Available online at www.sciencedirect.com

Origin and evolution of X inactivation

Joost Gribnau and J Anton Grootegoed

Evolution of the mammalian sex heavily impacts chromosome pair (Figure 1a) [3]. It is thought that the

on the expression of X-encoded , both in marsupials and region around Sry gained male beneficial genes and other

placental mammals. The loss of genes from the modifications such as inversions, step by step limiting the

forced a two-fold upregulation of dose sensitive X-linked options for meiotic recombination between the evolving

homologues. As a corollary, female cells would experience a heterologous sex chromosomes. As a result, the male

lethal dose of X-linked genes, if this upregulation was not specific region of Y (MSY), clonally inherited from father

counteracted by evolution of inactivation (XCI) to son, never meets a meiotic pairing partner. The X

that allows for only one active X chromosome per diploid chromosome, however, spends two-third of its time in

genome. Marsupials rely on imprinted XCI, which inactivates XX females, where it pairs and recombines in meiotic

always the paternally inherited X chromosome. In placental prophase of oogenesis. Further to the advantage of the

mammals, random XCI (rXCI) is the predominant form, X chromosome, X-linked genes are hemizygous in males,

inactivating either the maternal or paternal X. In this review, we leading to rapid fixation of mutations associated with

discuss recent new insights in the regulation of XCI. Based on positive natural and sexual selection. In anthropomorphic

these findings, we propose an X inactivation center (Xic), terms, an arms race is going on between X and Y, resulting

composed of a cis-Xic and trans-Xic that encompass all in regression of the Y and enrichment of the content of

elements and factors acting to control rXCI either in cis or in the X. Many of the events shaping the heterologous sex

trans. We also highlight that XCI may have evolved from a very chromosomes have occurred within a relatively short time

small nucleation site on the X chromosome in the vicinity of the span after formation of the proto-X and proto-Y, before

Sox3 gene. Finally, we discuss the possible evolutionary road radiation of the placental mammals [4–6]. Originating from

maps that resulted in imprinted XCI and rXCI as observed in the spectacular evolutionary history of these particular

present day mammals. chromosomes and its consequences, the X and Y chromo-

somes of present-day placental mammals undergo marked

Address dynamic changes of their activities in development and

Department of Reproduction and Development, Erasmus MC, University

gametogenesis. These activity changes have a biological

Medical Center, Rotterdam, The Netherlands

function to support growth and fertility of both females and

males. In the present review, we aim to outline the overall

Corresponding author: Gribnau, Joost ([email protected])

scheme of events and mechanisms, although we focus on X

chromosome inactivation (XCI) in female somatic cells.

Current Opinion in Cell Biology 2012, 24:397–404

Sex chromosomes and the need for gene

This review comes from a themed issue on

Nucleus and gene expression dosage compensation

Edited by Asifa Akhtar and Karla Neugebauer The human X chromosome carries more than 1000 genes

[7]. By contrast, only around 100 single-copy and multi-

Available online 14th March 2012

copy genes are found on the human Y chromosome,

0955-0674/$ – see front matter where the MSY carries 78 genes encoding 27 different

# 2012 Elsevier Ltd. All rights reserved.

proteins [8]. Hence, in diploid male somatic cells, there is

an X: gene dosage imbalance for most genes on

DOI 10.1016/j.ceb.2012.02.004

the X. Problems resulting from this imbalance can be

prevented by a two-fold transcriptional upregulation of

the X chromosome. Indeed, this has been observed for

Introduction mouse and human, by comparing micro-array expression

Meiotic recombination keeps the organized data sets [9–11]. Another study, using an RNA-sequen-

in homologous pairs, stabilizing the diploid genome of cing approach, challenged these findings [12], but the

 

mammalian species. For one pair of autosomes, this controversy is now explained [13 ,14 ]. When genes with

advantage of the diploid state was partly lost, when this low transcriptional activity are excluded from the analysis,

pair of autosomes started to become the present mamma- it is observed that the moderately to highly expressed X-

 

lian X and Y chromosomes around 160 million years ago, chromosomal genes are two-fold upregulated [13 ,14 ].

shortly before separation of the metatherians (marsupials) In addition to excluding biological noise from leaky gene

and eutherians (placental mammals) [1,2]. The initial expression, genes expressed at stochastically fluctuating

event probably has been a mutational change of one allele low levels may not require dosage compensation. How-

of the Sox3 gene, resulting in the male sex-determining ever, genes expressed at higher levels will include dose

gene Sry (sex-determining region Y) on the proto-Y of this sensitive genes. From this, we suggest that evolutionary

www.sciencedirect.com Current Opinion in Cell Biology 2012, 24:397–404

398 Nucleus and gene expression

Figure 1

genes in female mammals is prevented by a second

compensatory mechanism, called XCI, which leads to

(a) degeneration of the Y inactivation of one of the two X chromosomes, resetting

the global X:autosome gene expression ratio at 1:1 in

recombination

female cells (Figure 1c). Looking at the end result, XCI

block

can be viewed as a mechanism which equalizes X-chro-

Sox3 Sox3 Sox3 Sry male mosomal gene dosage between males and females. beneficial

genes

X chromosome inactivation: cis and trans proto mechanisms

autosomes X Y X Y X Y

In the female mouse embryo, an imprinted form of XCI

160 million years ago now

(iXCI) is initiated very early during pre-implantation

development, around the 4–8 cell stage, always targeting

(b) two-fold up-regulation in male

the X inherited from the father [16,17]. Following iXCI,

2x 2x this paternal X (Xp) remains inactive in the extra-embryo-

2x nic tissues, but is reactivated in the developing inner cell

mass that gives rise to the embryo proper [18]. This

proto reactivation asks for rapid intervention, which comes from

autosomes X Y X Y X Y

a strong wave of random XCI (rXCI), targeting either the

maternal X (Xm) or Xp, just after implantation. The

(c) X inactivation in female fascinating history of the discovery of rXCI, spearheaded

by Susumu Ohno and Mary Lyon around half a century

Xist RNA

ago, is highlighted in several recent reviews [19–21].

