate filaments mediate actin-driven positioning of the nucleus. J. Cell Sci.

124, 865–872.

Fan, S.S., and Ready, D.F. (1997). Glued participates in distinct microtubule-

based activities in Drosophila eye development. Development 124, 1497–

1507.

Folker, E.S., Ostlund, C., Luxton, G.W., Worman, H.J., and Gundersen, G.G.

(2011). Lamin A variants that cause striated muscle disease are defective in

anchoring transmembrane actin-associated nuclear lines for nuclear move-

ment. Proc. Natl. Acad. Sci. USA 108, 131–136.

Folker, E.S., Schulman, V.K., and Baylies, M.K. (2012). Muscle length and

myonuclear position are independently regulated by distinct Dynein pathways.

Development 139, 3827–3837.

Fridolfsson, H.N., and Starr, D.A. (2010). Kinesin-1 and dynein at the nuclear

envelope mediate the bidirectional migrations of nuclei. J. Cell Biol. 191,

115–128.

Fridolfsson, H.N., Ly, N., Meyerzon, M., and Starr, D.A. (2010). UNC-83 coor-

dinates kinesin-1 and dynein activities at the nuclear envelope during nuclear

migration. Dev. Biol. 338, 237–250.

Friedl, P., Wolf, K., and Lammerding, J. (2011). Nuclear mechanics during cell

migration. Curr. Opin. Cell Biol. 23, 55–64.

Gladfelter, A., and Berman, J. (2009). Dancing genomes: fungal nuclear posi-

tioning. Nat. Rev. Microbiol. 7, 875–886.

Gleeson, J.G., Allen, K.M., Fox, J.W., Lamperti, E.D., Berkovic, S., Scheffer, I.,

Cooper, E.C., Dobyns, W.B., Minnerath, S.R., Ross, M.E., and Walsh, C.A.

(1998). Doublecortin, a brain-specific gene mutated in human X-linked lissen-

cephaly and double cortex syndrome, encodes a putative signaling protein.

Cell 92, 63–72.

Godin, J.D., Thomas, N., Laguesse, S., Malinouskaya, L., Close, P., Malaise,

O., Purnelle, A., Raineteau, O., Campbell, K., Fero, M., et al. (2012).

p27(Kip1) is amicrotubule-associated protein that promotesmicrotubule poly-

merization during neuron migration. Dev. Cell 23, 729–744.

Goetz, S.C., and Anderson, K.V. (2010). The primary cilium: a signalling centre

during vertebrate development. Nat. Rev. Genet. 11, 331–344.

Gomes, E.R., Jani, S., and Gundersen, G.G. (2005). Nuclear movement regu-

lated by Cdc42, MRCK, myosin, and actin flow establishes MTOC polarization

in migrating cells. Cell 121, 451–463.

Gonczy, P., Pichler, S., Kirkham, M., and Hyman, A.A. (1999). Cytoplasmic

dynein is required for distinct aspects of MTOC positioning, including centro-

some separation, in the one cell stage Caenorhabditis elegans embryo. J. Cell

Biol. 147, 135–150.

Gonzalez, J.M., Navarro-Puche, A., Casar, B., Crespo, P., and Andres, V.

(2008). Fast regulation of AP-1 activity through interaction of lamin A/C,

ERK1/2, and c-Fos at the nuclear envelope. J. Cell Biol. 183, 653–666.

Grill, S.W., Howard, J., Schaffer, E., Stelzer, E.H., and Hyman, A.A. (2003). The

distribution of active force generators controls mitotic spindle position.

Science 301, 518–521.

Gros-Louis, F., Dupre, N., Dion, P., Fox, M.A., Laurent, S., Verreault, S., Sanes,

J.R., Bouchard, J.P., and Rouleau, G.A. (2007). Mutations in SYNE1 lead to

a newly discovered form of autosomal recessive cerebellar ataxia. Nat. Genet.

39, 80–85.

Hale, C.M., Shrestha, A.L., Khatau, S.B., Stewart-Hutchinson, P.J., Hernan-

dez, L., Stewart, C.L., Hodzic, D., and Wirtz, D. (2008). Dysfunctional

connections between the nucleus and the actin and microtubule networks in

laminopathic models. Biophys. J. 95, 5462–5475.

Haque, F., Lloyd, D.J., Smallwood, D.T., Dent, C.L., Shanahan, C.M., Fry,

A.M., Trembath, R.C., and Shackleton, S. (2006). SUN1 interacts with nuclear

lamin A and cytoplasmic nesprins to provide a physical connection between

the nuclear lamina and the cytoskeleton. Mol. Cell. Biol. 26, 3738–3751.

