- Research article
- Open Access
Retinoic acid accelerates downregulation of the Xist repressor, Oct4, and increases the likelihood of Xist activation when Tsix is deficient
BMC Developmental Biology volume 10, Article number: 90 (2010)
Imbalances in X-linked gene dosage between the sexes are resolved by transcriptionally silencing one of two X-chromosomes in female cells of the early mammalian embryo. X-inactivation is triggered by expression of the non-coding Xist gene. In turn, Xist is dually regulated by the antisense Tsix RNA and by the Oct4 pluripotency factor. Although there is general agreement that Tsix is an inhibitor of Xist, some laboratories have observed ectopic Xist induction in differentiating male ES cells when Tsix is mutated, whereas we have not observed significant changes in Xist. These observational differences have led to fundamentally diverse models of X-chromosome counting. Here, we investigate if different methods of cell differentiation and use of all -trans retinoic acid (RA) could be causative factors and how they might impact Xist expression.
We compared suspension and cell-adhesion cultures in the presence or absence of RA and find that RA significantly impacts Xist expression in Tsix-mutant male cells. Whereas the standard embryoid body method infrequently leads to ectopic Xist expression, adding RA generates a significant number of Xist-positive male cells. However, while normal Xist clouds in wild-type female cells are robust and well-circumscribed, those found in the RA-treated mutant males are loosely dispersed. Furthermore, ectopic Xist expression does not generally lead to complete gene silencing. We attribute the effect of RA on Xist to RA's repressive influence on Oct4, a pluripotency factor recently shown to regulate Tsix and Xist. RA-treated ES cells exhibit accelerated decreases in Oct4 RNA levels and also display accelerated loss of binding to Xist intron 1. When Tsix is deficient, the faster kinetics of Oct4 loss tip the equilibrium towards Xist expression. However, the aberrant Xist clusters are unlikely to explain elevated cell death, as X-linked silencing does not necessarily correlate with the qualitatively aberrant Xist clusters.
We conclude that RA treatment leads to premature downregulation of Oct4 and partial derepression of Xist irrespective of X-chromosome counting. RA-induced Xist clusters in male cells do not result in global or stable silencing, and excess cell death is not observed. These data and RA's known pleiotropic effects on ES transcription networks suggest that RA differentation bypasses normal X-inactivation controls and should be used judiciously. We propose that the likelihood of Xist expression is determined by a balance of multiple Xist activators and repressors, and that levels of Oct4 and Tsix are crucial toward achieving this balance.
Sex dosage compensation ensures equal X-linked gene expression between XX and XY individuals. In mammals, this balance is achieved by transcriptionally silencing an entire X-chromosome in females through a process called X-chromosome inactivation (XCI) . XCI is accomplished independently in each cell primarily by the Xist/Tsix pair of sense/antisense non-coding RNAs located within the X-inactivation center (Xic) [2–5]. Xist upregulation and cis-coating of an X-chromosome represent important steps in the X-inactivation process [6, 7], and are followed by recruitment of the PRC2 complex [8–10] and other silencing factors to initiate chromosome-wide inactivation and compaction into heterochromatin. Because X-inactivation occurs very early in embryonic development, cell culture models have been developed to facilitate analysis. Female mouse embryonic stem (ES) cells can undergo random X-inactivation when differentiated in vitro  and have therefore served as a powerful system with which to study this phenomenon. Xist expression remains low on both Xs in undifferentiated ES cells, but upon differentiation, Xist becomes upregulated only on the future inactive X [11, 12]. Xist thus serves as the trigger for the silencing step during the X-inactivation program.
Xist's central nature to XCI has led to intensive investigation of how this gene is regulated. One established regulator is Tsix, the antisense repressor of Xist [13, 14]. Tsix RNA is expressed from both Xs in undifferentiated female cells, but its expression becomes monoallelic during the process of cell differentiation and XCI. The chromosome that transiently retains Tsix expression becomes the future active X (Xa), while the chromosome that extinguishes Tsix expression first becomes the future inactive X (Xi). Tsix has been proposed to regulate X-chromosome pairing, counting, and the mutually exclusive choice of Xa and Xi [15–17]. Indeed, various knockouts of Tsix (and its upstream enhancer Xite ) have led to nonrandom XCI as well as effects on counting and pairing [13–16].
Although Tsix's repressive role in female cells seems clear, there has been debate over Tsix's role in male cells. The original 3.7 kb deletion encompassing the major Tsix promoter in male ES cells (TsixΔCpG/Y ) did not cause a significant degree of ectopic Xist upregulation upon differentiation (<1%). TsixΔCpG/Y ES cells were phenotypically normal and capable of generating chimeric mice with germline transmission, and male offspring were born at expected frequencies . This finding led to the conclusion that, whereas Tsix represses Xist, an additional factor - missing in male ES cells and present only in female cells - is required to induce Xist expression in cells with supernumerary X-chromosomes. It was therefore proposed that X-chromosome counting involves two factors: a 'blocking factor' that blocks Xist expression through Tsix on the Xa, and a 'competence factor' that induces Xist expression on Xi .
Significantly, however, other Tsix mutations show variable phenotypes in male cells (Fig. 1). Sado et al. also showed that Tsix+/- female mice can yield viable male offspring carrying the mutation, but 8-21% of cells isolated from E7.5 male embryos showed ectopic Xist expression, an observation that could be explained by either ectopic Xist expression in the embryo proper or by presence of extraembryonic tissue (which is subject to imprinted XCI and therefore be severely affected by the maternal Tsix deletion) . Another Tsix-specific mutation showed upregulation of Xist RNA in as many as 38% of differentiating male ES cells in culture [19, 20]. A 65 kb deletion of the region 3' to Xist (including Tsix) also resulted in significant induction of Xist RNA clouds when XO (XLD ) or XY (CK35Δ65 ) cells underwent differentiation. Furthermore, smaller mutations including only Tsix elements were also found to have ectopic Xist cloud formation in a subset of differentiating male cells (CK35Δ34 and CK35ΔAV ). These observations have led to the alternative hypothesis that Xist is activated by default when Tsix is mutated, obviating the need for a competence factor involved in female-specific expression of Xist. Differences with regard to Xist expression in Tsix-deficient male cells (hereafter referred to as XΔ/Y ES cells) have therefore led to fundamentally different models for X-chromosome counting [3, 13, 15, 22, 23].
