The Mbd1-Atf7ip-Setdb1 pathway contributes to the maintenance of X chromosome inactivation
© Minkovsky et al.; licensee BioMed Central Ltd. 2014
Received: 2 March 2014
Accepted: 5 June 2014
Published: 26 June 2014
X chromosome inactivation (XCI) is a developmental program of heterochromatin formation that initiates during early female mammalian embryonic development and is maintained through a lifetime of cell divisions in somatic cells. Despite identification of the crucial long non-coding RNA Xist and involvement of specific chromatin modifiers in the establishment and maintenance of the heterochromatin of the inactive X chromosome (Xi), interference with known pathways only partially reactivates the Xi once silencing has been established. Here, we studied ATF7IP (MCAF1), a protein previously characterized to coordinate DNA methylation and histone H3K9 methylation through interactions with the methyl-DNA binding protein MBD1 and the histone H3K9 methyltransferase SETDB1, as a candidate maintenance factor of the Xi.
We found that siRNA-mediated knockdown of Atf7ip in mouse embryonic fibroblasts (MEFs) induces the activation of silenced reporter genes on the Xi in a low number of cells. Additional inhibition of two pathways known to contribute to Xi maintenance, DNA methylation and Xist RNA coating of the X chromosome, strongly increased the number of cells expressing Xi-linked genes upon Atf7ip knockdown. Despite its functional importance in Xi maintenance, ATF7IP does not accumulate on the Xi in MEFs or differentiating mouse embryonic stem cells. However, we found that depletion of two known repressive biochemical interactors of ATF7IP, MBD1 and SETDB1, but not of other unrelated H3K9 methyltransferases, also induces the activation of an Xi-linked reporter in MEFs.
Together, these data indicate that Atf7ip acts in a synergistic fashion with DNA methylation and Xist RNA to maintain the silent state of the Xi in somatic cells, and that Mbd1 and Setdb1, similar to Atf7ip, play a functional role in Xi silencing. We therefore propose that ATF7IP links DNA methylation on the Xi to SETDB1-mediated H3K9 trimethylation via its interaction with MBD1, and that this function is a crucial feature of the stable silencing of the Xi in female mammalian cells.
X chromosome inactivation (XCI) is an X chromosome-wide process of gene silencing in female mammalian cells that is a mechanism for dosage compensation of X-linked gene products between the sexes. Due to the random nature of XCI, normal adult female mammalian organisms have a roughly equal distribution of cells with an inactive X chromosome (Xi) of maternal or paternal origin . Random XCI is initiated in early embryonic development in pluripotent epiblast cells shortly after the blastocyst implants. Once established, XCI becomes remarkably stable and is maintained for the lifetime of somatic cells, but is reversed in the cells that give rise to the germline [2–5]. Experimentally, the initiation of XCI can be recapitulated in differentiating female mouse embryonic stem cells (ESCs). Notably, despite its extreme stability in differentiated cells in vivo, the reactivation of the Xi can be achieved experimentally by induction of the pluripotent state in somatic cells either through transcription factor-induced reprogramming to induced pluripotent stem cells (iPSCs), cell fusion with ESCs, or somatic cell nuclear transfer [6–8].
XCI begins with the upregulation and spread of the long non-coding RNA Xist on the future Xi [2–5]. Xist, itself encoded on the X chromosome, is absolutely required for the initiation of XCI and its coating of the chromosome initiates a cascade of silencing events including the exclusion of RNA polymerase II from the Xist RNA- coated domain, the gain of histone H3K9me2, loss of active chromatin marks such as histone H3K4 methylation and histone acetylation, the recruitment of polycomb repressive complexes (PRC) 1 and 2 and deposition of their respective histone marks H2AK119ub1 and H3K27me3, and the gain of H4K20me1 [2–5]. After the establishment of the Xi, XCI transitions to the maintenance phase in which the repressed state of the X chromosome is stabilized and permanently locked in. The maintenance phase is marked by incorporation of the histone variant macroH2A, accumulation of promoter CpG methylation, and transition to late replication in S-phase [9–12].
Many of the protein factors implicated in XCI were identified based on their nuclear enrichment on the Xi by immunofluorescence . Notably, one of these Xi-enriched proteins is the matrix protein hnRNP U, which has been shown to be required for the Xist RNA coating of the Xi and the initiation of X-inactivation . Yet, various chromatin modifiers implicated in XCI, including Eed (the structural subunit of PRC2), Ring1b (the E3 ligase of PRC1), G9a (a histone methyltransferase mediating H3K9 methylation), and Dnmt3a/b (the de novo DNA methyltransferases) were found to be dispensable for both initiation and maintenance of X chromosome silencing in mice and in in vitro cell culture systems [15–18]. Therefore, the exact role for some of these factors in random XCI still remains unclear.
Interestingly, in the course of XCI initiation, the cascade of chromatin and transcriptional changes depends on continued Xist expression and remains reversible upon experimentally induced Xist shutdown . In contrast, the maintenance phase of XCI is characterized by almost complete resistance of the Xi to reactivation upon Xist deletion [19, 20]. To explain this switch in Xist dependence with XCI phase, studies in differentiated female cells have described synergism between Xist RNA, DNA methylation, histone variants, and histone hypoacetylation in maintaining XCI [20–22]. For instance, assaying primary mouse embryonic fibroblasts (MEFs) harboring a GFP reporter on the Xi that is subject to X-inactivation, showed reactivation in approximately 11% of cells 13 days after simultaneous deletion of Xist and Dnmt1. The contribution of Xist to silencing is considerably smaller than that of Dnmt1 as Xist deletion alone only doubled the low ‘spontaneous’ background rate of reactivation to 0.05%, while Dnmt1 deletion alone led to approximately 5% reactivation . Thus, multiple epigenetic layers act together to maintain the silenced state of the Xi and Xist retains some role in gene silencing in the maintenance phase that is appreciated only when other repressive modifications are inhibited. Another instance of this cooperative functional role of repressive modifications in XCI involves the histone variant macroH2A as well as Cullin3 and SPOP (both members of an E3 ligase complex that ubiquitinates macroH2A) [21, 22]. Knockdown of any one of these three proteins alone does not induce activation of the Xi-linked GFP reporter, but rates of activation increase when knockdown is sensitized by a DNA demethylating agent and a histone deacetylase inhibitor [21, 22].