Horn, H.F., Brownstein, Z., Lenz, D.R., Shivatzki, S., Dror, A.A., Dagan-Rose-

nfeld, O., Friedman, L.M., Roux, K.J., Kozlov, S., Jeang, K.-T., et al. (2013). The

LINC complex is essential for hearing. J. Clin. Invest. 123, 740–750.

Ji, J.Y., Lee, R.T., Vergnes, L., Fong, L.G., Stewart, C.L., Reue, K., Young,

S.G., Zhang, Q., Shanahan, C.M., and Lammerding, J. (2007). Cell nuclei

spin in the absence of lamin B1. J. Biol. Chem. 282, 20015–20026.

Keays, D.A., Tian, G., Poirier, K., Huang, G.J., Siebold, C., Cleak, J., Oliver,

P.L., Fray, M., Harvey, R.J., Molnar, Z., et al. (2007). Mutations in alpha-tubulin

cause abnormal neuronal migration in mice and lissencephaly in humans. Cell

128, 45–57.

Khatau, S.B., Hale, C.M., Stewart-Hutchinson, P.J., Patel, M.S., Stewart, C.L.,

Searson, P.C., Hodzic, D., and Wirtz, D. (2009). A perinuclear actin cap regu-

lates nuclear shape. Proc. Natl. Acad. Sci. USA 106, 19017–19022.

King, M.C., Drivas, T.G., and Blobel, G. (2008). A network of nuclear envelope

membrane proteins linking centromeres to microtubules. Cell 134, 427–438.

Koizumi, H., Higginbotham, H., Poon, T., Tanaka, T., Brinkman, B.C., and

Gleeson, J.G. (2006). Doublecortin maintains bipolar shape and nuclear trans-

location during migration in the adult forebrain. Nat. Neurosci. 9, 779–786.

Kracklauer, M.P., Banks, S.M., Xie, X., Wu, Y., and Fischer, J.A. (2007).

Drosophila klaroid encodes a SUN domain protein required for Klarsicht local-

ization to the nuclear envelope and nuclear migration in the eye. Fly (Austin) 1,

75–85.

Lammerding, J., Schulze, P.C., Takahashi, T., Kozlov, S., Sullivan, T., Kamm,

R.D., Stewart, C.L., and Lee, R.T. (2004). Lamin A/C deficiency causes defec-

tive nuclear mechanics and mechanotransduction. J. Clin. Invest. 113,

370–378.

Leckband, D.E., le Duc, Q., Wang, N., and de Rooij, J. (2011). Mechanotrans-

duction at cadherin-mediated adhesions. Curr. Opin. Cell Biol. 23, 523–530.

Lee, J.S., Hale, C.M., Panorchan, P., Khatau, S.B., George, J.P., Tseng, Y.,

Stewart, C.L., Hodzic, D., and Wirtz, D. (2007). Nuclear lamin A/C deficiency

induces defects in cell mechanics, polarization, and migration. Biophys. J.

93, 2542–2552.

Lei, K., Zhang, X., Ding, X., Guo, X., Chen, M., Zhu, B., Xu, T., Zhuang, Y., Xu,

R., and Han, M. (2009). SUN1 and SUN2 play critical but partially redundant

roles in anchoring nuclei in skeletal muscle cells in mice. Proc. Natl. Acad.

Sci. USA 106, 10207–10212.

Levy, J.R., and Holzbaur, E.L. (2008). Dynein drives nuclear rotation during

forward progression of motile fibroblasts. J. Cell Sci. 121, 3187–3195.

Liu, Q., Pante, N., Misteli, T., Elsagga, M., Crisp, M., Hodzic, D., Burke, B., and

Roux, K.J. (2007). Functional association of Sun1 with nuclear pore

complexes. J. Cell Biol. 178, 785–798.

Luxton, G.W., and Gundersen, G.G. (2011). Orientation and function of the

nuclear-centrosomal axis during cell migration. Curr. Opin. Cell Biol. 23,

579–588.

Luxton, G.W., Gomes, E.R., Folker, E.S., Vintinner, E., and Gundersen, G.G.

(2010). Linear arrays of nuclear envelope proteins harness retrograde actin

flow for nuclear movement. Science 329, 956–959.

Luxton, G.W., Gomes, E.R., Folker, E.S., Worman, H.J., and Gundersen, G.G.

(2011). TAN lines: a novel nuclear envelope structure involved in nuclear posi-

tioning. Nucleus 2, 173–181.