Here, we seek to understand why analyses of male cells in different laboratories should lead to differing results. Although the nature of the mutations could lead to disparate results, another potential difference among the studies might be the method of cell differentiation used to induce XCI. Whereas we differentiate ES cells by the embryoid body method (EB) in the absence of LIF, many laboratories also add retinoic acid (RA) to the differentiation medium. Although some studies have used both the EB and EB+RA methods and found ectopic Xist RNA in both cases [22, 24], it remains formally possible that the combination of induction methods, culture conditions, and nature of the mutations could affect the Xist phenotype.
The idea that RA affects Xist expression finds support in two recent studies that demonstrated a role for Oct4 in repressing Xist by direct and indirect means [25, 26]. One study showed that Oct4 positively regulates Tsix and that downregulation of Oct4 during cell differentiation triggers X-chromosome pairing, leads to monoallelic repression of Tsix, and consequently enables Xist upregulation on one chromosome . The other study showed that Oct4 also associates with intron 1 of the Xist gene and proposed that Oct4 directly represses Xist . Combined, these studies suggest that Xist is linked to cell differentiation through Oct4's effects on Tsix and Xist. Because it is known that Oct4 is rapidly downregulated in the presence of retinoic acid [27–30] (by binding the regulatory nuclear hormone receptors, RAR and RXR ), we consider the possibility that RA treatment to induce cell differentiation may have unanticipated effects on XCI and thereby confound XCI analysis. Here, we set out to test the idea that RA may affect Xist levels through Oct4 by studying one well-characterized Tsix mutation, TsixΔCpG, in male ES cells. Our results indicate that differentiation methods indeed influence Xist expression in these ES cells.
Ectopic Xist upregulation in TsixΔCpG male ES cells differentiated in the presence of retinoic acid
We differentiated wild-type and TsixΔCpG male and female ES cells using variations of four published in vitro differentiation techniques: (1) embryoid body formation (EB method), whereby ES cell clusters are cultured in suspension for 4 days, in the absence of LIF, and then plated onto a solid matrix to form EB outgrowths; (2) EB method in the presence of 100 nM of all-trans retinoic acid (EB+RA method); (3) adherent cultures whereby ES cells are plated at low-density on gelatinized tissue culture plates without feeders and LIF (TC method); and (4) TC method in the presence of 100 nM all-trans retinoic acid (TC+RA method).
By Xist RNA FISH (fluorescence in situ hybridization), we detected a slight increase in the percentage of cells with Xist clouds using the embryoid body (EB) versus the low-density cell adhesion (TC) method (Fig. 2A, 32% in X/X EB compared to 18% in X/X TC at day 6). Consistent with previous analysis , Xist clouds were rare in XΔ/Y cells in the absence of RA (<2-5%, n>200 cells) and were never observed in wild-type male cells under any differentiation condition and time point examined.
However, while no significant differences were seen for most cell lines between EB and EB+RA or between TC and TC+RA methods, we did detect differences for XΔ/Y cells (Fig. 2A,B). The differences were significant for all differentiation days (bracketed pairwise comparisons, P < 0.05). Ectopic Xist RNA clusters could be seen in 10-30% of XΔ/Y cells (n > 200) when induced to differentiate with RA. The ectopic Xist clouds were almost always more dispersed than wild-type Xist clouds, which are generally intense, compact, and well-defined (Fig. 2B, lower panels). This observation suggested a problem with either full Xist induction or transcript localization in mutant cells when differentiated with RA. We refer to these abnormal clouds as 'dispersed clusters' to emphasize these qualitative differences from wild-type Xist RNA clouds. X- and Y-chromosome DNA FISH confirmed that all ES cell lines used in this study exhibited the correct XX or XY constitution (Fig. 2C).
We conclude that the method of cell differentiation can impact Xist expression in XΔ/Y ES cells. Specifically, differentiation in the presence of RA leads to the appearance of ectopic Xist clusters only when Tsix is deleted and, when they do appear, the RNA is loosely clustered.
RA induction accelerates loss of Oct4 binding to Xist
Recent studies suggest that Xist is negatively regulated not only by Tsix but also by Oct4 [25, 26]. Because RA has been shown to negatively regulate the murine Oct4 promoter during ES cell differentiation [27–30], we asked if increased Xist expression might be due to accelerated Oct4 loss; i.e., in the absence of the second regulatory arm (Tsix), RA-treated cells would ectopically express Xist. By real-time RT-PCR, we compared relative Xist and Oct4 expression levels in differentiating male ES cells using the four described differentiation methods. In wild-type male cells, Xist was rapidly downregulated upon cell differentiation in a manner that correlated with downregulation of Oct4. In XΔ/Y cells differentiated by the EB method, Xist RNA was mildly upregulated upon cell differentiation when Oct4 levels dropped (Fig. 3), but this modest upregulation of Xist did not result in spreading of Xist RNA along the X, consistent with prior findings . Thus, Tsix represses Xist even in male ES cells, but its absence only modestly increases Xist levels [13, 14].
Differentiation with RA revealed two differences. First, while wild-type male cells appropriately repressed Xist expression, XΔ/Y ES cells displayed a significant rise in overall Xist RNA levels using the RA method (Fig. 3). This is consistent with the observation that, when differentiated via the EB method, <2-5% of XΔ/Y cells showed Xist clusters, but this percentage increased to 10-30% in the presence of RA (Fig. 2A). Similar results were obtained for both EB+RA and TC+RA conditions. At no time of differentiation up to day 6, however, did Xist levels increase by more than 2- to 10-fold in mutant male cells. Thus, Xist upregulation is considerably blunted in RA-induced Tsix-deficient male cells when compared to the 30- to 100-fold induction that typically accompanies XCI in normal female cells (data not shown) [13, 32–34]. These results showed that RA differentiation led to a higher likelihood of Xist induction in Tsix-deficient male ES cells.
Second, Oct4 downregulation generally occurred more quickly in both wild-type and mutant male ES cells when compared to the same cells grown in the absence of RA (Fig. 3; see pairwise t-tests done for each cell line comparing between EB/EB+RA (red asterisks) and TC/TC+RA (green asterisks)). While Oct4 RNA levels were similar in wild-type and mutant ES cells in the undifferentiated state, differentiation revealed a significant difference in Oct4 mRNA levels at days 4 and/or 6 dependent on whether RA was added to the culture or not. RA accelerated the rate of Oct4 decrease regardless of whether the ES cells were grown as EBs (compare EB versus EB+RA) or as adherent cells (compare TC versus TC+RA). The greatest difference was evident when adherent cells were used. Although Oct4 mRNA differences are small, cells at this early stage of differentiation may be more sensitive to subtle changes in levels since a specific range of Oct4 concentration is required to maintain ES cells as pluripotent stem cells . We note that the cell-adhesion method of differentiation has also been used and shown to form ectopic Xist RNA in Tsix-deficient male cells [19, 22].