In summary, these studies demonstrate that the Xi in somatic cells is relatively resistant to reactivation by interference with single known factors and that seemingly distinct silencing mechanisms act in a combinatorial fashion to ‘lock-in’ the heterochromatin state. Notably, among the chromatin mechanisms tested for a functional role in Xi-maintenance, DNA methylation so far appears to be the most critical pathway for maintaining the silent state of the Xi in differentiated cells . However, one fundamental repressive mechanism, histone H3K9 methylation, also enriches on the Xi but its importance in XCI has not yet been clearly established, raising the question of whether this methylation mark and the enzymes involved in its deposition functionally contribute to Xi-maintenance [23–27]. Furthermore, since in many developmental systems DNA methylation and histone H3K9 methylation pathways can be dependent on one another, an interesting question is whether these two chromatin regulatory mechanisms would also be linked on the Xi .
In this study, we focused on these questions by dissecting the functional contribution of the nuclear protein ATF7IP (also known as MCAF1 for Mbd-1 Chromatin Associated Factor 1) to XCI. We chose to test the function of ATF7IP in XCI for two reasons. First, though described to function as both a transcriptional activator and repressor, ATF7IP is a particularly interesting candidate factor for a regulator of XCI because in its repressive context, it has been shown to mediate MBD1-dependent transcriptional silencing through the recruitment and catalytic activation of the histone H3K9 methyltransferase SETDB1 [29–31]. Since MBD1 (methyl-CpG DNA Binding Domain Protein 1) binds methylated DNA, this function of ATF7IP could thereby bridge the DNA methylation and histone H3K9 methylation pathways on the Xi . Second, ATF7IP is implicated to interact with CDYL, a chromodomain-containing transcriptional co-repressor that is recruited to the Xi by H3K9me2, as ATF7IP was recently identified by mass spectrometry in a CDYL pulldown in undifferentiated and differentiating mouse ESCs .
Our study identifies ATF7IP and its previously described biochemical interactors MBD1 and SETDB1 as regulators of the Xi in differentiated cells. We reveal this function of Atf7ip, Mbd1, and Setdb1 in a ‘sensitized’ approach, where we add low concentrations of 5-aza-2’-dC to Xi-reactivation assays with the rationale of destabilizing DNA methylation and therefore XCI, so that interference with candidate gene function would increase rates of Xi-reactivation in somatic cells.
Results and discussion
Depletion of Atf7ip leads to Xi-reactivation in somatic cells
Recently, the chromodomain-containing transcriptional repressor CDYL was found to associate with the Xi, requiring the induction of differentiation and both histone H3K9me2 and H3K27me3 for its Xi-enrichment . Based on its biochemical interaction with the H3K9 methyltransferase G9A, it was proposed that CDYL anchors G9A to facilitate the propagation and maintenance of the H3K9me2 mark on the Xi and ensure the perpetuation of the H3K9me2 on the Xi during S-phase . Based on the timing of Xi association and its independence of the silencing domain of Xist, it was suggested that CDYL-regulated processes play a role in the maintenance of XCI rather than the initiation of X-inactivation. Interestingly, Heard and colleagues also found ATF7IP (MCAF1) and its homolog ATF7IP2 (MCAF2) as one of 20 high-confidence proteins interacting with CDYL in both undifferentiated and differentiating mouse ESCs , leading us to test the hypothesis that ATF7IP, a protein known to link the DNA methylation and H3K9 methylation pathways in autosomal gene silencing contexts, could also function in XCI.
To monitor the functional effect of Atf7ip knockdown on the maintenance of the silent state of the Xi, we developed a new Xi-linked reporter that is highly sensitive for Xi-reactivation. Specifically, the firefly luciferase open reading frame under control of the constitutive CAG promoter was targeted into the Hprt locus on the single X chromosome in male mouse ESCs using Flp-mediated recombination, an approach similar to one described previously  (Additional file 2: Figure S2). Correct targeting was confirmed by southern blotting (data not shown). Subsequently, transgenic mice were generated by blastocyst injection of a targeted ESC clone. Upon germline transmission, skewing of XCI to the luciferase-bearing X chromosome was ensured by breeding male reporter mice with female mice heterozygous for an Xist knockout (ΔXist) allele . It has been previously demonstrated that the loss of Xist on one X chromosome in female mice leads to inactivation of the wild-type X chromosome in 100% of the cells, thereby resulting in the Xist-deleted X chromosome remaining active (the Xa) in all cells .
As expected, XiCAG-LuciferaseXaΔXist MEFs derived from these animals (carrying the luciferase reporter on the Xi in all cells) did not show luciferase expression in normal culture conditions (Additional file 3: Figure S3Ai, untreated condition). By contrast, female MEFs carrying wild-type Xist on both X chromosomes and heterozygous for CAG-luciferase in the Hprt locus displayed strong luciferase expression due to random XCI (Additional file 3: Figure S3Aii). These results indicate that our new reporter is subject to developmental XCI and becomes efficiently silenced when located on the Xi. In agreement with this finding, the luciferase reporter gene including its promoter was found to be highly methylated on the Xi in XiCAG-LuciferaseXaΔXist MEFs as determined by reduced-representation bisulfite sequencing (Additional file 4: Figure S4A, B). Luciferase expression in XiCAG-LuciferaseXaΔXist MEFs increased when the DNA methylation machinery was impaired by either depleting Dnmt1 with siRNAs or treating cells with the demethylating agent 5-aza-2’-dC (Figure 1Di, Additional file 3: Figure S3Ai). Accordingly, levels of DNA methylation at the luciferase reporter decreased upon treatment with a high concentration of 5-aza-2’-dC (10 uM) (Additional file 4: Figure S4B, C). Thus, the luciferase reporter in the Hprt locus on the Xi can be activated efficiently when a known pathway of Xi maintenance is impaired. Therefore, we used these fibroblasts to examine the role of Atf7ip in XCI.