2x 2x

Regarding the mechanism of rXCI, genetic studies per-

formed in the 1980s, involving X-to-autosomal transloca-

proto tions and X truncations, indicated a region of 3

autosomes X X X X Xa Xi

megabases on the X, called the X inactivation center

Current Opinion in Cell Biology

(Xic), instrumental in the initiation of rXCI [22]. Sub-

sequent studies focusing on this region revealed a non-

Evolution of mammalian sex chromosomes. (a) Degeneration of the Y

coding gene, Xist in mouse and XIST in human, as the key

chromosome was triggered by changes on the Y, including the origin of

player in the XCI process [23–25]. Xist transcription is

Sry and the inclusion of male beneficial genes in a non-recombining

upregulated on the future inactive X, and the spliced and

region. (b) The loss of Y-encoded genes was compensated by a two-

fold upregulation of expression of dose sensitive homologous X- poly-adenylated non-coding RNA molecules spread in cis,

encoded genes. (c) A two-fold upregulation of X-encoded genes would thereby recruiting remodelling complexes that

be lethal to female cells, and this has driven the evolution of XCI that

render the X an inactive chromatin domain (reviewed in

silences one X chromosome in every female cell.

[26]). From what is known for mouse, Xist activation is

counteracted by the non-coding Tsix gene, which fully

overlaps with Xist but is transcribed anti-sense to Xist [27].

selection has employed mechanisms to keep the X:auto- To repress Xist transcription, transcription of Tsix needs to

some gene expression level at a ratio near 1:1, concomi- proceed through sequences overlapping with the Xist

tant with the evolution of the X and Y chromosomes, promoter [28]. This points to a transcriptional interfer-

leading to a global two-fold transcriptional upregulation ence mechanism, but the repression may also involve

of the X chromosome (Figure 1b). If the mechanism RNA-mediated recruitment of de novo methyltransferase

leading to this two-fold upregulation has become an DNMT3A [29,30]. With this basic machinery in place,

inherent and sex independent property of the X chromo- the hunt was on to find elements, genes and factors able to

some, female cells encounter a 2:1 X:autosome gene exert control over Xist and Tsix, in cis or in trans.

expression ratio. In some species, this is tolerated, as

for instance in the red flour beetle, Tribolium castaneum, Genuine promoter and enhancer sequences control Xist

where increased expression of both X chromosomes is and Tsix transcription, but flanking non-coding genes also

observed in XX female cells [15]. In mammals, the play an important role in activation of both genes.

developing oocyte contains two active X chromosomes, Located upstream of Tsix, the genes Xite and Tsx posi-

and a certain level of tolerance might also be found at tively regulate Tsix expression (Figure 2) [31,32].

some later steps of development and in specific tissues Chromosome conformation capture (3C) studies

and cell types. However, a general 2:1 X:autosome gene suggested that Xite and Tsx interact with Tsix within an

expression ratio is not tolerated by any of the mammalian active chromatin compartment or hub (ACH) [33]. Like-

species for which this was studied thus far. Overexpres- wise, the non-coding genes Jpx and Ftx, located upstream

sion of ‘male-dosage-compensated’ X-chromosomal of Xist, act as positive regulators of Xist (Figure 2) [34,35].

Current Opinion in Cell Biology 2012, 24:397–404 www.sciencedirect.com

X chromosome inactivation Gribnau and Grootegoed 399

Figure 2

(a) cis-Xic (b) trans-Xic

RNF12 + + +

boundary mouse Xist Jpx Ftx Cnbp2 Xpr Rnf12 Xist Jpx FtxCnbp2 Xpr Rnf12

cen tel

Tsx Xite Tsix Tsx Xite

Tsix – – – OCT4 OCT4 + + + OCT4 SOX2 SOX2 SOX2 NANOG NANOG Tsix ACH Xist ACH OCT4 CCTCF SOX2 YYY1 KLF4 REX1R cis-Xic cMYC KLF4 trans-Xic trans-Xic

Current Opinion in Cell Biology

The X inactivation center: cis and trans. (a) The cis-acting X inactivation center (cis-Xic) encompasses all cis-acting elements regulating Xist and Tsix.

The cis-Xic includes non-coding genes located upstream of Xist and Tsix, which are involved in setting up two active chromatin hub (ACH) structures.

(b) The trans-acting Xic includes all X-linked elements and genes that act in trans to activate Xist. Parts of the trans-Xic most probably are found at

more regions of the X, and the trans-Xic contains all XCI-activator genes including Rnf12, which seems to be a more potent XCI-activator (thick line)

than Jpx and Xpr (thin line) and the putative XCI activator Ftx (dashed line). Also shown are the binding sites, and the mode of action, of autosomally

encoded inhibitors of XCI.