To determine whether the accelerated overall decrease in Oct4 levels has a direct impact on Xist expression, we asked if the kinetics of Oct4 loss at the Xist locus were altered in RA-containing versus RA-free cultures. Oct4 was previously shown to bind Xist intron 1 in ES cells . It was proposed that Oct4 binding represses Xist in undifferentiated cells and that the loss of Oct4 binding during cell differentiation enables Xist upregulation. We therefore investigated whether cell differentiation triggered by RA might lead to accelerated loss of Oct4 binding to Xist intron 1 and increased Xist expression. By performing chromatin immunoprecipitations (ChIPs) using α-Oct4 antibodies on wild-type and Tsix-deficient male cells differentiated in parallel, we found that indeed Oct4-binding to Xist intron 1B was lost significantly more quickly when cells were differentiated in the presence of RA (Fig. 4; P < 0.05, see bracketed pairwise comparisons). We found that indeed Oct4-binding to Xist intron 1B was lost significantly more quickly when cells were differentiated in the presence of RA (Fig. 4; P < 0.05, see bracketed pairwise comparisons). Differences were evident as early as day 2. The difference was most evident for WT EB with versus without RA treatment; however, we note that the difference between mutant EB plus versus minus RA treatment might have been masked by somewhat lower Oct4 binding by d2. As a negative control, we looked at a region within intron 1 of Xist (intron 1A, located 0.6 kb away from intron 1B), which is not known to bind Oct4 , and found no statistically significant differences between EB and EB+RA methods. Thus, RA treatment leads to a more rapid loss of Oct4 binding to Xist intron 1 than is typically seen when ES cells are differentiated without RA.
Taken together, these data indicated that RA treatment has little effect on Xist RNA upregulation when Tsix is present, but increases the likelihood with which Xist will be induced when Tsix is deficient. We propose that RA's effect in the absence of Tsix is related to the rapid downregulation of Oct4 by RA, which in turn leads to accelerated loss of Oct4 binding to Xist intron 1. We suggest that, when these events occur in a Tsix-deficient background, Xist is more likely to become derepressed. Overall Oct4 levels are also downregulated in wildtype cells when treated with RA, but the persistence of Tsix would antagonize the derepressive forces. These findings are consistent with the idea that Oct4 and Tsix act in parallel to regulate Xist expression [25, 26].
X-chromosome silencing in mutant cells with dispersed Xist clusters
We next addressed whether the 2- to 10-fold increases in Xist levels and formation of dispersed Xist RNA clusters led to X-chromosome silencing in RA-treated Tsix-mutant cells. First, to correlate Xist expression with gene silencing on a cell-by-cell basis, we carried out RNA FISH using probes to simultaneously detect Xist and nascent transcription from the X-linked Pgk1 locus. In wild-type X/X cells that have undergone XCI, nascent Pgk1 RNA signals are generally not observed on the Xi, whereas they can be observed in ~60% on the Xa at a given time point  (Fig. 5A and data not shown). We reasoned that, if ectopic Xist clusters led to gene silencing in XΔ/Y cells, Pgk1 RNA would not be detectable on the Xist-coated XΔ. On the other hand, if the aberrant Xist clusters could not silence genes in cis, 60% of XΔ would continue to express Pgk1. When differentiated using the RA method, 30% (21/70) of XΔ/Y cells with Xist accumulation showed nascent Pgk1 signals (Fig. 5A), suggesting that complete silencing did not always follow Xist accumulation at this time point, though a fraction of EB+RA cells appeared to undergo inactivation. By contrast, when differentiated using the EB method, XΔ/Y cells did not express Xist, and Pgk1 expression was observed at the expected frequency of ~60% (Fig. 5A). Thus, although Xist was partially derepressed by RA treatment of XΔ/Y cells, silencing of genes in cis did not proceed to completion.
With respect to XCI, the X can be divided into genic versus repetitive element fractions [36, 37]. To assess whether repetitive element silencing proceeded to completion, we carried out RNA FISH using probes from the Cot-1 fraction , the genomic fraction containing the most highly repetitive elements. In cells where XCI is firmly established, the inactive X lies either completely or partially within a "Cot-1 hole" (i.e., the inactive X is depleted for repetitive element expression) [36–39]. In wild-type EB cells on day 6 of differentiation, 22.2% of Xist clouds lay entirely within a Cot-1 hole and 58.7% were partially within a Cot-1 hole; only 19% did not appear to overlap with a Cot-1 hole at all (Fig. 5B). By contrast, in RA-differentiated XΔ/Y EB, 18.8% of aberrant Xist RNA clusters were located within a Cot-1 hole and 44.9% partially within a Cot-1 hole, whereas 36.3% did not overlap with a Cot-1 hole at all (Fig. 5B). Because the ectopic Xist RNA clusters in XΔ/Y cells were very diffuse, Xist RNA could often be found in Cot-1+ regions (and were scored as 'partially within a Cot-1 hole'). Interestingly though, Xist RNA clusters scored as 'entirely within a Cot-1 hole' appeared more compact, suggesting that perhaps only those cells were able to sufficiently silence their X-chromosomes. Taken together, these results demonstrated that a significant fraction of RA-treated mutant male cells had not completed repeat silencing, suggesting either a delay or a failure to do so. Thus, both genic and repeat silencing were at least partially compromised in XΔ/Y cells with ectopic and diffuse Xist expression induced by RA-mediated cell differentiation.
Next we reasoned that, if Xist induction by RA led to XCI in part, differentiating XΔ/Y cells would show cell death in excess of wild-type cells. However, an equivalent level of cell death was observed for mutant and wild-type cells, suggesting that the XΔ chromosome could not have been silenced to a significant extent (Fig. 5C). Notably, regardless of the Tsix genotype, the magnitude of cell death was consistently greater and the population doubling time was substantially slower when cells were differentiated using the RA method (Fig. 5C). Comparison of EB versus EB+RA conditions showed that <50% of cells - either wild-type or mutant - survived RA treatment during the first 6 days. Comparison of TC versus TC+RA showed that addition of RA caused a >95% drop in cell viability over the same period. These results demonstrate that the Tsix mutation does not have a major effect on the viability of male cells and that RA alone - irrespective of the Tsix deficiency and RA's partial effect on Xist induction - has major effects on cell viability that are unrelated to the XCI pathway. This effect of RA raises the possibility of selection artifacts when mutations are analyzed in the presence of RA.