We found that Atf7ip knockdown alone yielded weak activation of the luciferase reporter in XiCAG-LuciferaseXaΔXist MEFs after three days of knockdown (at 72 h of treatment) (Figure 1Di). While this effect was often seen, it did not consistently pass significance tests. Since it is known that various repressive mechanisms synergistically act to maintain the silent state of the Xi, we reasoned that combining Atf7ip knockdown with conditions of DNA demethylation may unmask an effect of Atf7ip depletion . We first used the demethylating agent 5-aza-2’-dC, which inhibits DNMT1 and impedes propagation of DNA methylation through cell division . On its own, 5-aza-2’-dC had a dose-dependent activity as increasing concentration induced higher rates of Xi-reactivation based on the luciferase reporter assay (Additional file 3: Figure S3Ai). In our Atf7ip knockdown experiments (Figure 1Di), we chose to add 0.2 uM 5-aza-2’-dC, since that amount of 5-aza-2’-dC induced only very little luciferase activity on its own (Additional file 3: Figure S3Ai). When combined with this low level of 5-aza-2’-dC, luciferase reporter reactivation due to Atf7ip knockdown was reproducibly enhanced up to 50-fold (Figure 1Di). To confirm that the Xi-reactivation effect of 5-aza-2’-dC treatment on Atf7ip knockdown was indeed due to changes in DNA methylation and not through potential off-target effects of 5-aza-2’-dC, we also depleted Dnmt1, the main enzyme responsible for the maintenance of DNA methylation, with siRNAs in place of 5-aza-2’-dC treatment (Additional file 3: Figure S3B). Again, we observed strong enhancement in reporter activation when Atf7ip was depleted concurrently with Dnmt1, while knockdown of either Atf7ip or Dnmt1 alone induced very low levels of luciferase activity (Figure 1Di). Addition of low levels of 5-aza-2’-dC did not significantly alter the efficacy of Atf7ip knockdown as assayed by western blot (compare Figure 1B, C to Additional file 1: Figure S1A, B). Therefore, we conclude that depletion of Atf7ip on its own has a limited effect on the Xi, but greatly induces Xi-reactivation when the DNA methylation maintenance pathway is impaired concomitantly. These data reveal a synergistic effect on the Xi between ATF7IP and the DNA methylation pathway.
We recapitulated these results with MEFs carrying another newly generated fluorescence reporter (histone H2B fused to citrine - H2B-citrine) in the same locus as the luciferase reporter (Hprt) (Figure 1Dii), and with MEFs carrying a previously published randomly integrated GFP transgene near the centromere on the X chromosome (Figure 1Diii) [20, 37], allowing us to quantify the number of cells with Xi-reactivation by FACS. In both cases, depletion of only Atf7ip induced reporter activation in a low but significantly higher number of cells than control treatments. Inhibition of DNA methylation either with a low concentration 5-aza-2’-dC or by knockdown of Dnmt1 dramatically increased this effect. Taken together, the activity of Atf7ip depletion against three Xi reporters strongly supports a role for Atf7ip in the maintenance of the Xi.
We noted that reactivation in response to knockdown of Atf7ip and DNA methylation inhibition occurred in approximately 12% of GFP-reporter MEFs and in only approximately 2% of H2B-citrine Xi-reporter MEFs based on flow cytometry measurements (Figure 1Dii/iii), which is potentially due to different levels of DNA methylation, other chromatin marks, and/or their integration sites. Notably, we also observed that the X-GFP reporter, when located on the Xa, was expressed at a uniformly high intensity across the cell population while, upon reactivation from the Xi, the GFP intensity was much less uniform and ranged from low to high expression levels (Additional file 5: Figure S5A, compare blue cell population). The H2B-citrine Xi-reporter displayed a similar gradient of reactivation levels upon treatment with Atf7ip knockdown and 5-aza-2’-dC (Additional file 5: Figure S5B). We believe that this gradient in fluorescence intensity under reactivation conditions is a reflection of partial reactivation of the Xi-linked reporter transgenes that is different in each cell, likely owing to incomplete demethylation of the reporters and the interaction with persistent silencing chromatin marks.
Given the different extent of reactivation observed with our two fluorescent reporter MEF lines and the cell-to-cell heterogeneity in Xi fluorescent reporter intensity, we next set out to determine whether Atf7ip knockdown also influenced the expression of an endogenously encoded X-linked gene that is subject to XCI. Specifically, we monitored the activation of Atrx by RNA fluorescence in situ hybridization (FISH) relative to the Xist RNA cloud under reactivation treatment conditions. In cells of normal ploidy, we noticed rare nuclei with two spots of nascent Atrx transcripts with one of the two spots being in proximity of the Xist RNA cloud (Figure 1E). Quantification of the FISH data suggests a rate of Atrx reactivation of 0.5% to 1.0% in response to high concentration 5-aza-2’-dC or low 5-aza-2’-dC combined with Atf7ip knockdown (Figure 1F), which contrasts the rates of Xi-reactivation of 2% to 12% that was observed with the fluorescent reporters. However, we note that the rare numbers of cells with Atrx reactivation on the Xi were not consistently observed across independent treatments and FISH counts. While the FISH data show a trend towards reactivation of an endogenous gene on the Xi in response to Atf7ip knockdown, we interpret the lower magnitude of reactivation as compared to the Xi-reporter approach as stemming from poorer sensitivity of FISH compared to reporter assays. We believe that almost complete reactivation of Atrx would be required to allow the detection of a Atrx signal at the site of transcription on the Xi with our FISH methodology, but is likely not attained upon our treatments . This conclusion is supported by the limited extent of reactivation of our fluorescent reporters described above (Additional file 5: Figure S5).
ATF7IP does not accumulate on the Xi
We next evaluated whether ATF7IP accumulates at later stages of XCI, by analyzing its nuclear pattern in female differentiating ESCs and female MEFs. In female ESCs induced to differentiate with retinoic acid for five days, the majority of H3K27me3/Xist RNA-demarcated X chromosomes was coated with H3K9me2 (as described in ), yet again no ATF7IP enrichment was observed (Figure 2B). Similarly, in female MEFs, where a smaller proportion of H3K27me3-demarcated inactive X chromosomes have H3K9me2 enrichment, ATF7IP displayed a nuclear staining pattern without any specific Xi enrichment (Figure 2C).
In summary, these data indicate that ATF7IP does not specifically accumulate on the Xi in any of the tested cellular systems, however this does not exclude the possibility that ATF7IP is recruited to the Xi. It is possible that ATF7IP can fulfill its critical function in the maintenance phase of XCI as part of a dynamic interaction with the Xi that cannot be captured by immunostaining approaches. Alternatively, these data may suggest that ATF7IP functions on the Xi without being actively recruited to the Xi by Xist RNA.
ATF7IP’s function in XCI is independent of Xist
ATF7IP was previously described to function not only as transcriptional repressor but also as transcriptional activator, as in addition to the repressive chromatin factors SETDB1 and MBD1, it was also found to interact and function with general transcription factors, RNA polymerase II, and the transcriptional activator SP1 [42, 43]. Here, we considered the possibility that ATF7IP could directly control Xist expression, particularly because the effect of Atf7ip depletion alone (as described in Figure 1) mirrors that of Xist loss: Xist deletion alone only induces very minimal Xi-reactivation, but significantly enhances Xi-reactivation caused by DNA demethylation . Therefore, it could be possible that ATF7IP functions as an activator of Xist in the context of XCI. In this scenario, depletion of Atf7ip could reduce Xist expression and coating of the chromosome by the RNA, which in turn could enhance Xi-reactivation in the presence of DNA methylation inhibitors.