This would be in agreement with Ftx, Jpx and Xist setting expected to be highly dose-dependent. With activators

up an Xist ACH, neighbouring the Tsix ACH (Figure 2a). being encoded by X-linked genes, in contrast to inhibitors

of XCI that are autosomally encoded, the cell would obtain

From the above it follows that transcription of the key a build-in stop mechanism to prevent inactivation of all X

genes Xist and Tsix additionally depends on an interplay chromosomes. Embryonic rXCI is forcefully initiated in

in cis with upstream non-coding genes, within two neigh- female cells only, which have a two-fold higher gene dose of

bouring chromatin domains. The Xist ACH and Tsix ACH X-encoded activator(s), allowing the XCI-activator activity

are separated by a boundary element marked by a CTCF to reach a threshold level required to overcome repression

binding site located just downstream of Xist. Indeed, by the XCI-inhibitors. Although allelic properties of the

removal of this CTCF binding site blocks Xist induction two X chromosomes will play a role, the initiation of rXCI



[36 ], possibly because both ACHs have accidently on either one of the two X chromosomes in principle is a

merged, although this needs further investigation. Regu- stochastic event. Once rXCI is silencing one of the two X

lation of Xist and Tsix transcription within the respective chromosomes, the ensuing rapid downregulation of XCI-

ACHs may involve direct inter-genic promoter and activator gene(s) in cis should prevent initiation of rXCI on

enhancer interactions, but the chromatin environment the second X. Mutual exclusive initiation of rXCI may be

may facilitate recruitment of factors implicated in tran- facilitated by spatial interaction between the two X

scription initiation also by a mass action mechanism. It is chromosomes, involving Tsix, Xite and Xpr (X pairing)

anticipated that the two ACHs are not engaged in com- sequences [38–40], although a causal role for a transient

petition for proper spatial folding, but rather act as rela- direct interaction still needs to be established.

tively independent domains to promote transcription of

either Xist or Tsix. We suggest that the two chromatin In the developing embryo proper, rXCI does not start

domains represent the cis acting machinery, determining before the onset of cell differentiation. Hence, it makes

the probability that XCI is initiated. This probability has biological sense that, in addition to the ubiquitously

been investigated in particular for rXCI, in relation to the expressed proteins CTCF and YY1, the most important

identification of trans-acting factors. XCI inhibitors identified thus far happen to be the key

pluripotency and reprogramming factors OCT4, SOX2,



Trans-acting factors are probably taking part in control of NANOG, REX1, KLF4 and MYC ([41–43,44 ];

rXCI, to create regulatory mechanisms for counting and reviewed by [45]). These factors are recruited to different

initiation. Trans-acting activators would promote XCI regions involved in repression of Xist, or activation of Tsix

through activation of Xist or repression of Tsix, being or Xite (Figure 2b). By contrast, only few XCI-activators

counteracted by trans-acting inhibitors [37]. In this model, have been identified. X-encoded RNF12 appears to be a

the activities of both the activators and the inhibitors are very potent activator of Xist, although the mechanism of

www.sciencedirect.com Current Opinion in Cell Biology 2012, 24:397–404

400 Nucleus and gene expression

action of this E3 ubiquitin ligase needs to be resolved in at a larger distance, so that the trans-XIC may include



more detail [46,47 ]. Rnf12 is located 500 kb upstream of various regions of the X chromosome.

Xist, and this close proximity to the source of Xist RNA

production, in combination with a short half-life of Evolution of XCI

RNF12, is thought to guarantee a fast and stringent stop Upon formation of the mammalian proto-X and proto-Y by

mechanism of XCI-activator production once rXCI has mutational change of Sox3 giving rise to the Sry gene with

started. Analysis of Rnf12+/À and Rnf12À/À female ES male-specific expression, the early non-recombining

cells showed reduced initiation of rXCI upon differen- region must have included Sox3 and probably some neigh-

tiation, although this effect seems to be influenced by the bouring genes on the brand-new X chromosome. At a high

 

genetic background [47 ,48 ]. The Rnf12 knockout dose, SOX3 is still capable of mimicking the action of SRY

mice indicated that RNF12 is essential also for iXCI in triggering transcription of Sox9, the immediate down-



[48 ]. It is concluded that RNF12 is essential for both stream gene that is required for testis formation [3].

iXCI and rXCI, but that more activators of XCI are Hence, Sox3 is a dose sensitive gene, which may have

present, such as the candidate XCI-activators Jpx and experienced transcriptional upregulation from the proto-

Ftx. Knockout and rescue studies have implicated a role X, leading to an ensuing need for dosage compensation

for Jpx RNA in trans, but also uncovered important cis specifically in female cells. Perhaps, this represents the

effects [35]. So far, regulation of XCI by Ftx was studied evolutionary origin of XCI, where the genes that are

only in male cells, which did not allow discrimination mechanistically involved in XCI have evolved in a small

between cis and trans effects. Rnf12 transgenic male ES region around Sox3 (Figure 1c). The non-coding Xist gene

cell lines showed a robust XCI phenotype of the single X originates, at least partly, from the Lnx3 gene [50], encod-

present in these cells, reinforcing the action of RNF12 in ing a member of an old family of E3 ubiquitin ligases that

trans. However, male cells carrying Jpx, Ftx, Xpr or Tsx dates far back to a metazoan ancestor [51]. The transcribed

transgenes did not show ectopic induction of XCI [46]. It and processed Xist RNA gained a completely new func-

is quite possible, therefore, that Jpx, Ftx, Xpr and Tsx tion, not related to the function of Lnx3. Spreading of Xist

exert their effect on XCI mainly in cis [46]. RNA on the early X may have been very localized,

reminiscent of the action of the non-coding

Taken together, the present model points out that XCI, as Kcnq1ot1 and Air, which are transcribed from one parental

investigated mainly for rXCI, is regulated by trans-acting allele of imprinted autosomal loci and mediate silencing of

activators that modulate Xist and Tsix directly or through a region 300 kb around the Kcnq1ot1 and Air genes,

the two neighbouring cis-acting chromatin domains, the respectively [52,53]. XCI may have evolved from such a