Here, we have shown that differentiation in the presence of RA results in partial upregulation of Xist RNA when Tsix is deleted in male ES cells. We believe that this results from accelerated loss of Oct4 and that the depletion of both Tsix and Oct4 (both repressors of Xist) in differentiating ES cells leads to ectopic Xist expression. However, the ectopic Xist clusters appear more dispersed than those typically seen in wild-type female cells, and the degree of Xist activation is an order of magnitude lower. Consistent with these aberrant qualities, the dispersed Xist clusters do not globally or stably silence the male X-chromosome. Thus, RA-induced Xist expression in a Tsix-deficient background does not lead to proper and complete XCI. We also observed that RA treatment uniformly causes massive cell death (>50%) regardless of the Tsix genotype. Nonetheless, the fraction of Tsix-deficient male cells that do initiate silencing may be selected against, thus enriching for cells that have not properly initiated XCI. In view of our findings, we suggest that RA differentiation of XCI mutants should be analyzed carefully because RA treatment could either create selection artifacts or indirectly affect XCI through RA's established pleiotropic effects.
It has long been known that RA regulates expression of the POU-domain transcription factor Oct4 (also referred to as Oct3/4 and Pou5f1), which is restricted to pluripotent stem cells and is downregulated when induced to differentiate by treatment with RA [27–30]. RA and other vitamin A derivatives have pervasive effects on developmental processes--including vertebrate embryogenesis, growth, differentiation, and homeostasis--primarily through binding to the RAR and RXR nuclear hormone receptors . RAR isoforms mediate transcriptional activation through binding of either 9-cis or all-trans RA [40, 41], while RXRs (α, β, γ) work specifically through 9-cis-RA [31, 42–44]. These receptors form homo- or heterodimeric complexes and bind RA-responsive elements in target genes to transcriptionally mediate expression [45–49]. Oct4 is regulated at the level of transcription  through binding of RARs , chick ovalbumin upstream promoter transcription factors (COUP-TFs) , germ cell nuclear factor (GCNF), as well as other proteins , to upstream enhancer regions (designated DE and PE for distal and proximal enhancers, respectively)  and an Sp1 site  located within the Oct4 promoter. Upon RA-induced differentiation, these specific factors bind to and downregulate Oct4 in ES and EC cells [50, 51, 54]. It is possible that these genome-wide and pleiotropic effects of RA would have indirect effects on XCI, which might thereby explain differences in the Xist phenotype observed in our Tsix-deficient background from those of others.
Xist differences in EB- versus RA-differentiated cultures can be summarized by the model shown in Figure 6. Our model supposes that Xist is independently repressed by Oct4 and Tsix. In undifferentiated wild-type male ES cells (Fig. 6A), high levels of Oct4--along with Nanog and Sox2--bind within intron 1 of the Xist gene [25, 26] and keep Xist expression at a "basal" level (<10 copies per cell) [8, 55, 56]. Oct4 also binds to Tsix/Xite to transactivate antisense transcription (not shown), which in turn also represses Xist [25, 26]. During differentiation into EBs, Oct4 expression decreases slowly and remains high enough during early differentiation to maintain Xist repression. At the same time, persistent Tsix expression stably switches off Xist ("off" state) [33, 55, 57–59]. This early stage corresponds to the reversible Xist-dependent time window for XCI . Later during differentiation, Oct4 is significantly downregulated, Tsix is extinguished, and Xist is safely and stably repressed, possibly by a transcriptional gene silencing (TGS)-like mechanism and by DNA methylation of the Xist promoter [59, 61] (after the Xist-dependent time window ). Under conditions of RA-induced differentiation and accelerated loss of Oct4, the outcome would be similar, as Tsix ensures continued repression of Xist.
In XΔ/Y cells, the outcome would be different (Fig. 6B). Ordinarily, the rapid drop in Oct4 levels during RA-induced differentiation would have little impact on Xist, as an intact Tsix would suppress full Xist activation. However, deletion of Tsix would eliminate the second arm of Xist regulation, resulting in ectopic Xist expression. The increase in Xist expression would be modest, would not yield robust Xist clouds, and would not necessarily be followed by chromosome-wide inactivation, explaining why we did not observe a noticeable increase in cell death.
Although we prefer this model, we cannot exclude other potential explanations for the behavior of TsixΔCpG/Y cells in general. It is clear that the Xist promoter is not properly silenced in RA-differentiated Tsix-mutant male cells. Therefore, it is possible that a very small minority of TsixΔCpG/Y cells that initiate XCI during the early time window (when the Xi can be established) are selected against and are rarely observed. Because Xist is not properly silenced in the absence of Tsix, cells might continue to upregulate Xist at later stages and form improper and abnormal Xist RNA clusters, which would not be capable of silencing the X. These cells might then survive longer and be observed by our assay. Although this scenario is possible, we do not believe that it can explain the observations related to RA made in this study. Firstly, ectopic Xist expression is rarely observed without RA treatment. Secondly, in the presence of RA, Xist RNA clusters are actually observed during the early time window during the silencing-competent phase of XCI. Thus, we believe that the more likely explanation is the one offered in Figure 6.
Interestingly, the percentage of RA-differentiated cells with ectopic Xist upregulation in a Tsix-mutant background is similar among other studies and our own (~10% at day 2) [19, 20, 22]. Our results are also consistent with those reported by Navarro et al. for an Oct4-null male ES line. Although these authors reported that Xist upregulation was not seen in Oct4-null male ES cells when differentiated by retinoic acid alone , this line contained a tetracycline-repressible Oct4 transgene which lacked the endogenous Oct4 promoter and upstream enhancer regions (PE and DE). As the retinoic acid repressible elements [50, 62] are located within this region, their Oct4-null male line would not be directly affected by RA treatment.
Our results suggest that the likelihood of Xist expression is determined by a balance of multiple Xist activators and repressors, consistent with other proposed models [13, 15, 23, 63–65]. Deletion, mutation or downregulation of repressive factors would predispose that X-chromosome to inactivation, but would ordinarily not lead to ectopic inactivation in male cells when other repressors are present. Thus, deletion of the Tsix locus alone in male cells would prevent binding of XCI repressive factors to the single X-chromosome but not lead to ectopic inactivation. However, the absence of Tsix in combination with premature loss of Oct4-binding to Xist intron 1 due to RA-induced differentiation in TsixΔCpG/Y male cells would be sufficient to derepress Xist expression independently from this sensing mechanism.
