The WSTF-ISWI Chromatin Remodeling Complex Transiently Associates with the Human Inactive X Chromosome during Late S-Phase Prior to BRCA1 and γ-H2AX

Replicating the genome prior to each somatic cell division not only requires precise duplication of the genetic information, but also accurately reestablishing the epigenetic signatures that instruct how the genetic material is to be interpreted in the daughter cells. The mammalian inactive X chromosome (Xi), which is faithfully inherited in a silent state in each daughter cell, provides an excellent model of epigenetic regulation. While much is known about the early stages of X chromosome inactivation, much less is understood with regards to retaining the Xi chromatin through somatic cell division. Here we report that the WSTF-ISWI chromatin remodeling complex (WICH) associates with the Xi during late S-phase as the Xi DNA is replicated. Elevated levels of WICH at the Xi is restricted to late S-phase and appears before BRCA1 and γ-H2A.X. The sequential appearance of WICH and BRCA1/γ-H2A.X implicate each as performing important but distinct roles in the maturation and maintenance of heterochromatin at the Xi.


Introduction
During each somatic cell division, DNA of the parent cell is precisely duplicated and distributed into two identical daughter cells. As a result of this process, however, more than just genetic information is inherited. The packaging of the DNA as heterochromatin and euchromatin must also be faithfully reestablished in order to maintain the epigenetic state of the genome, or the epigenome [1]. Failure to successfully reinstitute the epigenetic signature is a common feature of disease [2]. Alteration of the epigenome to switch genes on or off is critical for the differentiation of cells in order to permit highly specialized function. To achieve gene silencing throughout the developmental process, facultative heterochromatin must be established and maintained. The most extensive example of facultative heterochromatin is the mammalian inactive X chromosome (Xi), the end product of X chromosome inactivation (XCI) [3], an essential process that balances the levels of X-linked gene expression between the sexes by silencing most gene expression chromosomewide [4,5]. Consequently, due to its size the Xi can be readily detected in female nuclei [6], providing a valuable cytological model to investigate facultative heterochromatin in general.
In eutherian mammals, an integral component of the process is the X-inactive specific transcript (XIST). XIST is a long noncoding RNA that is expressed exclusively from the chosen Xi [7,8], and remains physically associated with the territory of the chromosome at interphase [9]. XIST RNA recruits various chromatin complexes to the Xi [10] that assist in repackaging the chromosome into facultative heterochromatin [11] characterized by increased CpG methylation [12,13], acquisition of repressive histone modifications/variants [14][15][16][17][18][19][20], and a substantial delay in DNA replication during S-phase relative to the active X chromosome (Xa) [21,22].
Once established, the Xi is remarkably stable and remains silent even in the absence of XIST/Xist [23,24,25]. Many investigations probing the role of chromatin features in maintaining gene silencing at the Xi have taken advantage of a mouse Xchromosome integrated green fluorescent protein (GFP) transgene [26], that is silenced when located on the Xi. Reactivation from the Xi can be measured by the appearance of GFP. However, the number of cells showing reactivation is very low [24,27,28,29], even when multiple chromatin features are disrupted simultaneously [30]. In fact, global reactivation has only been achieved through fusion of female mammalian cells with mouse embryonal carcinoma cells [31][32][33][34][35], or through forced expression in mouse of exogenous factors as cells revert to a pluripotent state [36]. These observations highlight the stability of facultative heterochromatin at the Xi.
Although much is known about the initiation of XCI and the resultant chromatin changes [10], our understanding of how the Xi chromatin is faithfully maintained throughout each cell cycle is limited. An attractive period in which facultative heterochromatin can be maintained is during S-phase as the underlying DNA sequence is replicated [37]. The Xi replicates in late S-phase [21,22], and as such one would anticipate transient association of chromatin complexes with the Xi during this period of the cell cycle. We have previously identified the Breast Cancer 1 gene (BRCA1) and the histone variant H2A.X phosphorylated at serine 139 (c-H2A.X) as features that transiently appear at the Xi during its replication in late S-phase [38]. BRCA1 is a multifunctional nuclear phosphoprotein [39] that has a central role in maintaining genome integrity as part of the DNA damage response pathway [40]. Other functions of BRCA1 include associating with a chromatin-remodeling complex [41], inducing alterations in chromatin condensation [42], and assisting in heterochromatin mediated gene silencing via ubiquitylation of histone H2A [43].  The histone variant H2A.X is phosphorylated in response to DNA damage [44] and is common to several different DNA damage response pathways throughout the cell cycle [45], including modulating homology-directed double strand break repair with BRCA1 during S-phase [46]. Although the nature of the relationship between BRCA1 and the Xi remains unresolved [47,48], many functional attributes of the protein support a potential role for BRCA1 in maintaining heterochromatin at the Xi during late S-phase [28]. Alternatively, it is also possible that BRCA1 and c-H2A.X are acting at the Xi during late S-phase in response to DNA damage incurred during replication [46]. Replication forks may stall or collapse when impeded, such as when secondary structure, transcription machinery, or a replication fork barrier is encountered [49]. Recent studies have highlighted that the DNA damage response and repair pathways not only rectify damage generated by the replication process, but also regulate the progression of replication forks [50,51].
The goal of this study is to examine the attributes of the protein termed Williams syndrome transcription factor (WSTF) [52,53], which has ties to both chromatin remodeling [54] and in mediating the DNA damage response [55]. Here we report the transient association of WSTF partnered with the ATPase subunit sucrose nonfermenting homolog (SNF2H) in the WSTF-ISWI Chromatin remodeling complex (WICH) [56] with the Xi during late S-phase as the chromosome is replicated. WSTF is encoded by the WBSCR9 gene at chromosome 7q11.23, and is one of approximately 28 genes that are lost due to a large deletion of 1.5-1.8 Mb from one copy of chromosome 7 in Williams syndrome (OMIM#194050) [57]. WSTF is a versatile nuclear protein that is a central component of several different chromatin remodeling complexes [58], and more recently was shown to be the kinase responsible for tyrosine 142 phosphorylation of H2A.X [55], a modification that modulates the H2A.X-mediated DNA damage response pathway. Like BRCA1, the many functional attributes of WSTF suggest that this protein may play a multifaceted role at newly replicated DNA.