Malone, C.J., Misner, L., Le Bot, N., Tsai, M.C., Campbell, J.M., Ahringer, J.,

and White, J.G. (2003). The C. elegans hook protein, ZYG-12, mediates the

essential attachment between the centrosome and nucleus. Cell 115,

825–836.

Maniotis, A.J., Chen, C.S., and Ingber, D.E. (1997). Demonstration of mechan-

ical connections between integrins, cytoskeletal filaments, and nucleoplasm

that stabilize nuclear structure. Proc. Natl. Acad. Sci. USA 94, 849–854.

McKenney, R.J., Vershinin, M., Kunwar, A., Vallee, R.B., and Gross, S.P.

(2010). LIS1 and NudE induce a persistent dynein force-producing state.

Cell 141, 304–314.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1387

McNiven, M.A. (2013). Breaking away: matrix remodeling from the leading

edge. Trends Cell Biol. 23, 16–21.

Metzger, T., Gache, V., Xu, M., Cadot, B., Folker, E.S., Richardson, B.E.,

Gomes, E.R., and Baylies, M.K. (2012). MAP and kinesin-dependent nuclear

positioning is required for skeletal muscle function. Nature 484, 120–124.

Mislow, J.M., Holaska, J.M., Kim, M.S., Lee, K.K., Segura-Totten, M., Wilson,

K.L., and McNally, E.M. (2002). Nesprin-1alpha self-associates and binds

directly to emerin and lamin A in vitro. FEBS Lett. 525, 135–140.

Morimoto, A., Shibuya, H., Zhu, X., Kim, J., Ishiguro, K., Han, M., and Wata-

nabe, Y. (2012). A conserved KASH domain protein associates with telomeres,

SUN1, and dynactin during mammalian meiosis. J. Cell Biol. 198, 165–172.

Morris, N.R., Efimov, V.P., and Xiang, X. (1998). Nuclear migration, nucleokine-

sis and lissencephaly. Trends Cell Biol. 8, 467–470.

Mosley-Bishop, K.L., Li, Q., Patterson, L., and Fischer, J.A. (1999). Molecular

analysis of the klarsicht gene and its role in nuclear migration within differenti-

ating cells of the Drosophila eye. Curr. Biol. 9, 1211–1220.

Muchir, A., Pavlidis, P., Decostre, V., Herron, A.J., Arimura, T., Bonne, G., and

Worman, H.J. (2007). Activation of MAPK pathways links LMNA mutations to

cardiomyopathy in Emery-Dreifuss muscular dystrophy. J. Clin. Invest. 117,

1282–1293.

Muchir, A., Wu, W., and Worman, H.J. (2009). Reduced expression of A-type

lamins and emerin activates extracellular signal-regulated kinase in cultured

cells. Biochim. Biophys. Acta 1792, 75–81.

Murciano, A., Zamora, J., Lopez-Sanchez, J., and Frade, J.M. (2002). Interki-

netic nuclear movement may provide spatial clues to the regulation of neuro-

genesis. Mol. Cell. Neurosci. 21, 285–300.

Nery, F.C., Zeng, J., Niland, B.P., Hewett, J., Farley, J., Irimia, D., Li, Y., Wiche,

G., Sonnenberg, A., and Breakefield, X.O. (2008). TorsinA binds the KASH

domain of nesprins and participates in linkage between nuclear envelope

and cytoskeleton. J. Cell Sci. 121, 3476–3486.

Norden, C., Young, S., Link, B.A., and Harris, W.A. (2009). Actomyosin is the

main driver of interkinetic nuclear migration in the retina. Cell 138, 1195–1208.

Osmani, N., Vitale, N., Borg, J.P., and Etienne-Manneville, S. (2006). Scrib

controls Cdc42 localization and activity to promote cell polarization during

astrocyte migration. Curr. Biol. 16, 2395–2405.

Ostlund, C., Folker, E.S., Choi, J.C., Gomes, E.R., Gundersen, G.G., and Wor-

man, H.J. (2009). Dynamics and molecular interactions of linker of nucleoske-

leton and cytoskeleton (LINC) complex proteins. J. Cell Sci. 122, 4099–4108.

Padiath, Q.S., Saigoh, K., Schiffmann, R., Asahara, H., Yamada, T., Koeppen,

A., Hogan, K., Ptacek, L.J., and Fu, Y.H. (2006). Lamin B1 duplications cause

autosomal dominant leukodystrophy. Nat. Genet. 38, 1114–1123.