This idea may explain why some investigators have observed the appearance of Xist clusters even when Tsix-deficient cells were differentiated using the EB method. In our hands, TsixΔCpG/Y ES cells rarely show Xist clusters (<5%). Sado et al. have reported that the TsixpAA2Δ1.7/Y mutant shows slightly more Xist clusters during early differentiation by the EB method (~10%) [14, 24], which may be explained by the severity of this Tsix mutation. The TsixpAA2Δ1.7 mutation is a severe truncation mutation with lethality being highly penetrant in the mouse [14, 24], whereas the TsixΔCpG mutation shows ~10% residual antisense expression and shows lower penetrance of lethality in mice [13, 66]. The combination of a severe TsixpAA2Δ1.7 mutation with RA-accelerated Oct4 depletion could explain why Xist upregulation is more frequently seen than in the case for our TsixΔCpG mutation. The 65 kb deletion of Avner and colleagues yields an even greater number of Xist+ cells when differentiated into EBs [19, 22]. Because the 65 kb deletion includes all of the 5' end of Tsix and Xite and deletes two of the three known Oct4-binding sites within the Xic [25, 26], we propose that combining it with accelerated decreases in both RA-induced Oct4 transcription and binding would further increase the likelihood of Xist expression.
We argue that the moderate elevation of Xist RNA levels in RA-treated XΔ/Y cells is not due to aberrant X-chromosome counting but to premature Oct4 downregulation and accelerated loss of Oct4-binding to Xist intron 1. RA also has general pleiotropic effects on transcription that could further affect the XCI process. Our observations suggest that cells lacking Tsix - when differentiated under more physiological conditions (such as that simulated by LIF withdrawal and formation of EBs) - remain capable of suppressing Xist and preserving viability. These ideas are consistent with the two-factor model, which proposes that a repressive "blocking" factor and an activating "competence" factor coordinately regulate the X-chromosome counting process [13, 15]. By this model, Xist RNA is induced and XCI is initiated only when the blocking factor is not bound in cis and a competence factor (expressed only in X/X cells) is present. Both X/Y and XΔ/Y cells would lack the competence factor and be incapable of fully initiating XCI even if Tsix and the blocking factor were eliminated. Apart from the consequences shown here, RA's acceleration of Oct4 downregulation would also be expected to alter the timing of X-chromosome pairing, an event proposed to regulate X-chromosome counting, and thereby account for why different studies observe pairing on different differentiation days [16, 17]. Relevantly, Oct4 has been shown to control the timing of X-chromosome pairing . We suggest, therefore, that RA should be used judiciously in studies of X-chromosome counting and other events surrounding the initiation of XCI.
ES cell lines and culture
Wild-type male (J1), female (EL16.7), and Tsix-mutant (CG7 male and 3F1 female) ES cell lines and culture conditions have been previously described .
ES cell differentiation
Each ES cell line was differentiated for six days using one of the following differentiation methods: (1) suspension cultures forming embryoid bodies (EB), (2) suspension cultures forming EBs + 100 nM of all-trans retinoic acid (EB+RA), (3) cells plated at low density on 0.2% gelatinized tissue culture plates (TC), and (4) cells plated at low density on 0.2% gelatinized tissue culture plates + 100 nM all-trans retinoic acid (TC+RA). LIF (leukemia inhibitory factor) was removed in all differentiation methods. EB and EB+RA cultures had a starting concentration of ~5 × 105 cells/60 cm2, while TC and TC+RA cultures had a starting concentration of ~1 × 105 cells/60 cm2. Culture medium was changed every two days, and 100 nM all-trans retinoic acid was added fresh where appropriate. All-trans retinoic acid (Sigma Cat # R2625) was diluted in DMSO as 10 mM stock solutions, stored in light-protected vials at -20°C, and diluted in culture media just prior to use. All experiments were performed three times. Viable cells were counted using a Cellometer Auto cell counter (Nexcelom Biosciences).
RNA fluorescence in situ hybridization
ES cells were trypsinized into single cells, cytospun onto glass slides, fixed in 4% paraformaldehyde/1xPBS, and FISH was carried out as described . Xist, Pgk1, and Cot-1 RNA were detected using double-stranded DNA probes labeled with Cy3-dUTP or FITC-dUTP either by nick-translation (Roche) or using the Prime-It Fluor Fluorescence Labeling Kit (Stratagene). Digital images were taken with a Nikon Eclipse 90i microscope (Nikon Instruments, Inc.) and processed using Volocity software (Improvision). In brief, z-sections were captured at 0.2 μm intervals, and 3-D images were projected onto a single 2-D plane. Cells were counted and scored for the presence or absence of an Xist cloud or cluster. X- and Y-chromosome paint was performed on undifferentiated cells as previously described  using StarFISH Cy3- or FITC-labeled probes from Cambio.
Total cellular RNA was isolated from TRIzol reagent (Invitrogen) by phenol/chloroform extraction and treated with TURBO DNase (Ambion). RNA (1 μg) was reverse transcribed into cDNA with random hexamers using the SuperScript III First-Strand Synthesis System (Invitrogen) following the manufacturer's instructions. Real-time PCR was performed on a Bio-Rad iCycler machine with SYBR-Green iQ Mix (Bio-Rad). After an initial denaturation at 95°C for 8:30 min, reactions were amplified for 40 cycles: 95°C for 30 s, 58°C for 30 s, 72°C for 30 s, followed by dissociation curve analyses at 55°C for 1 min and 55°C + 0.5°C every 20 s. Xist primers amplified exons 1-3 of the cDNA and were as follows: NS66, 5'-GCTGGTTCGTCTATCTTGTGGG-3' and NS33, 5'-CAGAGTAGCGAGGACTTGAAGAG-3' . Oct 4 and GAPDH primers were as follows: Oct4F, 5'-GAAGCAGAAGAGGATCACCTTG-3'; Oct4R, 5'-TTCTTAAGGCTGAGCTGCAAG-3'; GAPDH1F, 5'-ATGAATACGGCTACAGCAACAGG-3'; GAPDH1R, 5'-CTCTTGCTCAGTGTCCTTGCTG-3'. All primers were added at a final concentration of 200 nM. The Ct (defined as the cycle number at which the fluorescence exceeds the threshold value) was determined for each reaction run in triplicate, and relative-fold differences were calculated using the 2-ΔΔCt method normalized to GAPDH levels and female fibroblasts (for Xist) or undifferentiated ES cells (for Oct4) as a reference sample .
Chromatin immunoprecipitations (ChIPs)
ChIP analyses were carried out using a modified protocol from Millipore, and the results were averaged for two independent biological replicates of wild-type and TsixΔCpG male cells at day 0, day 2, day 4, and day 6 of differentiation; error bars denote standard errors of the means.