ACHs, separated by a boundary element. Composition local silencing, step by step extending into larger regions

and structure of these domains is a key determinant in of the X chromosome during eutherian evolution, by

expression of Xist and Tsix. The Xic can be viewed as a recruiting more factors to participate in the process.

functional entity, but quite a large and complex entity,

composed of a cis-acting Xic (cis-Xic) and a trans-acting From the model discussed above, it is anticipated that

Xic (trans-Xic). Next to Xist and Tsix, the cis-Xic includes several genes involved in the control of XCI might have

all cis regulatory regions involved in regulation of Xist and been present in close proximity to Sox3 right from the start.

Tsix directly or by modulating the activity of the ACHs. Are Xist and Rnf12 located in close proximity to Sox3, in

Polymorphisms of the DNA incorporated into the ACHs placental mammals? When examined for human, mouse

might affect their distinctive activities, leading to skewed and cow, this is not the case (Figure 3a), but when we

XCI on a hybrid genetic background. Skewed XCI in searched for the chromosomal location of Lnx3 relative to

mouse is genetically linked to the X choosing element the locations of Sox3 and Rnf12 in birds, we found these

(Xce) . The Xce has been mapped to a 1.8 megabase three genes close together, on chromosomes 4A and 4 of

region including Xist [49], and may represent any nucleo- zebra finch and chicken, respectively (Figure 3a). The

tide change in the cis-Xic that extends into the Xist-ACH. syntenic region including these genes was on the autosome

The requirement for a feedback stop mechanism may pair that became the proto-X and proto-Y of mammals [2],

have promoted the evolutionary selection of genes encod- so that this finding would be in agreement with the present

ing XCI-activators in close vicinity to the cis-Xic, as found hypothesis of the evolutionary origin of XCI in the vicinity

for Rnf12, but this is probably not a rule of thumb. of Sox3. Possibly, orthologs of the oldest activators of XCI

Temporal spreading of global silencing from the Xic in placental mammals, other than Rnf12 and still unknown,

region is not strictly linear along the X [11], which would remain to be found close to Sox3 on chicken chromosome 4

allow the trans-Xic to be located at various sites on the X. (or on zebra finch chromosome 4A).

The abovementioned early genetic studies defining the

Xic indicated that the most important sequences control- In marsupials, who have maintained Lnx3 and hence lack

ling XCI are retained within a 3 megabase region, but Xist, imprinted inactivation of the paternally inherited X

this does not exclude that some genes contributing to a chromosome (iXp) is the only form of XCI, at all stages of

robust threshold of XCI-activator total activity are located development and in all tissues. However, the imprinted

Current Opinion in Cell Biology 2012, 24:397–404 www.sciencedirect.com

X chromosome inactivation Gribnau and Grootegoed 401

Figure 3

(a) Xist Rnf12 Sox3

human chr. X

Tsix 0,7mb 68mb

Sox3 Xist Rnf12

mouse chr. X 48mb Tsix 0.5mb

0,7mb Tsix

cow chr. X

Rnf12 Xist Sox3 55mb

Lnx3 Rnf12

zebra finch chr. 4A 0,1mb 1.2mb Sox3

Lnx3 Rnf12

chicken chr. 4 Sox3 1.7mb 0,1mb

(b) 200 mya 180 160 140 now

monotremes

iXp marsupials

iXp?

Sox3 -> Sry Lnx3 -> Xist placentals (Lnx3 imprinted?) rXCI ? iXp-Xist? iXCl ?

Current Opinion in Cell Biology

Evolution of X chromosome inactivation. (a) Sox3 is located at a quite large distance from Rnf12 and Xist on the X chromosome of mammals. By

contrast, Lnx3, the ancestral gene giving rise to Xist, is located in close proximity of Rnf12 and Sox3 on the X syntenic region of chromosomes 4A and 4

of zebra finch and chicken, respectively. (b) Timing of the evolution of Sry, Xist, imprinted inactivation of Xp and rXCI. The evolution of imprinted

inactivation of Xp may have occurred at least two times: (1) iXp in the marsupial lineage; and (2) iXp-Xist in the eutherian lineage, based on imprinted

Lnx3 or Xist. The evolution of rXCI might have been driven by relaxation of iXp-Xist, during placental radiation. The iXCI in mouse species might

represent a re-invention of iXp-Xist.

inactivation of Xp in mouse (and possibly in few or many incompletely synapsed X and Y chromosomes in male

other placental mammals), indicated herein with iXCI, meiotic prophase, forming the so-called XY body, inde-

is dependent on Xist, to obtain stable silencing [54,55]. pendent of Xist [57]. This silencing mechanism

It has been hypothesized that the Xp is inherited in a represents a specialized form of an evolutionary ancient

pre-inactive state, in marsupials and eutherians, as a mechanism leading to meiotic silencing of unsynapsed

consequence of meiotic sex chromosome inactivation chromatin regions (MSUC) [58,59]. However, MSUC

(MSCI) in spermatogenesis [16,56]. MSCI silences the may not target small unsynapsed regions effectively. In

www.sciencedirect.com Current Opinion in Cell Biology 2012, 24:397–404

402 Nucleus and gene expression

studies on mice carrying X-to-autosome translocations Can it be excluded that iXCI as observed in mouse in

resulting in meiotically unsynapsed regions of 40 mega- fact is a recent mechanism, turning up next to rXCI? It has

bases or even larger regions, MSUC of those regions was been described that rXCI in the mouse, where only few

found only in a proportion of meiotic prophase cells X-linked genes escape from silencing, is much more