Since these data suggested that Atf7ip does not control Xist expression or coating, we next determined the consequences of combined knockdown of Atf7ip and deletion of Xist in MEFs on the Xi-linked H2B-citrine reporter. We hypothesized that if ATF7IP is acting in the capacity of a transcriptional repressor directly on Xi linked genes, we would see an increase in the number of cells with Xi-reactivation when both Atf7ip and Xist are inhibited compared to the deletion of only one of these factors. To measure the relative contributions of Xist and Atf7ip to Xi-maintenance, we employed female MEFs carrying the histone H2B-Citrine reporter in the Hprt locus in cis to a conditional allele of the Xist gene on the Xi and assayed reactivation by FACS (Figure 3C). Xist was deleted in half of these MEFs by infection with a Cre-recombinase-encoding adenovirus (Ad-Cre), while the other half was treated with control adenovirus (Ad-Null) (Figure 3C). Efficient loss of Xist coating in almost all of the Cre-treated cells was confirmed by FISH (Figure 3D). MEFs positive and negative for Xist, respectively, were then subjected to Atf7ip knockdown and 5-aza-2’-dC treatment for 72 h to assess the contribution of Xist to Xi silencing under these conditions (Figure 3C, E).
Consistent with a previous report, Xist deletion led to a roughly two-fold increase in the number of cells displaying Xi-reactivation in MEFs treated with 5-aza-2’-dC at a high concentration (10 uM) (Figure 3E) , confirming the role of Xist in Xi-maintenance. Reporter activation was also enhanced by Xist deletion when MEFs were depleted for Atf7ip, both with and without 5-aza-2’-dC at low concentration (0.2 um) (Figure 3E). Furthermore, with approximately 14% cells being positive for histone H2B-citrine expression at 72 h of treatment, maximal reactivation of the H2B-citrine reporter was obtained when 10 uM (high) 5-aza-2’-dC, siAtf7ip, and loss of Xist were combined (Figure 3E). These findings further validate ATF7IP as a critical regulator of the Xi in somatic cells. Importantly, these observations argue against a role for ATF7IP in transcriptional activation of Xist and strongly point to an independent role in the repression of Xi-linked genes.
ATF7IP’s repressive binding partners MBD1 and SETDB1 function in XCI
To further investigate whether the Mbd1-Atf7ip-Setdb1 pathway is a specific gene silencing mechanism that is involved in the maintenance of XCI, and to examine whether other related silencing factors also have a role in XCI, we screened a selected panel of siRNAs validated to target other mediators of transcriptional silencing including heterochromatin proteins (Cbx3 and Cbx5) and other H3K9 methyltransferases (Ehmt2 (G9a) and Suv39h1) (Figure 4C) . Despite efficient knockdown (Additional file 7: Figure S7B), our data do not implicate any of these additional factors in the maintenance of XCI even when tested with 5-aza-2’-dC or siDnmt1 sensitization (Figure 4C). This finding supports the conclusion that the three factors Mbd1, Atf7ip, and Setdb1 play a specific role in XCI.
Taken together, our results suggest that DNA methylation, the Mbd1-Atf7ip-Setdb1 pathway, and Xist RNA cooperate in maintaining the silent state of Xi-linked genes in differentiated female cells.
In summary, we show that depletion of Atf7ip or of its known biochemical interactors Mbd1 and Setdb1 leads to Xi-reactivation. We also assayed the interaction of Atf7ip with two of the known major mediators of Xi-maintenance, Xist expression and DNA methylation. Our data reveal a strongly enhanced reactivation of the Xi when all these mechanisms are inhibited and implicate the Mbd1-Atf7ip-Setdb1 pathway, described to couple DNA methylation and histone 3 lysine 9 trimethylation, in the maintenance of XCI in differentiated cells.
It remains unclear from our data whether ATF7IP is necessary for XCI in vivo, as, to our knowledge, no knockout mouse model has been described yet. Atf7ip’s pattern of expression is consistent with a role in XCI maintenance as in situ hybridization shows that it is ubiquitously expressed in d9.5 embryos with more specific tissue distribution at later time points . Notably, ATF7IP’s repressive partners MBD1 and SETDB1 have very different knockout phenotypes, namely mild spatial learning defects and autoimmune disease in adult mice lacking Mbd1, and early embryonic lethality at 3.5 to 5.5 days post conception for the Setdb1 deletion [48–50]. Given these different knockout phenotypes, it would be interesting to see whether Atf7ip tethers the two proteins, one binding methylated DNA and one mediating histone H3K9 trimethylation, only in specific developmental contexts, such as XCI. Notably, the effect of Atf7ip, Setdb1, or Mbd1 depletion on the Xi was more clearly unmasked when other repressive mechanisms were also inhibited, suggesting that the knockout animals may not reveal problems with XCI due to redundancy of silencing mechanisms acting on the Xi.
The identification of Atf7ip as a factor involved in XCI maintenance suggests that the downstream effect of its binding partner Setdb1, namely enzymatic conversion of H3K9me2 to H3K9me3, may also be required on the Xi. H3K9 methylation has previously been reported to play a role in XCI on the basis of enrichment on the Xi detected with a pan-methyl H3K9 antibody in mouse and human cells, though it is unclear if the antibody used in these studies had cross-reactivity to H3K27me3 [23–25]. Other studies comparing H3K27me3 and H3K9me2 by ChIP and by immunostaining established that both marks are increased on the Xi in MEFs relative to the active X [26, 27] (Figure 2). In human cells, the Xi has an H3K9me3 domain distinct from the Xist and H3K27me3-enriching territory of the Xi . Interaction of these domains is proposed to occur through the HBiX protein bridging HP1, a protein binding HK9me3, and SMCHD1, a protein required for methylation of CpG islands on the Xi and recruited by Xist to the H3K27me3 domains . This bridging of H3K27me3 and H3K9me3 domains is thought to be necessary for Xi compaction, however, depletion of HBiX or SMCHD1 by siRNA did not lead to upregulation of Xi-linked genes , leaving an open question as to whether H3K9me3 has silencing function in XCI and whether that function is dependent on HP1 and HBiX binding.