Palazzo, A.F., Joseph, H.L., Chen, Y.J., Dujardin, D.L., Alberts, A.S., Pfister,

K.K., Vallee, R.B., and Gundersen, G.G. (2001). Cdc42, dynein, and dynactin

regulateMTOC reorientation independent of Rho-regulatedmicrotubule stabi-

lization. Curr. Biol. 11, 1536–1541.

Parsons, J.T., Horwitz, A.R., and Schwartz, M.A. (2010). Cell adhesion: inte-

grating cytoskeletal dynamics and cellular tension. Nat. Rev. Mol. Cell Biol.

11, 633–643.

Patterson, K., Molofsky, A.B., Robinson, C., Acosta, S., Cater, C., and Fischer,

J.A. (2004). The functions of Klarsicht and nuclear lamin in developmentally

regulated nuclear migrations of photoreceptor cells in the Drosophila eye.

Mol. Biol. Cell 15, 600–610.

Razafsky, D., and Hodzic, D. (2009). Bringing KASH under the SUN: the many

faces of nucleo-cytoskeletal connections. J. Cell Biol. 186, 461–472.

Reiner, O., Carrozzo, R., Shen, Y., Wehnert, M., Faustinella, F., Dobyns, W.B.,

Caskey, C.T., and Ledbetter, D.H. (1993). Isolation of a Miller-Dieker lissence-

phaly gene containing G protein beta-subunit-like repeats. Nature 364,

717–721.

Reinsch, S., and Gonczy, P. (1998). Mechanisms of nuclear positioning. J. Cell

Sci. 111, 2283–2295.

1388 Cell 152, March 14, 2013 ª2013 Elsevier Inc.

Roux, K.J., Crisp,M.L., Liu, Q., Kim, D., Kozlov, S., Stewart, C.L., andBurke, B.

(2009). Nesprin 4 is an outer nuclear membrane protein that can induce kine-

sin-mediated cell polarization. Proc. Natl. Acad. Sci. USA 106, 2194–2199.

Salpingidou, G., Smertenko, A., Hausmanowa-Petrucewicz, I., Hussey, P.J.,

and Hutchison, C.J. (2007). A novel role for the nuclear membrane protein

emerin in association of the centrosome to the outer nuclear membrane.

J. Cell Biol. 178, 897–904.

Schaar, B.T., and McConnell, S.K. (2005). Cytoskeletal coordination during

neuronal migration. Proc. Natl. Acad. Sci. USA 102, 13652–13657.

Schenk, J., Wilsch-Brauninger, M., Calegari, F., and Huttner, W.B. (2009).

Myosin II is required for interkinetic nuclear migration of neural progenitors.

Proc. Natl. Acad. Sci. USA 106, 16487–16492.

Schmoranzer, J., Fawcett, J.P., Segura, M., Tan, S., Vallee, R.B., Pawson, T.,

and Gundersen, G.G. (2009). Par3 and dynein associate to regulate local

microtubule dynamics and centrosome orientation during migration. Curr.

Biol. 19, 1065–1074.

Schreiber, K.H., and Kennedy, B.K. (2013). When lamins go bad: Nucleur

structure and disease. Cell 152, this issue, 1365–1375.

Shu, T., Ayala, R., Nguyen, M.D., Xie, Z., Gleeson, J.G., and Tsai, L.H. (2004).

Ndel1 operates in a common pathway with LIS1 and cytoplasmic dynein to

regulate cortical neuronal positioning. Neuron 44, 263–277.

Solecki, D.J., Model, L., Gaetz, J., Kapoor, T.M., and Hatten, M.E. (2004).

Par6alpha signaling controls glial-guided neuronal migration. Nat. Neurosci.

7, 1195–1203.

Solecki, D.J., Trivedi, N., Govek, E.E., Kerekes, R.A., Gleason, S.S., and Hat-

ten, M.E. (2009). Myosin II motors and F-actin dynamics drive the coordinated

movement of the centrosome and soma during CNS glial-guided neuronal

migration. Neuron 63, 63–80.

Sosa, B.A., Rothballer, A., Kutay, U., and Schwartz, T.U. (2012). LINC

complexes form by binding of three KASH peptides to domain interfaces of

trimeric SUN proteins. Cell 149, 1035–1047.

Splinter, D., Tanenbaum, M.E., Lindqvist, A., Jaarsma, D., Flotho, A., Yu, K.L.,

Grigoriev, I., Engelsma, D., Haasdijk, E.D., Keijzer, N., et al. (2010). Bicaudal

D2, dynein, and kinesin-1 associate with nuclear pore complexes and regulate

centrosome and nuclear positioning during mitotic entry. PLoS Biol. 8,

e1000350.