[60,61]. As proposed above, XCI may have originated complete than rXCI in women [67]. Possibly, driven by

in a small non-recombining region of the early X chromo- a short generation time and strong natural selection,

some, and that region would have been too small to be mouse species may have reached an advanced level of

recognized by the MSUC machinery. In transgene stu- rXCI not seen in many other placental mammals. Inter-

dies, it was found that all sequences required for estingly, a Tsix gene that fully overlaps with Xist seems to

imprinted inactivation of autosomes with an Xist/Tsix be a unique feature of the mouse, as all other eutherian

transgene integration site are located within the trans- species examined have a Tsix gene that overlaps only with

gene, which is also in agreement with the notion that the last two exons of Xist [68]. Hence, Tsix might be less

MSCI is not required for iXCI [17]. Furthermore, detailed functional, or act through a different mechanism, in

studies on mouse have shown that many Xp-linked genes eutherian species other than mice. In addition to devel-

undergo zygotic activation [62,63]. Taken together, the oping an optimized rXCI mechanism, natural selection

weight of the evidence suggests that MSCI has no may have pushed mouse species to call on a mechanism

mechanistic impact on iXCI in mouse, and that imprinted for iXCI, either new or re-invented. Exploiting an

inactivation of Xp has not evolved from MSCI. advanced rXCI mechanism in combination with iXCI

covering dosage compensation in the very early embryo,

Parent-of-origin expression, by genomic imprinting, mouse species might be exceptionally well equipped for

originates from shortly before the separation of marsupials balanced growth of both male and female embryos shar-

and placental mammals [64]. Possibly, Lnx3 was an ing the same uterine horn during a short gestation period.

imprinted gene before its transformation to Xist

(Figure 3b), so that Xist may have maintained this geno- Conclusion and more questions

mic imprint during the beginning of the eutherian radi- The mammalian solution to the evolution of heterologous

ation. The first placental mammals may have shown sex chromosomes appears to involve the co-evolution of

imprinted X inactivation, probably targeting Xp rather two layers of gene dosage compensation: a two-fold upre-

than Xm to oppose accidental inactivation of the single X gulation of dose sensitive X-linked genes that were lost

from maternal origin in male embryos. It is anticipated from the Y chromosome, and X chromosome inactivation in

that several events have occurred within a relatively short female cells. Although important progress has been made

period, around the marsupial-eutherian split (Figure 3b). deciphering the mechanisms driving XCI in mouse, which

These events include mutation of Sox3 to proto-Sry, is the best studied placental mammal regarding these

mutation of Lnx3 to proto-Xist, and the onset of X inac- aspects of development, important factors and regulatory

tivation. It is not at all unlikely that two mechanisms for X pathways still need to be identified. Which factors have

inactivation came into existence: Xist independent contributed to evolution of XCI after the separation of

imprinted inactivation of Xp in the marsupial lineage, marsupials and eutherians? How many XCI activators are

and Xist dependent imprinted inactivation of Xp in the involved and how do they act on Xist or Tsix? What is the

eutherian lineage. In mouse, many autosomal genes precise nature and role of the proposed active chromatin

maintain regulatory genomic imprints in extra-embryonic hubs? How special is the mouse, employing both iXCI and

tissues, in regulation of placental growth. However, for a rXCI? What will we learn from studying XCI in other

proportion of these genes, the expression becomes more mammalian species, including human? The forthcoming

relaxed, independent from the original imprint, in the years of research will be challenging and exciting.

developing embryo proper [65]. The early eutherian

imprinted XCI mechanism may have undergone a similar Acknowledgements

We would like to thank all lab. members for stimulating discussions. J.G.

relaxation during evolution of the placental mammals,

was supported by NWO VICI and ERC starting grants.’

leading to reactivation of Xp in the embryo proper, forcing

selection towards the invention of rXCI.

References and recommended reading

Papers of particular interest, published within the period of review,

Recent studies indicate that XCI in pre-implantation

have been highlighted as:

human and rabbit embryos shows quite some differences,

 of special interest

with respect to timing and mechanism, from what is

 of outstanding interest



known for the mouse [66 ]. No strong evidence for iXCI

in human and rabbit has been obtained. Mice experience 1. Luo ZX, Yuan CX, Meng QJ, Ji Q: A Jurassic eutherian mammal

and divergence of marsupials and placentals. Nature 2011,

zygotic genome activation at an earlier time point in pre-

476:442-445.

implantation development than human and rabbit, and

2. Marshall Graves JA: Weird animal genomes and the evolution of

such an early genome activation might have been an

vertebrate sex and sex chromosomes. Annu Rev Genet 2008,



important factor driving the evolution of iXCI [66 ]. 42:565-586.

Current Opinion in Cell Biology 2012, 24:397–404 www.sciencedirect.com

X chromosome inactivation Gribnau and Grootegoed 403

3. Sutton E, Hughes J, White S, Sekido R, Tan J, Arboleda V, 21. Lee JT: Gracefully ageing at 50, X-chromosome inactivation

Rogers N, Knower K, Rowley L, Eyre H et al.: Identification of becomes a paradigm for RNA and chromatin control. Nat Rev

SOX3 as an XX male sex reversal gene in mice and humans. J Mol Cell Biol 2011, 12:815-826.