It remains unclear whether knockdown of Setdb1, though a known histone H3K9 trimethyltransferase, would deplete H3K9me3 on the Xi, since the Setdb1 knockout has not been shown to lead to global alterations in H3K9me3 levels nor DNA methylation in early embryonic mouse cells when assayed by immunostaining and bisulfite sequencing of IAP elements, respectively . In addition, the H3K9 methyl reader heterochromatin proteins Cbx1, Cbx3, and/or Cbx5 are thought to mediate transcriptional repression downstream to H3K9 methylation, yet, in our study, knockdown of Cbx3 or Cbx5 individually had no effect on the reactivation of the Xi in somatic cells (Figure 4C). Furthermore, we did not see strong Xi enrichment of Atf7ip or H3K9me3 by immunostaining (Figure 2, Additional file 8: Figure S8). However, we observed that in a fraction of cells a subdomain of the H3K27me3-demarcated Xi co-localizes with a punctum of H3K9me3 enrichment (Additional file 8: Figure S8). ChIP-seq studies of H3K9me3 and the analysis of DNA methylation patterns with and without knockdown of Atf7ip, Setdb1, and Mbd1 will be necessary to refine the role of the Atf7ip/Setdb1/Mbd1 pathway on the Xi and to determine whether genes on the Xi have different propensities for regulation through this pathway. To explain the combinatorial effect of DNA methylation with interference of the Mbd1-Atf7ip-Setdb1 pathway, we hypothesize that some DNA methylation loss is necessary to ‘unlock’ the stably silenced chromatin. However, the relationship between H3K9me3 and DNA methylation may be reciprocal and H3K9me3 could also recruit the DNA methylation machinery .
Finally, ATF7IP has also been shown to interact with CDYL, a protein implicated in the maintenance of H3K9me2 on the Xi, which requires H3K27me3 and H3K9me2 for its Xi accumulation on the Xi . Therefore, Atf7ip may contribute to an extensive regulatory network that links various repressive chromatin pathways to contribute to the maintenance of the Xi.
Cell culture methods
Mouse ESCs were grown on irradiated DR4 MEFs in standard media (DMEM supplemented with 15% FBS, nonessential amino acids, L-glutamine, penicillin-streptomycin, β-mercaptoethanol, and 1000 U/mL LIF). For induction of differentiation, cells were feeder-depleted for 45 min and subsequently plated at a density of 5.0 × 104 cells/ 6-well in MEF media (same as ESC media except that 10% FBS were used and LIF was omitted). One day later, MEF media was supplemented with 1 μM all-trans retinoic acid (Sigma) and refreshed every 2 days, for a total of 5 days. Male ESCs carrying the tet-inducible promoter in front of the endogenous Xist allele were induced to differentiate with 1 μg/mL doxycycline in standard ESC media for 21 h . MEFs were cultured in MEF media and derived at embryonic day 14.5 from timed matings of mice with the appropriate genotypes (see below).
Generation of X-linked reporter MEFs
Hprt-CAG-Luciferase and Hprt-CAG-H2B-citrine reporters were engineered in male V6.5 ESCs. These ESCs were modified by two-step targeting as described in Additional file 2: Figure S2, closely following a strategy originally described for the Col1A locus in . First, an FRT-flanked neomycin resistance cassette and a hygromycin resistance gene that has an FRT site embedded in its 5’coding region but lacks a promoter and an ATG initiation codon, were placed into Hprt locus by homologous recombination, deleting 1,313 bp of Hprt sequence between homology arms including the third exon of Hprt. To place the ‘FRT-neo-FRT-hygro’ cassette into the Hprt locus, the ColA1 homology arms of the Col1A-FRT-hygro-pA ‘homing’ vector described in  were replaced with homology regions targeting the Hprt locus. The Hprt-FRT-hygro-pA ‘homing’ vector was introduced into V6.5 ESCs by electroporation followed by selection with 350 ug/mL G418. DNA from picked clones was analyzed for proper targeting by southern blotting using a BglII digest as outlined in Additional file 2: Figure S2. Second, using FLPe-mediated recombination, the CAG-luciferase or CAG-H2B-citrine transgene was introduced into one ESC clone carrying the Hprt- FRT hygro-pA ‘homing’ site, specifically into the FRT site in front of the hygromycin open reading frame, using the plasmids pBS32-CAG-luciferase or pBS32-CAG-H2B-citrine as shown in Additional file 2: Figure S2. Co-electroporation of these plasmids with a plasmid expressing Flpe recombinase was followed by hygromycin treatment selecting for ESC clones that have the pgk promoter driving the hygromycin resistance gene due to integration of the entire pBS32 plasmid sequence. DNA from selected clones was digested with BglII and screened by southern blotting using 3’ external probe for correct FLP recombination (Additional file 2: Figure S2, data not shown). The pBS32 vector was made by exchanging the tetracycline-responsive operating binding sequence in the pgkATGfrt vector described in  with the constitutive CAG promoter. Luciferase was introduced into pBS32 by Gateway® cloning (Life Technologies). To make pBS32 vector compatible for Gateway cloning (pBS32-GW), a Gateway cassette with attR sites and a ccdB gene was flanked by SgrAI restriction sites and ligated into a unique EcoRI site on pBS32. One-step BP and LR cloning was performed with pDonr221 entry vector, attB-primer amplified firefly luciferase from pGL3 vector (Promega) and pBS32-GW. The pBS32-CAG-H2B-citrine was constructed analogously using a H2B-citrine template kindly provided by Michael Elowitz (Caltech). Luciferase and H2B-citrine reporter ESCs were microinjected by the UCLA transgenic core into C57BL/6 blastocysts to produce chimeric mice following standard procedures. High agouti coat color male chimeras were bred with C57BL/6 females for germline transmission. For skewing of XCI to the luciferase-bearing X chromosome, X-linked reporter male mice were bred with female mice heterozygous for an Xist knockout (ΔXist) allele . For the experiments requiring simultaneous deletion of Xist and reporter quantification, 2lox Xist and Hprt-H2B-Citrine alleles were bred until a spontaneous recombination event brought the alleles in cis. The recombinant males were subsequently bred to ΔXist females to skew X inactivation to the reporter-bearing X chromosome in the embryos. All animal experiments were in accordance with the legislation of the UCLA Animal Research Committee.