Starr, D.A., and Han, M. (2002). Role of ANC-1 in tethering nuclei to the actin

cytoskeleton. Science 298, 406–409.

Starr, D.A., and Fridolfsson, H.N. (2010). Interactions between nuclei and the

cytoskeleton are mediated by SUN-KASH nuclear-envelope bridges. Annu.

Rev. Cell Dev. Biol. 26, 421–444.

Sullivan, T., Escalante-Alcalde, D., Bhatt, H., Anver, M., Bhat, N., Nagashima,

K., Stewart, C.L., and Burke, B. (1999). Loss of A-type lamin expression

compromises nuclear envelope integrity leading to muscular dystrophy.

J. Cell Biol. 147, 913–920.

Swan, A., Nguyen, T., and Suter, B. (1999). Drosophila Lissencephaly-1 func-

tions with Bic-D and dynein in oocyte determination and nuclear positioning.

Nat. Cell Biol. 1, 444–449.

Tanabe, L.M., Kim, C.E., Alagem, N., andDauer,W.T. (2009). Primary dystonia:

molecules and mechanisms. Nat. Rev. Neurol. 5, 598–609.

Tassin, A.M., Maro, B., and Bornens, M. (1985). Fate ofmicrotubule-organizing

centers during myogenesis in vitro. J. Cell Biol. 100, 35–46.

Taverna, E., and Huttner, W.B. (2010). Neural progenitor nuclei IN motion.

Neuron 67, 906–914.

Tomlinson, A., and Ready, D.F. (1986). Sevenless: a cell-specific homeotic

mutation of the Drosophila eye. Science 231, 400–402.

Tran, P.T., Marsh, L., Doye, V., Inoue, S., and Chang, F. (2001). A mechanism

for nuclear positioning in fission yeast based on microtubule pushing. J. Cell

Biol. 153, 397–411.

Tsai, L.H., and Gleeson, J.G. (2005). Nucleokinesis in neuronal migration.

Neuron 46, 383–388.

Tsai, F.C., and Meyer, T. (2012). Ca2+ pulses control local cycles of lamellipo-

dia retraction and adhesion along the front of migrating cells. Curr. Biol. 22,

837–842.

Tsai, J.W., Chen, Y., Kriegstein, A.R., and Vallee, R.B. (2005). LIS1 RNA inter-

ference blocks neural stem cell division, morphogenesis, and motility at

multiple stages. J. Cell Biol. 170, 935–945.

Tsai, J.W., Bremner, K.H., and Vallee, R.B. (2007). Dual subcellular roles for

LIS1 and dynein in radial neuronal migration in live brain tissue. Nat. Neurosci.

10, 970–979.

Tsai, J.W., Lian, W.N., Kemal, S., Kriegstein, A.R., and Vallee, R.B. (2010).

Kinesin 3 and cytoplasmic dynein mediate interkinetic nuclear migration in

neural stem cells. Nat. Neurosci. 13, 1463–1471.

Tsujikawa, M., Omori, Y., Biyanwila, J., and Malicki, J. (2007). Mechanism of

positioning the cell nucleus in vertebrate photoreceptors. Proc. Natl. Acad.

Sci. USA 104, 14819–14824.

Warren, D.T., Tajsic, T., Mellad, J.A., Searles, R., Zhang, Q., and Shanahan,

C.M. (2010). Novel nuclear nesprin-2 variants tether active extracellular

signal-regulated MAPK1 and MAPK2 at promyelocytic leukemia protein

nuclear bodies and act to regulate smooth muscle cell proliferation. J. Biol.

Chem. 285, 1311–1320.

Welte, M.A. (2004). Bidirectional transport along microtubules. Curr. Biol. 14,

R525–R537.

Wilhelmsen, K., Litjens, S.H., Kuikman, I., Tshimbalanga, N., Janssen, H., van

denBout, I., Raymond, K., and Sonnenberg, A. (2005). Nesprin-3, a novel outer

nuclear membrane protein, associates with the cytoskeletal linker protein

plectin. J. Cell Biol. 171, 799–810.

Wilkie, G.S., Korfali, N., Swanson, S.K., Malik, P., Srsen, V., Batrakou, D.G., de

las Heras, J., Zuleger, N., Kerr, A.R., Florens, L., et al. (2011). Several novel

nuclear envelope transmembrane proteins identified in skeletal muscle have

cytoskeletal associations. Mol. Cell. Proteomics, 10, M110.003129.