Clin Invest 2011, 121:328-341.

22. Brown CJ, Lafreniere RG, Powers VE, Sebastio G, Ballabio A,

4. Brawand D, Soumillon M, Necsulea A, Julien P, Csardi G, Pettigrew AL, Ledbetter DH, Levy E, Craig IW, Willard HF:

Harrigan P, Weier M, Liechti A, Aximu-Petri A, Kircher M et al.: The Localization of the X inactivation centre on the human X

evolution of gene expression levels in mammalian organs. chromosome in Xq13. Nature 1991, 349:82-84.

Nature 2011, 478:343-348.

23. Borsani G, Tonlorenzi R, Simmler MC, Dandolo L, Arnaud D,

5. Potrzebowski L, Vinckenbosch N, Kaessmann H: The emergence Capra V, Grompe M, Pizzuti A, Muzny D, Lawrence C et al.:

of new genes on the young therian X. Trends Genet 2009, 26:1-4. Characterization of a murine gene expressed from the inactive

X chromosome. Nature 1991, 351:325-329.

6. Zhang YE, Vibranovski MD, Landback P, Marais GA, Long M:

Chromosomal redistribution of male-biased genes in 24. Brockdorff N, Ashworth A, Kay GF, McCabe VM, Norris DP,

mammalian evolution with two bursts of gene gain on the X Cooper PJ, Swift S, Rastan S: The product of the mouse Xist

chromosome. PLoS Biol 2010, 8. gene is a 15 kb inactive X-specific transcript containing no

conserved ORF and located in the nucleus. Cell 1992,

7. Ross MT, Grafham DV, Coffey AJ, Scherer S, McLay K, Muzny D, 71:515-526.

Platzer M, Howell GR, Burrows C, Bird CP et al.: The DNA

sequence of the human X chromosome. Nature 2005, 25. Brown CJ, Ballabio A, Rupert JL, Lafreniere RG, Grompe M,

434:325-337. Tonlorenzi R, Willard HF: A gene from the region of the human X

inactivation centre is expressed exclusively from the inactive

8. Skaletsky H, Kuroda-Kawaguchi T, Minx PJ, Cordum HS, Hillier L, X chromosome. Nature 1991, 349:38-44.

Brown LG, Repping S, Pyntikova T, Ali J, Bieri T et al.: The male-

26. Wutz A: Gene silencing in X-chromosome inactivation:

specific region of the human Y chromosome is a mosaic of

advances in understanding facultative heterochromatin

discrete sequence classes. Nature 2003, 423:825-837.

formation. Nat Rev Genet 2011, 12:542-553.

9. Johnston CM, Lovell FL, Leongamornlert DA, Stranger BE,

27. Lee JT, Davidow LS, Warshawsky D: Tsix, a gene antisense to

Dermitzakis ET, Ross MT: Large-scale population study of

Xist at the X-inactivation centre. Nat Genet 1999, 21:400-404.

human cell lines indicates that dosage compensation is

virtually complete. PLoS Genet 2008, 4:e9.

28. Ohhata T, Hoki Y, Sasaki H, Sado T: Crucial role of antisense

transcription across the Xist promoter in Tsix-mediated Xist

10. Nguyen DK, Disteche CM: Dosage compensation of the active X

chromatin modification. Development 2008, 135:227-235.

chromosome in mammals. Nat Genet 2006, 38:47-53.

29. Shibata S, Yokota T, Wutz A: Synergy of Eed and Tsix in the

11. Lin H, Gupta V, Vermilyea MD, Falciani F, Lee JT, O’Neill LP,

repression of Xist gene and X-chromosome inactivation.

Turner BM: Dosage compensation in the mouse balances up-

EMBO J 2008, 27:1816-1826.

regulation and silencing of X-linked genes. PLoS Biol 2007,

5:e326.

30. Sun BK, Deaton AM, Lee JT: A transient heterochromatic state

in Xist preempts X inactivation choice without RNA

12. Xiong Y, Chen X, Chen Z, Wang X, Shi S, Wang X, Zhang J, He X:

stabilization. Mol Cell 2006, 21:617-628.

RNA sequencing shows no dosage compensation of the active

X-chromosome. Nat Genet 2010, 42:1043-1047.

31. Anguera MC, Ma W, Clift D, Namekawa S, Kelleher RJ 3rd, Lee JT:

Tsx produces a long noncoding RNA and has general

13. Deng X, Hiatt JB, Nguyen DK, Ercan S, Sturgill D, Hillier LW,

functions in the germline, stem cells, and brain. PLoS Genet

 Schlesinger F, Davis CA, Reinke VJ, Gingeras TR et al.: Evidence

2011, 7:e1002248.

for compensatory upregulation of expressed X-linked genes in

mammals, Caenorhabditis elegans and Drosophila

32. Ogawa Y, Lee JT: Xite, X-inactivation intergenic transcription

melanogaster. Nat Genet 2011, 43:1179-1185.

elements that regulate the probability of choice. Mol Cell 2003,

Using a new strategy to analyse Chip and RNA-seq data, this manuscript 11:731-743.

settles the debate about X to autosomal gene expression levels in

mammalian cells, by showing that X-linked genes are upregulated com- 33. Tsai CL, Rowntree RK, Cohen DE, Lee JT: Higher order

pared to autosomal genes. chromatin structure at the X-inactivation center via looping

DNA. Dev Biol 2008, 319:416-425.