Knockdown, overexpression, and chemical treatments of MEFs
For Xi-reactivation assays, MEFs at passage 1 or 2 post-derivation were seeded at a density of 60,000 cells per 12-well well and chemicals in MEF media were added and incubated for 72 h before further analysis (by FISH, immunostaining, or reporter assays). 5-aza-2’-dC (Sigma A3656) was dissolved in DMSO and added at indicated concentrations. Across all conditions in a given experiment, final DMSO concentration (always below 0.1%) and final volumes of media were kept constant across samples. Knockdown with siRNA was performed by reverse transfection at 25 nM final concentration of siRNA. Briefly, a cell suspension was added to a pre-incubated mixture of Lipofectamine RNAimax, 100 uL of reduced serum Opti-MEM media, and siRNA. siRNAs used in this study included Atf7ip (Ambion, AM16706), Dnmt1 (Ambion, AM161526), Mbd1 (Dharmacon, MU-056829-01), Setdb1 (Dharmacon, D-040815-04), Suv39h1 (Dharmacon MU-046141-10), Ehmt2 (Dharmacon, MU-053728-03), Cbx3 (Dharmacon MU-044218-02), Cbx5 (Dharmacon MU-040799-02), Luciferase (Dharmacon, D-001210-02), Scramble (Ambion, AM4636), and GFP (Dharmacon, P-002048-01). For experiments involving multiple knockdowns, control siRNA (against either Scramble, GFP, or Luciferase depending upon experiment as indicated in figure legends) was added to equalize the final siRNA concentration across all conditions. For overexpression of Flag-tagged Atf7ip, the Atf7ip cDNA was introduced into the pMX retroviral vector by In-Fusion® cloning (Clontech) of the Atf7ip cDNA and incorporation of the FLAG tag by PCR at the N-terminal end. The pMX retrovirus was generated in transfected platE cells, and MEFs were transduced as previously described .
Chemical and/or siRNA treatments of MEFs were performed in triplicate 12-well wells for 72 h and cells were lysed with 200 uL passive lysis buffer (PLB, Promega) for 20 min at room temperature on an orbital shaker. Lysates were cleared by 30 s of centrifugation at 12,000 × g and 20 uL were assayed for luciferase activity with 50 uL of LARI reagent (Promega) on a GloMax microplate luminometer (Promega).
Cells were harvested from a 6-well format in Trizol (Invitrogen) and RNA purification was performed with the RNeasy kit (Qiagen) according to manufacturer’s instructions with on-column DNAse treatment (Qiagen). cDNA was prepared using SuperScript III (Invitrogen) with random hexamers and RT-qPCR was performed using a Stratagene Mx3000 thermocycler with primers listed in Additional file 9: Table S1. Results were normalized to Gapdh by the ΔCt method.
Western blot analysis
Xi-luciferase MEFs were treated as described above and one well was used to confirm luciferase reactivation. Cells in the other well were lysed in RIPA buffer and sonicated (30 s on, 30 s off, cycled for 5 min). Equal volume from each sonicated lysate was loaded onto a 4% to 12% 10-well 1.5 mm Bis-Tris gel (NuPAGE). Proteins were transferred onto a nitrocellulose membrane overnight at 10 V in 4°C. Membranes were incubated overnight at 4°C with 1:2,500 dilution of polyclonal rabbit-anti-Atf7ip antibody (Nakao lab). Monoclonal mouse anti-alpha-Tubulin (Calbiochem CP06) was used as loading control at 1:1,000 dilution overnight at 4°C. Quantification was performed on the Odyssey scanner (Li-Cor) with Image Studio software (Li-Cor) and band intensity was normalized to alpha-Tubulin loading control.
Immunofluorescence, FISH, and chromosome paint analysis
Cells were plated on glass coverslips, washed once with PBS, and fixed for 10 min in 4% paraformaldehyde . Immunostaining with antibodies against ATF7IP (Abcam 84497), H3K27me3 (Active Motif 39155), H3K9me2 (Cosmo Bio MCA-MABI0007-100-EX), H3K9me3 (Active Motif 39161, AbCam ab8898), Ash2L (Bethyl a300-107a), FLAG M2 (Sigma F3165), GFP (Nacalai USA Inc., 04404–84), and FISH with DNA probes against Xist and Atrx and IF/FISH combinations thereof were performed as previously reported and mounted with Prolong Gold reagent with DAPI . BAC templates for Atrx FISH probes were obtained from the BACPAC resource (Atrx RP23-265D6).
Cells were trypsinized, washed in PBS, loaded through cell strainer caps (BD Biosciences) and analyzed on a FACSDiva machine (BD Biosciences) with FlowJo software (Tree Star, Inc.).
We thank Dr. Mitsuyoshi Nakao of Kumamoto University for kindly providing ATF7IP antibody. We thank Ryan Schmidt, Amy Pandya-Jones, and Vincent Pasque for experimental advice and reagent sharing. We are also grateful to Caroline Beard and Konrad Hochedlinger for help with the Hprt targeting strategy, and the UCLA transgenic core for blastocyst injection. KP is supported by the NIH (DP2OD001686 and P01 GM099134), CIRM (RN1-00564, RB3-05080, and RB4-06133), the Jonsson Comprehensive Cancer Center and the Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research at UCLA. AM is supported by National Research Service Award AG039179, and SP by a CIRM training grant TG2-01169. KP and AM were supported for this work by funds from the Iris Cantor-UCLA Women’s Health Center Executive Advisory Board. Funding bodies played no role in design, collection, analysis, or interpretation of data, nor in the writing of manuscript or decision to submit manuscript for publication.