Xie, Z., Moy, L.Y., Sanada, K., Zhou, Y., Buchman, J.J., and Tsai, L.H. (2007).

Cep120 and TACCs control interkinetic nuclear migration and the neural

progenitor pool. Neuron 56, 79–93.

Yu, J., Lei, K., Zhou,M., Craft, C.M., Xu, G., Xu, T., Zhuang, Y., Xu, R., and Han,

M. (2011). KASH protein Syne-2/Nesprin-2 and SUN proteins SUN1/2 mediate

nuclear migration during mammalian retinal development. Hum. Mol. Genet.

20, 1061–1073.

Zhang, Q., Skepper, J.N., Yang, F., Davies, J.D., Hegyi, L., Roberts, R.G.,

Weissberg, P.L., Ellis, J.A., and Shanahan, C.M. (2001). Nesprins: a novel

family of spectrin-repeat-containing proteins that localize to the nuclear

membrane in multiple tissues. J. Cell Sci. 114, 4485–4498.

Zhang, Q., Ragnauth, C.D., Skepper, J.N., Worth, N.F., Warren, D.T., Roberts,

R.G., Weissberg, P.L., Ellis, J.A., and Shanahan, C.M. (2005). Nesprin-2 is

a multi-isomeric protein that binds lamin and emerin at the nuclear envelope

and forms a subcellular network in skeletal muscle. J. Cell Sci. 118, 673–687.

Zhang, Q., Bethmann, C., Worth, N.F., Davies, J.D., Wasner, C., Feuer, A.,

Ragnauth, C.D., Yi, Q., Mellad, J.A., Warren, D.T., et al. (2007a). Nesprin-1

and -2 are involved in the pathogenesis of Emery Dreifuss muscular dystrophy

and are critical for nuclear envelope integrity. Hum. Mol. Genet. 16, 2816–

2833.

Zhang, X., Xu, R., Zhu, B., Yang, X., Ding, X., Duan, S., Xu, T., Zhuang, Y., and

Han, M. (2007b). Syne-1 and Syne-2 play crucial roles in myonuclear

anchorage and motor neuron innervation. Development 134, 901–908.

Zhang, X., Lei, K., Yuan, X., Wu, X., Zhuang, Y., Xu, T., Xu, R., and Han, M.

(2009). SUN1/2 and Syne/Nesprin-1/2 complexes connect centrosome to

the nucleus during neurogenesis and neuronal migration in mice. Neuron 64,

173–187.

Zhao, T., Graham, O.S., Raposo, A., and St Johnston, D. (2012). Growing

microtubules push the oocyte nucleus to polarize the Drosophila dorsal-

ventral axis. Science 336, 999–1003.

Zhou, X., Graumann, K., Evans, D.E., and Meier, I. (2012a). Novel plant SUN-

KASH bridges are involved in RanGAP anchoring and nuclear shape determi-

nation. J. Cell Biol. 196, 203–211.

Zhou, Z., Du, X., Cai, Z., Song, X., Zhang, H., Mizuno, T., Suzuki, E., Yee, M.R.,

Berezov, A., Murali, R., et al. (2012b). Structure of Sad1-UNC84 homology

(SUN) domain defines features of molecular bridge in nuclear envelope.

J. Biol. Chem. 287, 5317–5326.

Cell 152, March 14, 2013 ª2013 Elsevier Inc. 1389

High chromatinmobility

Pre-RCassembly

Timing decisionpoint (TDP)

Timing decision point (TDP)

Limited pool

Cdc45 loading stimulated byFkh1/2 and inhibited by

Rif1, Taz1, Yku70 orRpd3 deacetylation

Initiation of Cdc45 loadedorigins and removal of Rif1/Taz1/Yku70/Rpd3-

mediated repression

Terminating replisomes release Cdc45 to activate

un�red origins

Prereplicativecomplex

Cdc45

Fkh1/2

Rif1, Taz1, Yku70 or Rpd3

Replisome

Replication

Early origin

Late origin

Early replication Late replication

Origin decisionpoint (ODP)

Restrictionpoint

Earlyreplication

Latereplication

Timing determinants lost

Domains repositioned at the TDP remain anchored for the remainder of interphaseChromosomecondensation