Briefly, 1-2 × 107 cells were trypsinized into a single-cell suspension and crosslinked with formaldehyde to a final concentration of 1% at 37°C for 10 min. Crosslinking was quenched with glycine (125 mM final), and cells were pelleted at 640 g for 5 min and washed twice with 1× phosphate buffered saline containing protease inhibitors (Roche mini cOmplete protease inhibitor cocktail tablets, EDTA-free #11-836-170-001). Nuclei were isolated from fixed cells by washing once with Buffer A-NP-40 (5 mM PIPES, pH 8.0, 85 mM KCl, 0.5% NP-40), incubated on ice for 10 min, then washed with Buffer A (5 mM PIPES, pH 8.0, 85 mM KCl) and Lysis Buffer (10 mM Tris-HCl, pH 8.0, 10 mM NaCl, 3 mM MgCl2, 0.5% NP-40). Pelleted nuclei were resuspended in MNase buffer (10 mM Tris-HCl, pH 8.0, 10 mM NaCl, 3 mM MgCl2, 1 mM CaCl2, 4% NP-40) containing protease inhibitors and 1% SDS. Lysates were sonicated using the Bioruptor XL (Diagenode) [15 min total with 30 s on- and 30 s off-cycles]. Sonicated lysates were centrifuged at 16,100 g for 10 min, and the supernatants were stored at -80°C.
All IP steps were performed at 4°C. For each immunoprecipitation, 50 μl of Dynal Protein G-magnetic beads (Invitrogen # 100-04D) were first incubated with 5 μg of appropriate antibody for 2 hours with rotation, then with sonicated supernatants (~2.0 × 106 cells) overnight. IP samples were washed twice each with Low Salt TSE 150 (20 mM Tris-HCl, pH 8.0, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 150 mM NaCl), High Salt TSE 500 (20 mM Tris-HCl, pH 8.0, 0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl), LiCl Buffer (10 mM Tris-HCl, pH 8.0, 250 mM LiCl, 1% NP-40, 1% deoxycholate, 1 mM EDTA), and TE Buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA). Protein/antibody complexes were eluted from the beads with freshly made Elution Buffer (50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1% SDS, 50 mM NaHCO3) incubated at 65°C for 10 min. Crosslinks were reversed by digestion with 80 μg proteinase K at 65°C for 4 hours, and DNA was recovered by phenol/chloroform extraction.
ChIP antibodies included goat polyclonal anti-Oct4 (Santa Cruz #sc8628) and normal rabbit IgG (Cell Signaling #2729). Quantitative PCR was performed using an iCycler iQ real-time PCR detection system (Bio-Rad) with primers specific to Xist intron 1 (site A: p63/p64, distanced 0.6 kb from site B: p65/p66). Primer sequences were as follows: p63, 5'-CTGAAGATGGTGATGGCGAGTTG-3'; p64, 5'-AAAGAGTTCCCCAAATTAGTGTCCTG-3'; p65, 5'-ATGTTTCCTTTTGAAGCAGTTACTTGTAC-3'; p66, 5-CATTGTCTGGCTCTCTAGGTGATAATAC-3' . Pairwise comparisons were made between corresponding samples (EB vs. EB+RA) normalized to the control IgG ChIP (background) for each time point. Statistical significance of each result was calculated using a paired, two-tailed t-test. * p < 0.05
Lyon MF: Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature. 1961, 190: 372-373. 10.1038/190372a0.
Payer B, Lee JT: X chromosome dosage compensation: how mammals keep the balance. Annu Rev Genet. 2008, 42: 733-772. 10.1146/annurev.genet.42.110807.091711.
Avner P, Heard E: X-chromosome inactivation: counting, choice and initiation. Nat Rev Genet. 2001, 2: 59-67. 10.1038/35047580.
Lucchesi JC, Kelly WG, Panning B: Chromatin remodeling in dosage compensation. Annu Rev Genet. 2005, 39: 615-651. 10.1146/annurev.genet.39.073003.094210.
Wutz A, Gribnau J: X inactivation Xplained. Curr Opin Genet Dev. 2007, 17: 387-393. 10.1016/j.gde.2007.08.001.
Clemson CM, McNeil JA, Willard HF, Lawrence JB: XIST RNA paints the inactive X chromosome at interphase: evidence for a novel RNA involved in nuclear/chromosome structure. J Cell Biol. 1996, 132: 259-275. 10.1083/jcb.132.3.259.
Penny GD, Kay GF, Sheardown SA, Rastan S, Brockdorff N: Requirement for Xist in X chromosome inactivation. Nature. 1996, 379: 131-137. 10.1038/379131a0.
Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT: Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science. 2008, 322: 750-756. 10.1126/science.1163045.
Plath K, Fang J, Mlynarczyk-Evans SK, Cao R, Worringer KA, Wang H, de la Cruz CC, Otte AP, Panning B, Zhang Y: Role of histone H3 lysine 27 methylation in X inactivation. Science. 2003, 300: 131-135. 10.1126/science.1084274.
Silva J, Mak W, Zvetkova I, Appanah R, Nesterova TB, Webster Z, Peters AH, Jenuwein T, Otte AP, Brockdorff N: Establishment of histone h3 methylation on the inactive X chromosome requires transient recruitment of Eed-Enx1 polycomb group complexes. Dev Cell. 2003, 4: 481-495. 10.1016/S1534-5807(03)00068-6.
Martin GR, Epstein CJ, Travis B, Tucker G, Yatziv S, Martin DW, Clift S, S C: X-chromosome inactivation during differentiation of female teratocarcinoma stem cells in vitro. Nature. 1978, 271: 329-333. 10.1038/271329a0.
Rastan S, Robertson E: X chromosome deletions in embryo-derived (EK) cell lines associated with the lack of X chromosome inactivation. J Embryol Exp Morphol. 1985, 90: 379-388.
Lee JT, Lu N: Targeted mutagenesis of Tsix leads to nonrandom X inactivation. Cell. 1999, 99: 47-57. 10.1016/S0092-8674(00)80061-6.
Sado T, Wang Z, Sasaki H, Li E: Regulation of imprinted X-chromosome inactivation in mice by Tsix. Development. 2001, 128: 1275-1286.
Lee JT: Regulation of X-chromosome counting by Tsix and Xite sequences. Science. 2005, 309: 768-771. 10.1126/science.1113673.
Xu N, Tsai CL, Lee JT: Transient homologous chromosome pairing marks the onset of X inactivation. Science. 2006, 311: 1149-1152. 10.1126/science.1122984.
Bacher CP, Guggiari M, Brors B, Augui S, Clerc P, Avner P, Eils R, Heard E: Transient colocalization of X-inactivation centres accompanies the initiation of X inactivation. Nat Cell Biol. 2006, 8: 293-299. 10.1038/ncb1365.