- Wu H, Luo J, Yu H, Rattner A, Mo A, Wang Y, Smallwood PM, Erlanger B, Wheelan SJ, Nathans J: Cellular resolution maps of X Chromosome inactivation: implications for neural development, function, and disease. Neuron. 2014, 81: 103-119.PubMed CentralView ArticlePubMedGoogle Scholar
- Brockdorff N: Chromosome silencing mechanisms in X-chromosome inactivation: unknown unknowns. Development. 2011, 138: 5057-5065.View ArticlePubMedGoogle Scholar
- Pollex T, Heard E: Recent advances in X-chromosome inactivation research. Curr Opin Cell Biol. 2012, 24: 825-832.View ArticlePubMedGoogle Scholar
- Lee JT: Gracefully ageing at 50, X-chromosome inactivation becomes a paradigm for RNA and chromatin control. Nat Rev Mol Cell Biol. 2011, 12: 815-826.View ArticlePubMedGoogle Scholar
- Barakat TS, Gribnau J: X chromosome inactivation in the cycle of life. Development. 2012, 139: 2085-2089.View ArticlePubMedGoogle Scholar
- Eggan K, Akutsu H, Hochedlinger K, Rideout W, Yanagimachi R, Jaenisch R: X-Chromosome inactivation in cloned mouse embryos. Science. 2000, 290: 1578-1581.View ArticlePubMedGoogle Scholar
- Maherali N, Sridharan R, Xie W, Utikal J, Eminli S, Arnold K, Stadtfeld M, Yachechko R, Tchieu J, Jaenisch R, Plath K, Hochedlinger K: Directly reprogrammed fibroblasts show global epigenetic remodeling and widespread tissue contribution. Cell Stem Cell. 2007, 1: 55-70.View ArticlePubMedGoogle Scholar
- Tada M, Takahama Y, Abe K, Nakatsuji N, Tada T: Nuclear reprogramming of somatic cells by in vitro hybridization with ES cells. Curr Biol. 2001, 11: 1553-1558.View ArticlePubMedGoogle Scholar
- Gendrel A-V, Apedaile A, Coker H, Termanis A, Zvetkova I, Godwin J, Tang YA, Huntley D, Montana G, Taylor S, Giannoulatou E, Heard E, Stancheva I, Brockdorff N: Smchd1-dependent and -independent pathways determine developmental dynamics of CpG island methylation on the inactive X chromosome. Dev Cell. 2012, 23: 265-279.PubMed CentralView ArticlePubMedGoogle Scholar
- Blewitt ME, Gendrel A-V, Pang Z, Sparrow DB, Whitelaw N, Craig JM, Apedaile A, Hilton DJ, Dunwoodie SL, Brockdorff N, Kay GF, Whitelaw E: SmcHD1, containing a structural-maintenance-of-chromosomes hinge domain, has a critical role in X inactivation. Nat Genet. 2008, 40: 663-669.View ArticlePubMedGoogle Scholar
- Casas-Delucchi CS, Brero A, Rahn H-P, Solovei I, Wutz A, Cremer T, Leonhardt H, Cardoso MC: Histone acetylation controls the inactive X chromosome replication dynamics. Nat Commun. 2011, 2: 222.PubMed CentralView ArticlePubMedGoogle Scholar
- Costanzi C, Pehrson JR: Histone macroH2A1 is concentrated in the inactive X chromosome of female mammals. Nature. 1998, 393: 599-601.View ArticlePubMedGoogle Scholar
- Heard E: Recent advances in X-chromosome inactivation. Curr Opin Cell Biol. 2004, 16: 247-255.View ArticlePubMedGoogle Scholar
- Hasegawa Y, Brockdorff N, Kawano S, Tsutui K, Tsutui K, Nakagawa S: The matrix protein hnRNP U is required for chromosomal localization of Xist RNA. Dev Cell. 2010, 19: 469-476.View ArticlePubMedGoogle Scholar
- Sado T, Okano M, Li E, Sasaki H: De novo DNA methylation is dispensable for the initiation and propagation of X chromosome inactivation. Development. 2004, 131: 975-982.View ArticlePubMedGoogle Scholar
- Kalantry S, Magnuson T: The Polycomb group protein EED is dispensable for the initiation of random X-chromosome inactivation. PLoS Genet. 2006, 2: e66.PubMed CentralView ArticlePubMedGoogle Scholar
- Leeb M, Wutz A: Ring1B is crucial for the regulation of developmental control genes and PRC1 proteins but not X inactivation in embryonic cells. J Cell Biol. 2007, 178: 219-229.PubMed CentralView ArticlePubMedGoogle Scholar
- Ohhata T, Tachibana M, Tada M, Tada T, Sasaki H, Shinkai Y, Sado T: X-inactivation is stably maintained in mouse embryos deficient for histone methyl transferase G9a. Genesis N Y N 2000. 2004, 40: 151-156.Google Scholar
- Wutz A, Jaenisch R: A shift from reversible to irreversible X inactivation is triggered during ES cell differentiation. Mol Cell. 2000, 5: 695-705.View ArticlePubMedGoogle Scholar
- Csankovszki G, Nagy A, Jaenisch R: Synergism of Xist RNA, DNA methylation, and histone hypoacetylation in maintaining X chromosome inactivation. J Cell Biol. 2001, 153: 773-784.PubMed CentralView ArticlePubMedGoogle Scholar
- Hernández-Muñoz I, Lund AH, van der Stoop P, Boutsma E, Muijrers I, Verhoeven E, Nusinow DA, Panning B, Marahrens Y, van Lohuizen M: Stable X chromosome inactivation involves the PRC1 Polycomb complex and requires histone MACROH2A1 and the CULLIN3/SPOP ubiquitin E3 ligase. Proc Natl Acad Sci U S A. 2005, 102: 7635-7640.PubMed CentralView ArticlePubMedGoogle Scholar
- Nusinow DA, Hernández-Muñoz I, Fazzio TG, Shah GM, Kraus WL, Panning B: Poly(ADP-ribose) Polymerase 1 Is Inhibited by a Histone H2A Variant, MacroH2A, and Contributes to Silencing of the Inactive X Chromosome. J Biol Chem. 2007, 282: 12851-12859.View ArticlePubMedGoogle Scholar
- Boggs BA, Cheung P, Heard E, Spector DL, Chinault AC, Allis CD: Differentially methylated forms of histone H3 show unique association patterns with inactive human X chromosomes. Nat Genet. 2002, 30: 73-76.View ArticlePubMedGoogle Scholar
- Heard E, Rougeulle C, Arnaud D, Avner P, Allis CD, Spector DL: Methylation of histone H3 at Lys-9 is an early mark on the X chromosome during X inactivation. Cell. 2001, 107: 727-738.View ArticlePubMedGoogle Scholar
- Mermoud JE, Popova B, Peters AHFM, Jenuwein T, Brockdorff N: Histone H3 lysine 9 methylation occurs rapidly at the onset of random X chromosome inactivation. Curr Biol. 2002, 12: 247-251.View ArticlePubMedGoogle Scholar
- Rougeulle C, Chaumeil J, Sarma K, Allis CD, Reinberg D, Avner P, Heard E: Differential Histone H3 Lys-9 and Lys-27 Methylation Profiles on the X Chromosome. Mol Cell Biol. 2004, 24: 5475-5484.PubMed CentralView ArticlePubMedGoogle Scholar
- Escamilla-Del-Arenal M, da Rocha ST, Spruijt CG, Masui O, Renaud O, Smits AH, Margueron R, Vermeulen M, Heard E: Cdyl, a new partner of the inactive X Chromosome and potential reader of H3K27me3 and H3K9me2. Mol Cell Biol. 2013, 33: 5005-5020.PubMed CentralView ArticlePubMedGoogle Scholar