M

MAMMALS

YEAST

G1 EARLY S LATE S G2 M

MAMMALS

Constitutively earlyGC rich

Gene richHigh nuclease sensitivity

Developmentally regulatedIntermediate sequence composition

Low nuclease sensitivityDynamic chromatin marks

Correlated with transcription

Constitutively lateAT rich

Gene poorLow nuclease sensitivity

Replicationfoci

Replicationfoci

1 µm

5 µm

YEAST

REPLICATION DOMAINS

Differentiation

Early replicating

Late replicating

Replication

Prereplicative complexReplisomesUnreplicated DNAReplicated DNA

N U C L E A R I N T E R I O R

N U C L E A R L A M I N A

x x xx

Interdomain interaction cis boundary

SnapShot: Replication TimingBenjamin D. Pope,1 Oscar M. Aparicio,2 and David M. Gilbert,1

1Department of Biological Science, Florida State University, Tallahassee, FL 32306, USA2Molecular and Computational Biology Program, University of Southern California, Los Angeles, CA 90089, USA

See online version for legend and references.1390 Cell 152, March 14, 2013 ©2013 Elsevier Inc. DOI http://dx.doi.org/10.1016/j.cell.2013.02.038

species Genome size s-phase length

Replication fork rate

number of potential origins

number of replicons per s

number of foci per s

Replicons/focus

number of domains Domain size

Yeast 12-14 Mb <1 hr 1-2 kb/min 500-1,000 100-200 15-30 4-7 40-70? ≤250 kb

Mammals 2-4 Gb 8-10 hr 1-2 kb/min >250,000 25,000-50,000 5,000-10,000 6-20 4,000-5,000 400-800 kb

1390.e1 Cell 152, March 14, 2013 ©2013 Elsevier Inc. DOI http://dx.doi.org/10.1016/j.cell.2013.02.038

SnapShot: Replication TimingBenjamin D. Pope,1 Oscar M. Aparicio,2 and David M. Gilbert,1

1Department of Biological Science, Florida State University, Tallahassee, FL 32306, USA2Molecular and Computational Biology Program, University of Southern California, Los Angeles, CA 90089, USA

All organisms use similar principles to duplicate DNA at replication forks (Yao and O’Donnell, 2010). However, eukaryotic cells contain large chromosomes with hundreds to thousands of replication origins and complex, heterogeneous chromatin. Conserved cell-cycle and checkpoint mechanisms ensure one complete round of replication (Labib, 2010), and additional mechanisms coordinate initiation at the many replicons (regions replicated from a single origin) in space and time.

A temporal order to genome replication balances replication with limiting cellular resources such as initiation factors and nucleotide pools (Aparicio, 2013; Rhind and Gil-bert, 2013). Chromatin features regulate replication timing by controlling the access of initiation factors to replication origins. In fact, replication timing is one of the few cellular functions that are clearly regulated at the level of large-scale/long-range chromatin folding. In mammals, this temporal order is regulated during development and is linked to transcriptional regulation (Nordman and Orr-Weaver, 2012; Rhind and Gilbert, 2013).

Replication foci in budding and fission yeast are nuclear sites of active replication that can be visualized by fluorescently tagged replication fork proteins or nucleotide ana-logs (Kitamura et al., 2006; Meister et al., 2007). Foci (yellow) in the displayed image (originally published in Trends Cell Biol., December 2001) exhibit a pattern that is typical of early S in budding yeast. Chromatin is counterstained (red). These foci are mobile and frequently fuse with other foci or split to form new foci, making precise measurements of their characteristics difficult. The ratio of replication foci to active replication forks indicates that several closely spaced replicons are active simultaneously within each focus, although the organization can only be modeled imprecisely at this point.

Replication foci in mammals are more numerous than in yeast and, relative to the size of the nucleus, less mobile (Maya-Mendoza et al., 2010; Rhind and Gilbert, 2013). An average focus replicates ~1 Mb of DNA in 45–60 min. In the displayed image (originally published in Genome Res., June 2010), cells were dual labeled in successive pulses to visualize both early (green) and late (red) foci simultaneously, highlighting the spatial compartmentalization of chromatin replicated at different times during S phase. Foci labeled in one cell cycle are stable in appearance for many cycles, indicating that they are structural units and most likely the cytological equivalents of replication domains measured by molecular genomics methods. Like foci, replication domains are chromosomal units that are replicated coordinately by synchronously firing clusters of replicons, which also approach megabase size in mammals. At least half of replication domains are regulated to replicate at different times in different tissues (Nordman and Orr-Weaver, 2012; Rhind and Gilbert, 2013). Replicons associated with the nuclear interior are replicated early in S phase, whereas those adjacent to the lamina replicate later. Domains that switch replication timing during differentiation move between subnuclear compartments, as indicated both by physical position and by changes in interdomain chromatin interactions (Takebayashi et al., 2012). Replication domain boundaries may insulate chromatin types from each other, facilitating the differential replication timing of adjacent chromatin domains.