14. Lin H, Halsall JA, Antczak P, O’Neill LP, Falciani F, Turner BM:

 Relative overexpression of X-linked genes in mouse 34. Chureau C, Chantalat S, Romito A, Galvani A, Duret L, Avner P,

embryonic stem cells is consistent with Ohno’s hypothesis. Rougeulle C: Ftx is a non-coding RNA which affects Xist

Nat Genet 2011, 43:1169-1170. expression and chromatin structure within the X-inactivation

This correspondence letter is in agreement with the findings reported in center region. Hum Mol Genet 2011, 20:705-718.

[13].

35. Tian D, Sun S, Lee JT: The long noncoding RNA, Jpx, is a

15. Prince EG, Kirkland D, Demuth JP: Hyperexpression of the X molecular switch for X chromosome inactivation. Cell 2011,

chromosome in both sexes results in extensive female bias of 143:390-403.

X-linked genes in the flour beetle. Genome Biol Evol 2010,

2:336-346. 36. Spencer RJ, del Rosario BC, Pinter SF, Lessing D, Sadreyev RI,

 Lee JT: A boundary element between Tsix and Xist binds the

16. Huynh KD, Lee JT: Inheritance of a pre-inactivated paternal X chromatin insulator Ctcf and contributes to initiation of X-

chromosome in early mouse embryos. Nature 2003, 426:857-862. chromosome inactivation. Genetics 2011, 189:441-454.

By demonstrating the presence of a boundary element, the data in this

17. Okamoto I, Otte AP, Allis CD, Reinberg D, Heard E: Epigenetic paper are instrumental to support the presence of Xist and Tsix active

dynamics of imprinted X inactivation during early mouse chromatin hubs, described in [33]. Removal of the boundary element

development. Science 2004, 303:644-649. disturbs Xist activation.

18. Mak W, Nesterova TB, de Napoles M, Appanah R, Yamanaka S, 37. Monkhorst K, Jonkers I, Rentmeester E, Grosveld F, Gribnau J: X

Otte AP, Brockdorff N: Reactivation of the paternal X inactivation counting and choice is a stochastic process:

chromosome in early mouse embryos. Science 2004, evidence for involvement of an X-linked activator. Cell 2008,

303:666-669. 132:410-421.

19. Augui S, Nora EP, Heard E: Regulation of X-chromosome 38. Augui S, Filion GJ, Huart S, Nora E, Guggiari M, Maresca M,

inactivation by the X-inactivation centre. Nat Rev Genet 2011, Stewart AF, Heard E: Sensing X, chromosome pairs before X

12:429-442. inactivation via a novel X-pairing region of the Xic. Science

2007, 318:1632-1636.

20. Morey C, Avner P: The demoiselle of X-inactivation: 50 years

old and as trendy and mesmerising as ever. PLoS Genet 2011, 39. Bacher CP, Guggiari M, Brors B, Augui S, Clerc P, Avner P, Eils R,

7:e1002212. Heard E: Transient colocalization of X-inactivation centres

www.sciencedirect.com Current Opinion in Cell Biology 2012, 24:397–404

404 Nucleus and gene expression

accompanies the initiation of X inactivation. Nat Cell Biol 2006, 54. Kalantry S, Purushothaman S, Bowen RB, Starmer J, Magnuson T:

8:293-299. Evidence of Xist RNA-independentinitiation of mouse imprinted

X-chromosome inactivation. Nature 2009, 460:647-651.

40. Xu N, Tsai CL, Lee JT: Transient homologous chromosome

pairing marks the onset of X inactivation. Science 2006, 55. Marahrens Y, Panning B, Dausman J, Strauss W, Jaenisch R: Xist-

311:1149-1152. deficient mice are defective in dosage compensation but not

spermatogenesis. Genes Dev 1997, 11:156-166.

41. Donohoe ME, Silva SS, Pinter SF, Xu N, Lee JT: The pluripotency

factor Oct4 interacts with Ctcf and also controls X- 56. Namekawa SH, VandeBerg JL, McCarrey JR, Lee JT: Sex

chromosome pairing and counting. Nature 2009, chromosome silencing in the marsupial male germ line. Proc

460:128-132. Natl Acad Sci USA 2007, 104:9730-9735.

42. Donohoe ME, Zhang LF, Xu N, Shi Y, Lee JT: Identification of a 57. McCarrey JR, Watson C, Atencio J, Ostermeier GC, Marahrens Y,

Ctcf cofactor, Yy1, for the X chromosome binary switch. Mol Jaenisch R, Krawetz SA: X-chromosome inactivation during

Cell 2007, 25:43-56. spermatogenesis is regulated by an Xist/Tsix-independent

mechanism in the mouse. Genesis 2002, 34:257-266.

43. Navarro P, Chambers I, Karwacki-Neisius V, Chureau C, Morey C,

Rougeulle C, Avner P: Molecular coupling of Xist regulation and 58. Baarends WM, Wassenaar E, van der Laan R, Hoogerbrugge J,

pluripotency. Science 2008, 321:1693-1695. Sleddens-Linkels E, Hoeijmakers JH, de Boer P, Grootegoed JA:

Silencing of unpaired chromatin and histone H2A

44. Navarro P, Oldfield A, Legoupi J, Festuccia N, Dubois A, Attia M,

ubiquitination in mammalian meiosis. Mol Cell Biol 2005,

 Schoorlemmer J, Rougeulle C, Chambers I, Avner P: Molecular 25:1041-1053.

coupling of Tsix regulation and pluripotency. Nature 2010,

468:457-460. 59. Turner JM, Mahadevaiah SK, Fernandez-Capetillo O,

Providing clear evidence for involvement of pluripotency factors in Nussenzweig A, Xu X, Deng CX, Burgoyne PS: Silencing of

repression of XCI, as a follow up of a previous study [43], these authors unsynapsed meiotic chromosomes in the mouse. Nat Genet

opened new avenues in our understanding of the link between cell 2005, 37:41-47.

differentiation and XCI.