- Cedar H, Bergman Y: Linking DNA methylation and histone modification: patterns and paradigms. Nat Rev Genet. 2009, 10: 295-304.View ArticlePubMedGoogle Scholar
- Ichimura T, Watanabe S, Sakamoto Y, Aoto T, Fujita N, Nakao M: Transcriptional repression and heterochromatin formation by MBD1 and MCAF/AM family proteins. J Biol Chem. 2005, 280: 13928-13935.View ArticlePubMedGoogle Scholar
- Wang H, An W, Cao R, Xia L, Erdjument-Bromage H, Chatton B, Tempst P, Roeder RG, Zhang Y: mAM facilitates conversion by ESET of dimethyl to trimethyl lysine 9 of histone H3 to cause transcriptional repression. Mol Cell. 2003, 12: 475-487.View ArticlePubMedGoogle Scholar
- Fujita N, Watanabe S, Ichimura T, Ohkuma Y, Chiba T, Saya H, Nakao M: MCAF mediates MBD1-dependent transcriptional repression. Mol Cell Biol. 2003, 23: 2834-2843.PubMed CentralView ArticlePubMedGoogle Scholar
- Wade PA: Methyl CpG binding proteins: coupling chromatin architecture to gene regulation. Oncogene. 2001, 20: 3166-3173.View ArticlePubMedGoogle Scholar
- Beard C, Hochedlinger K, Plath K, Wutz A, Jaenisch R: Efficient method to generate single-copy transgenic mice by site-specific integration in embryonic stem cells. Genesis N Y N 2000. 2006, 44: 23-28.Google Scholar
- Marahrens Y, Panning B, Dausman J, Strauss W, Jaenisch R: Xist-deficient mice are defective in dosage compensation but not spermatogenesis. Genes Dev. 1997, 11: 156-166.View ArticlePubMedGoogle Scholar
- Marahrens Y, Loring J, Jaenisch R: Role of the Xist Gene in X Chromosome choosing. Cell. 1998, 92: 657-664.View ArticlePubMedGoogle Scholar
- Yang X, Lay F, Han H, Jones PA: Targeting DNA methylation for epigenetic therapy. Trends Pharmacol Sci. 2010, 31: 536-546.PubMed CentralView ArticlePubMedGoogle Scholar
- Hadjantonakis AK, Gertsenstein M, Ikawa M, Okabe M, Nagy A: Generating green fluorescent mice by germline transmission of green fluorescent ES cells. Mech Dev. 1998, 76: 79-90.View ArticlePubMedGoogle Scholar
- Itzkovitz S, van Oudenaarden A: Validating transcripts with probes and imaging technology. Nat Methods. 2011, 8: S12-S19.PubMed CentralView ArticlePubMedGoogle Scholar
- 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.View ArticlePubMedGoogle Scholar
- Silva J, Mak W, Zvetkova I, Appanah R, Nesterova TB, Webster Z, Peters AHFM, 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.View ArticlePubMedGoogle Scholar
- Engreitz JM, Pandya-Jones A, McDonel P, Shishkin A, Sirokman K, Surka C, Kadri S, Xing J, Goren A, Lander ES, Plath K, Guttman M: The Xist lncRNA exploits three-dimensional genome architecture to spread across the X Chromosome. Science. 2013, 341: 1237973.PubMed CentralView ArticlePubMedGoogle Scholar
- Liu L, Ishihara K, Ichimura T, Fujita N, Hino S, Tomita S, Watanabe S, Saitoh N, Ito T, Nakao M: MCAF1/AM is involved in Sp1-mediated maintenance of cancer-associated telomerase activity. J Biol Chem. 2009, 284: 5165-5174.View ArticlePubMedGoogle Scholar
- Chang L-K, Chung J-Y, Hong Y-R, Ichimura T, Nakao M, Liu S-T: Activation of Sp1-mediated transcription by Rta of Epstein-Barr virus via an interaction with MCAF1. Nucleic Acids Res. 2005, 33: 6528-6539.PubMed CentralView ArticlePubMedGoogle Scholar
- Masui O, Heard E: RNA and protein actors in X-chromosome inactivation. Cold Spring Harb Symp Quant Biol. 2006, 71: 419-428.View ArticlePubMedGoogle Scholar
- Pullirsch D, Härtel R, Kishimoto H, Leeb M, Steiner G, Wutz A: The Trithorax group protein Ash2l and Saf-A are recruited to the inactive X chromosome at the onset of stable X inactivation. Development. 2010, 137: 935-943.PubMed CentralView ArticlePubMedGoogle Scholar
- De Graeve F, Bahr A, Chatton B, Kedinger C: A murine ATFa-associated factor with transcriptional repressing activity. Oncogene. 2000, 19: 1807-1819.View ArticlePubMedGoogle Scholar
- Li E: Chromatin modification and epigenetic reprogramming in mammalian development. Nat Rev Genet. 2002, 3: 662-673.View ArticlePubMedGoogle Scholar
- Dodge JE, Kang Y-K, Beppu H, Lei H, Li E: Histone H3-K9 methyltransferase ESET is essential for early development. Mol Cell Biol. 2004, 24: 2478-2486.PubMed CentralView ArticlePubMedGoogle Scholar
- Zhao X, Ueba T, Christie BR, Barkho B, McConnell MJ, Nakashima K, Lein ES, Eadie BD, Willhoite AR, Muotri AR, Summers RG, Chun J, Lee K-F, Gage FH: Mice lacking methyl-CpG binding protein 1 have deficits in adult neurogenesis and hippocampal function. Proc Natl Acad Sci U S A. 2003, 100: 6777-6782.PubMed CentralView ArticlePubMedGoogle Scholar
- Waterfield M, Khan IS, Cortez JT, Fan U, Metzger T, Greer A, Fasano K, Martinez-Llordella M, Pollack JL, Erle DJ, Su M, Anderson MS: The transcriptional regulator Aire coopts the repressive ATF7ip-MBD1 complex for the induction of immunotolerance. Nat Immunol. 2014, 15: 258-265.PubMed CentralView ArticlePubMedGoogle Scholar
- Chadwick BP, Willard HF: Multiple spatially distinct types of facultative heterochromatin on the human inactive X chromosome. Proc Natl Acad Sci U S A. 2004, 101: 17450-17455.PubMed CentralView ArticlePubMedGoogle Scholar
- Nozawa R-S, Nagao K, Igami K-T, Shibata S, Shirai N, Nozaki N, Sado T, Kimura H, Obuse C: Human inactive X chromosome is compacted through a PRC2-independent SMCHD1-HBiX1 pathway. Nat Struct Mol Biol. 2013, 20: 566-573.View ArticlePubMedGoogle Scholar
- Lehnertz B, Ueda Y, Derijck AAHA, Braunschweig U, Perez-Burgos L, Kubicek S, Chen T, Li E, Jenuwein T, Peters AHFM: Suv39h-mediated histone H3 lysine 9 methylation directs DNA methylation to major satellite repeats at pericentric heterochromatin. Curr Biol. 2003, 13: 1192-1200.View ArticlePubMedGoogle Scholar
- Csankovszki G, Panning B, Bates B, Pehrson JR, Jaenisch R: Conditional deletion of Xist disrupts histone macroH2A localization but not maintenance of X inactivation. Nat Genet. 1999, 22: 323-324.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.