Ogawa Y, Lee JT: Xite, X-inactivation intergenic transcription elements that regulate the probability of choice. Mol Cell. 2003, 11: 731-743. 10.1016/S1097-2765(03)00063-7.
Vigneau S, Augui S, Navarro P, Avner P, Clerc P: An essential role for the DXPas34 tandem repeat and Tsix transcription in the counting process of X chromosome inactivation. Proc Natl Acad Sci USA. 2006, 103: 7390-7395. 10.1073/pnas.0602381103.
Luikenhuis S, Wutz A, Jaenisch R: Antisense transcription through the Xist locus mediates Tsix function in embryonic stem cells. Mol Cell Biol. 2001, 21: 8512-8520. 10.1128/MCB.21.24.8512-8520.2001.
Clerc P, Avner P: Role of the region 3' to Xist exon 6 in the counting process of X-chromosome inactivation. Nat Genet. 1998, 19: 249-253. 10.1038/924.
Morey C, Navarro P, Debrand E, Avner P, Rougeulle C, Clerc P: The region 3' to Xist mediates X chromosome counting and H3 Lys-4 dimethylation within the Xist gene. Embo J. 2004, 23: 594-604. 10.1038/sj.emboj.7600071.
Monkhorst K, Jonkers I, Rentmeester E, Grosveld F, Gribnau J: X inactivation counting and choice is a stochastic process: evidence for involvement of an X-linked activator. Cell. 2008, 132: 410-421. 10.1016/j.cell.2007.12.036.
Sado T, Li E, Sasaki H: Effect of TSIX disruption on XIST expression in male ES cells. Cytogenet Genome Res. 2002, 99: 115-118. 10.1159/000071582.
Navarro P, Chambers I, Karwacki-Neisius V, Chureau C, Morey C, Rougeulle C, Avner P: Molecular coupling of Xist regulation and pluripotency. Science. 2008, 321: 1693-1695. 10.1126/science.1160952.
Donohoe ME, Silva SS, Pinter SF, Xu N, Lee JT: The pluripotency factor Oct4 interacts with Ctcf and also controls X-chromosome pairing and counting. Nature. 2009, 460: 128-132. 10.1038/nature08098.
Scholer HR, Balling R, Hatzopoulos AK, Suzuki N, Gruss P: Octamer binding proteins confer transcriptional activity in early mouse embryogenesis. Embo J. 1989, 8: 2551-2557.
Okamoto K, Okazawa H, Okuda A, Sakai M, Muramatsu M, Hamada H: A novel octamer binding transcription factor is differentially expressed in mouse embryonic cells. Cell. 1990, 60: 461-472. 10.1016/0092-8674(90)90597-8.
Rosner MH, Vigano MA, Ozato K, Timmons PM, Poirier F, Rigby PW, Staudt LM: A POU-domain transcription factor in early stem cells and germ cells of the mammalian embryo. Nature. 1990, 345: 686-692. 10.1038/345686a0.
Schoorlemmer J, Kruijer W: Octamer-dependent regulation of the kFGF gene in embryonal carcinoma and embryonic stem cells. Mech Dev. 1991, 36: 75-86. 10.1016/0925-4773(91)90074-G.
Allenby G, Bocquel MT, Saunders M, Kazmer S, Speck J, Rosenberger M, Lovey A, Kastner P, Grippo JF, Chambon P, et al: Retinoic acid receptors and retinoid X receptors: interactions with endogenous retinoic acids. Proc Natl Acad Sci USA. 1993, 90: 30-34. 10.1073/pnas.90.1.30.
Morey C, Arnaud D, Avner P, Clerc P: Tsix-mediated repression of Xist accumulation is not sufficient for normal random X inactivation. Hum Mol Genet. 2001, 10: 1403-1411. 10.1093/hmg/10.13.1403.
Sheardown SA, Duthie SM, Johnston CM, Newall AE, Formstone EJ, Arkell RM, Nesterova TB, Alghisi GC, Rastan S, Brockdorff N: Stabilization of Xist RNA mediates initiation of X chromosome inactivation. Cell. 1997, 91: 99-107. 10.1016/S0092-8674(01)80012-X.
Stavropoulos N, Lu N, Lee JT: A functional role for Tsix transcription in blocking Xist RNA accumulation but not in X-chromosome choice. Proc Natl Acad Sci USA. 2001, 98: 10232-10237. 10.1073/pnas.171243598.
Ovitt CE, Scholer HR: The molecular biology of Oct-4 in the early mouse embryo. Mol Hum Reprod. 1998, 4: 1021-1031. 10.1093/molehr/4.11.1021.
Namekawa SH, Payer B, Huynh KD, Jaenisch R, Lee JT: Two-step imprinted X-inactivation: Repeat vs genic silencing in the mouse. Mol Cell Biol.
Clemson CM, Hall LL, Byron M, McNeil J, Lawrence JB: The X chromosome is organized into a gene-rich outer rim and an internal core containing silenced nongenic sequences. Proc Natl Acad Sci USA. 2006, 103: 7688-7693. 10.1073/pnas.0601069103.
Hall LL, Byron M, Sakai K, Carrel L, Willard HF, Lawrence JB: An ectopic human XIST gene can induce chromosome inactivation in postdifferentiation human HT-1080 cells. Proc Natl Acad Sci USA. 2002, 99: 8677-8682. 10.1073/pnas.132468999.
Huynh KD, Lee JT: Inheritance of a pre-inactivated paternal X chromosome in early mouse embryos. Nature. 2003, 426: 857-862. 10.1038/nature02222.
Kastner P, Perez A, Lutz Y, Rochette-Egly C, Gaub MP, Durand B, Lanotte M, Berger R, Chambon P: Structure, localization and transcriptional properties of two classes of retinoic acid receptor alpha fusion proteins in acute promyelocytic leukemia (APL): structural similarities with a new family of oncoproteins. Embo J. 1992, 11: 629-642.
Chambon P: The molecular and genetic dissection of the retinoid signalling pathway. Gene. 1993, 135: 223-228. 10.1016/0378-1119(93)90069-F.
Heyman RA, Mangelsdorf DJ, Dyck JA, Stein RB, Eichele G, Evans RM, Thaller C: 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell. 1992, 68: 397-406. 10.1016/0092-8674(92)90479-V.
Levin AA, Sturzenbecker LJ, Kazmer S, Bosakowski T, Huselton C, Allenby G, Speck J, Kratzeisen C, Rosenberger M, Lovey A, et al: 9-cis retinoic acid stereoisomer binds and activates the nuclear receptor RXR alpha. Nature. 1992, 355: 359-361. 10.1038/355359a0.
Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE, Kakizuka A, Evans RM: Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev. 1992, 6: 329-344. 10.1101/gad.6.3.329.
Yu VC, Delsert C, Andersen B, Holloway JM, Devary OV, Naar AM, Kim SY, Boutin JM, Glass CK, Rosenfeld MG: RXR beta: a coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell. 1991, 67: 1251-1266. 10.1016/0092-8674(91)90301-E.
Bugge TH, Pohl J, Lonnoy O, Stunnenberg HG: RXR alpha, a promiscuous partner of retinoic acid and thyroid hormone receptors. Embo J. 1992, 11: 1409-1418.
Kliewer SA, Umesono K, Heyman RA, Mangelsdorf DJ, Dyck JA, Evans RM: Retinoid X receptor-COUP-TF interactions modulate retinoic acid signaling. Proc Natl Acad Sci USA. 1992, 89: 1448-1452. 10.1073/pnas.89.4.1448.
Leid M, Kastner P, Lyons R, Nakshatri H, Saunders M, Zacharewski T, Chen JY, Staub A, Garnier JM, Mader S, et al: Purification, cloning, and RXR identity of the HeLa cell factor with which RAR or TR heterodimerizes to bind target sequences efficiently. Cell. 1992, 68: 377-395. 10.1016/0092-8674(92)90478-U.
Marks MS, Hallenbeck PL, Nagata T, Segars JH, Appella E, Nikodem VM, Ozato K: H-2RIIBP (RXR beta) heterodimerization provides a mechanism for combinatorial diversity in the regulation of retinoic acid and thyroid hormone responsive genes. Embo J. 1992, 11: 1419-1435.
Pikarsky E, Sharir H, Ben-Shushan E, Bergman Y: Retinoic acid represses Oct-3/4 gene expression through several retinoic acid-responsive elements located in the promoter-enhancer region. Mol Cell Biol. 1994, 14: 1026-1038.
Schoorlemmer J, van Puijenbroek A, van Den Eijnden M, Jonk L, Pals C, Kruijer W: Characterization of a negative retinoic acid response element in the murine Oct4 promoter. Mol Cell Biol. 1994, 14: 1122-1136.
Gu P, LeMenuet D, Chung AC, Mancini M, Wheeler DA, Cooney AJ: Orphan nuclear receptor GCNF is required for the repression of pluripotency genes during retinoic acid-induced embryonic stem cell differentiation. Mol Cell Biol. 2005, 25: 8507-8519. 10.1128/MCB.25.19.8507-8519.2005.
Marin M, Karis A, Visser P, Grosveld F, Philipsen S: Transcription factor Sp1 is essential for early embryonic development but dispensable for cell growth and differentiation. Cell. 1997, 89: 619-628. 10.1016/S0092-8674(00)80243-3.
Sylvester I, Scholer HR: Regulation of the Oct-4 gene by nuclear receptors. Nucleic Acids Res. 1994, 22: 901-911. 10.1093/nar/22.6.901.
Sun BK, Deaton AM, Lee JT: A transient heterochromatic state in Xist preempts X inactivation choice without RNA stabilization. Mol Cell. 2006, 21: 617-628. 10.1016/j.molcel.2006.01.028.
Buzin CH, Mann JR, Singer-Sam J: Quantitative RT-PCR assays show Xist RNA levels are low in mouse female adult tissue, embryos and embryoid bodies. Development. 1994, 120: 3529-3536.
Panning B, Dausman J, Jaenisch R: X chromosome inactivation is mediated by Xist RNA stabilization. Cell. 1997, 90: 907-916. 10.1016/S0092-8674(00)80355-4.
Navarro P, Pichard S, Ciaudo C, Avner P, Rougeulle C: Tsix transcription across the Xist gene alters chromatin conformation without affecting Xist transcription: implications for X-chromosome inactivation. Genes Dev. 2005, 19: 1474-1484. 10.1101/gad.341105.
Ogawa Y, Sun BK, Lee JT: Intersection of the RNA interference and X-inactivation pathways. Science. 2008, 320: 1336-1341. 10.1126/science.1157676.
Wutz A, Jaenisch R: A shift from reversible to irreversible X inactivation is triggered during ES cell differentiation. Mol Cell. 2000, 5: 695-705. 10.1016/S1097-2765(00)80248-8.
Sado T, Hoki Y, Sasaki H: Tsix silences Xist through modification of chromatin structure. Dev Cell. 2005, 9: 159-165. 10.1016/j.devcel.2005.05.015.
Minucci S, Botquin V, Yeom YI, Dey A, Sylvester I, Zand DJ, Ohbo K, Ozato K, Scholer HR: Retinoic acid-mediated down-regulation of Oct3/4 coincides with the loss of promoter occupancy in vivo. Embo J. 1996, 15: 888-899.
Monkhorst K, de Hoon B, Jonkers I, Mulugeta Achame E, Monkhorst W, Hoogerbrugge J, Rentmeester E, Westerhoff HV, Grosveld F, Grootegoed JA, Gribnau J: The probability to initiate X chromosome inactivation is determined by the X to autosomal ratio and X chromosome specific allelic properties. PLoS One. 2009, 4: e5616-10.1371/journal.pone.0005616.
Starmer J, Magnuson T: A new model for random X chromosome inactivation. Development. 2009, 136: 1-10. 10.1242/dev.025908.
Jonkers I, Barakat TS, Achame EM, Monkhorst K, Kenter A, Rentmeester E, Grosveld F, Grootegoed JA, Gribnau J: RNF12 is an X-Encoded dose-dependent activator of X chromosome inactivation. Cell. 2009, 139: 999-1011. 10.1016/j.cell.2009.10.034.
Lee JT: Disruption of imprinted X inactivation by parent-of-origin effects at Tsix. Cell. 2000, 103: 17-27. 10.1016/S0092-8674(00)00101-X.
Panning B, Jaenisch R: DNA hypomethylation can activate Xist expression and silence X-linked genes. Genes Dev. 1996, 10: 1991-2002. 10.1101/gad.10.16.1991.
Livak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001, 25: 402-408. 10.1006/meth.2001.1262.
We would like to thank all members of the Lee lab for stimulating discussion and helpful comments. This work was supported by the National Institutes of Health [RO1-GM58839] to JTL and the Michael and Anna Vranos Graduate Fellowship Fund in the Life Sciences to JYA. JTL is an investigator of the Howard Hughes Medical Institute.
JYA performed all of the experiments. JYA and JTL designed the experiments, analyzed the data, and wrote the paper. All authors read and approved the final manuscript.