60. Schoenmakers S, Wassenaar E, van Cappellen WA, Derijck AA, de

45. Navarro P, Avner P: An embryonic story: analysis of the gene Boer P, Laven JS, Grootegoed JA, Baarends WM: Increased

regulative network controlling Xist expression in mouse frequency of asynapsis and associated meiotic silencing of

embryonic stem cells. Bioessays 2010, 32:581-588. heterologous chromatin in the presence of irradiation-

induced extra DNA double strand breaks. Dev Biol 2008,

46. Jonkers I, Barakat TS, Achame EM, Monkhorst K, Kenter A, 317:270-281.

Rentmeester E, Grosveld F, Grootegoed JA, Gribnau J: RNF12 is

an X-Encoded dose-dependent activator of X chromosome 61. Turner JM, Mahadevaiah SK, Ellis PJ, Mitchell MJ, Burgoyne PS:

inactivation. Cell 2009, 139:999-1011. Pachytene asynapsis drives meiotic sex chromosome

inactivation and leads to substantial postmeiotic repression in

47. Barakat TS, Gunhanlar N, Pardo CG, Achame EM, Ghazvini M,

spermatids. Dev Cell 2006, 10:521-529.

 Boers R, Kenter A, Rentmeester E, Grootegoed JA, Gribnau J:

RNF12 activates Xist and is essential for X chromosome 62. Patrat C, Okamoto I, Diabangouaya P, Vialon V, Le Baccon P,

inactivation. PLoS Genet 2011, 7:e1002001. Chow J, Heard E: Dynamic changes in paternal X-chromosome

Following identification of RNF12 as a trans-acting XCI-activator [46], this activity during imprinted X-chromosome inactivation in mice.

study describes the molecular mechanism by which RNF12 exerts its role Proc Natl Acad Sci USA 2009, 106:5198-5203.

as an activator.

63. Okamoto I, Arnaud D, Le Baccon P, Otte AP, Disteche CM,

48. Shin J, Bossenz M, Chung Y, Ma H, Byron M, Taniguchi-Ishigaki N, Avner P, Heard E: Evidence for de novo imprinted X-

 Zhu X, Jiao B, Hall LL, Green MR et al.: Maternal Rnf12/RLIM is chromosome inactivation independent of meiotic inactivation

required for imprinted X-chromosome inactivation in mice. in mice. Nature 2005, 438:369-373.

Nature 2010, 467:977-981.

64. Renfree MB, Hore TA, Shaw G, Graves JA, Pask AJ: Evolution of

Similarities and differences between iXCI and rXCI are a complex field of

genomic imprinting: insights from marsupials and

study in which this manuscript provides clear evidence for a role of RNF12

monotremes. Annu Rev Genomics Hum Genet 2009, 10:241-262.

in iXCI.

65. Ferron SR, Charalambous M, Radford E, McEwen K, Wildner H,

49. Chadwick LH, Pertz LM, Broman KW, Bartolomei MS, Willard HF:

Hind E, Morante-Redolat JM, Laborda J, Guillemot F, Bauer SR

Genetic control of X chromosome inactivation in mice:

et al.: Postnatal loss of Dlk1 imprinting in stem cells and niche

definition of the Xce candidate interval. Genetics 2006,

173:2103-2110. astrocytes regulates neurogenesis. Nature 2011, 475:381-385.

66. Okamoto I, Patrat C, Thepot D, Peynot N, Fauque P, Daniel N,

50. Duret L, Chureau C, Samain S, Weissenbach J, Avner P: The Xist

 Diabangouaya P, Wolf JP, Renard JP, Duranthon V et al.:

RNA gene evolved in eutherians by pseudogenization of a

Eutherian mammals use diverse strategies to initiate X-

protein-coding gene. Science 2006, 312:1653-1655.

chromosome inactivation during development. Nature 2011,

51. Flynn M, Saha O, Young P: Molecular evolution of the LNX gene 472:370-374.

family. BMC Evol Biol 2011, 11:235. Studies with pre-implantation human and mouse embryos indicate

important differences in the XCI mechanism between mouse and other

52. Pandey RR, Mondal T, Mohammad F, Enroth S, Redrup L,

mammalian species, and therefore this paper emphasizes the importance

Komorowski J, Nagano T, Mancini-Dinardo D, Kanduri C:

of studying iXCI and rXCI in an evolutionary context.

Kcnq1ot1 antisense noncoding RNA mediates lineage-

specific transcriptional silencing through chromatin-level 67. Berletch JB, Yang F, Disteche CM: Escape from X inactivation in

regulation. Mol Cell 2008, 32:232-246. mice and humans. Genome Biol 2010, 11:213.

53. Nagano T, Mitchell JA, Sanz LA, Pauler FM, Ferguson-Smith AC, 68. Horvath JE, Sheedy CB, Merrett SL, Diallo AB, Swofford DL,

Feil R, Fraser P: The Air noncoding RNA epigenetically silences Program NCS, Green ED, Willard HF: Comparative analysis of the

transcription by targeting G9a to chromatin. Science 2008, primate X-inactivation center region and reconstruction of the

322:1717-1720. ancestral primate XIST locus. Genome Res 2011, 21:850-862.

Current Opinion in Cell Biology 2012, 24:397–404 www.sciencedirect.com