In yeast, the extent to which clusters of origins form replication domains is controversial. At least four large (~250 kb) regions in budding yeast have distinctly late timing (McCune et al., 2008). In addition, each chromosome could be considered to contain several domains—the centromere, each arm, and the telomeres—possibly with a few addi-tional subdomains. The range in number of potential origins reflects differences in budding (~500) and fission yeast (~1,000), wherein origin efficiencies are generally higher in budding yeast, probably resulting in similar replicon numbers.

Exiting mitosis, the chromatin of mammalian cells is highly mobile and lacks determinants for replication timing. Within 1–2 hr, cells reach the timing decision point (TDP), when replication domains/foci anchor in their respective subnuclear positions for the remainder of interphase and simultaneously acquire the ability to dictate a replication-timing program (Rhind and Gilbert, 2013). Establishing this timing program occurs upstream of specifying which sites will be used for initiation (origin decision point) and the activation of S phase Cdk activity (restriction point). The replication-timing program is executed during S phase through the firing of several sequential groups of internally localized foci (green), followed by replication at the nuclear and nucleolar periphery (red) and, finally, a few sites of internally localized heterochromatin (not shown). Chromatin in G2 phase lacks determinants for replication timing, suggesting that such determinants are lost during replication.

Coincident with the TDP (between mitosis and start in yeast), Fkh1/2 and Rif1/Taz1/Yku70/Rpd3 organize early and late-replicating chromatin, respectively. Fkh1/2 bind con-sensus DNA elements near early replicating origins and through interaction with the origin recognition complex (ORC) and/or through Fkh1/2 dimerization bring the origins into proximity (Aparicio, 2013). Fkh1/2 stimulates Cdc45 recruitment and loading onto prereplication complexes (pre-RCs) at early origins during G1 phase, facilitating early initiation. Rif1/Taz1/Yku70 position and tether telomeres and some internal sequences to the nuclear periphery, whereas Rpd3 modifies chromatin, isolating late-replicating chromatin from Cdc45 (and other initiation factors). The incorporation of the limited pool of Cdc45 (and other factors) into early replisomes delays firing of Rif1/Taz1/Yku70/Rpd3-repressed origins until termination of early replicons recycles Cdc45.

RefeRences

Aparicio, O.M. (2013). Location, location, location: it’s all in the timing for replication origins. Genes Dev. 27, 117–128.

Kitamura, E., Blow, J.J., and Tanaka, T.U. (2006). Live-cell imaging reveals replication of individual replicons in eukaryotic replication factories. Cell 125, 1297–1308.

Labib, K. (2010). How do Cdc7 and cyclin-dependent kinases trigger the initiation of chromosome replication in eukaryotic cells? Genes Dev. 24, 1208–1219.

Maya-Mendoza, A., Olivares-Chauvet, P., Shaw, A., and Jackson, D.A. (2010). S phase progression in human cells is dictated by the genetic continuity of DNA foci. PLoS Genet. 6, e1000900.

McCune, H.J., Danielson, L.S., Alvino, G.M., Collingwood, D., Delrow, J.J., Fangman, W.L., Brewer, B.J., and Raghuraman, M.K. (2008). The temporal program of chromosome rep-lication: genomewide replication in clb5Delta Saccharomyces cerevisiae. Genetics 180, 1833–1847.

Meister, P., Taddei, A., Ponti, A., Baldacci, G., and Gasser, S.M. (2007). Replication foci dynamics: replication patterns are modulated by S-phase checkpoint kinases in fission yeast. EMBO J. 26, 1315–1326.

Nordman, J., and Orr-Weaver, T.L. (2012). Regulation of DNA replication during development. Development 139, 455–464.

Rhind, N., and Gilbert, D.M. (2013). Replication timing. In DNA Replication and Human Disease, S.D. Bell, M. Mechali, and M.L. DePamphilis, eds. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). http://dx.doi.org/10.1101/cshperspect.a010132.

Takebayashi, S., Dileep, V., Ryba, T., Dennis, J.H., and Gilbert, D.M. (2012). Chromatin-interaction compartment switch at developmentally regulated chromosomal domains reveals an unusual principle of chromatin folding. Proc. Natl. Acad. Sci. USA 109, 12574–12579.

Yao, N.Y., and M. O’Donnell. (2010). SnapShot: The replisome. Cell 141, 1088–e1081.