Materials and Methods
Cell culture and BrdU labeling  Rabbit anti-macroH2A.1 antibody was described previously [59]. Alexa-FluorH conjugated secondary antibodies were obtained from Life Technologies Corporation (Invitrogen).

Immunofluorescence, immuno-DNA FISH, and immuno-RNA FISH
Approximately 200,000 cells were grown directly on slides. Slides were washed once with 1X phosphate buffered saline (PBS) for 2 minutes at room temperature (RT) and fixed in 4% formaldehyde/1X PBS/0.1% Triton X-100 for 10 minutes at  RT. Slides were washed twice in 1X PBS for 2 minutes each at RT and blocked in 3% BSA/1X PBS/0.1% Tween 20 for 30 minutes at RT. Slides were washed three times with 1X PBS for 2 minutes each at RT. Slides were incubated with primary antibodies in 1% BSA/1X PBS/0.1% Tween 20 for 1 hour at RT. Slides were washed three times with 1X PBS for 2 minutes each at RT before incubation with secondary antibodies in 1% BSA/1X PBS/0.1% Tween 20 for 1 hour at RT. Slides were washed three times with 1X PBS for 2 minutes each at RT, fixed as above, and finally washed three times with 1X PBS for 2 minutes each at RT. DNA was counterstained using ProLongH Gold antifade reagent supplemented with 49,6-diamidino-2phenylindole (DAPI)(Life Technologies Corporation). Detection of BrdU incorporation was performed after staining and fixing. Antibody access was achieved by treating cells in 4N HCl for 10 min at RT prior to neutralizing in 1x PBS and staining with anti-BrdU-FITC for 1 hour in 1%BSA/1xPBS/0.1% Tween 20. Slides were then washed, fixed, and washed again as described above. For immunofluorescence combined with DNA FISH, immunofluorescence was first performed as described above. Bacterial Artificial Chromosomes (BACs) were obtained from the BACPAC resource center (Children's Hospital Oakland Research Institute). P1-artificial chromosome (PAC) RP5-1174A5, which contains rDNA and detects the Nucleolar Organizing Regions (NOR) [60] was obtained from Dr. P. Finelli. The FISH probe was denatured in the presence of human Cot-1 DNA (15279-011, Life Technologies Corporation) at 78uC for 10 minutes followed by 30 minutes at 37uC to block repetitive sequences before applying to the slides. Slides were dehydrated in 70% and 100% ethanol for 3 minutes each at 4uC and then denatured in 50% formamide/2X SSC for 10 minutes at 83uC. Slides were dehydrated once more and air-dried before applying the FISH probe, sealing under cover glass, and incubating overnight at 37uC. Slides were then washed twice in 50% formamide/2X SSC at 43uC for 8 minutes each and once in 2X SSC at 43uC for 8 minutes. Antifade containing DAPI was added to the slides. Immunofluorescence combined with RNA FISH was performed essentially as described [61].
Fluorescence in situ hybridization probe preparation Direct-labeled FISH probes were prepared using a Nick Translation kit with SpectrumGreen dUTP according to the manufacturer's recommendations (Abbott Molecular). RP11-423G05 BAC was obtained from the BAC-PAC Resource Center, Children's Hospital Oakland Research Institute.

Image acquisition
Images were collected using a DeltaVision pDV. Delta Vision images were deconvolved with softWoRx 3.7.0 (Applied Precision) and compiled with Adobe Photoshop CS2 (Adobe Systems).

Transient association of WSTF with the human Xi
A previous screen investigating the distribution of a variety of chromatin proteins with the human Xi revealed that most proteins were either enriched, deficient, or showed comparable levels at the Xi relative to their general nuclear distribution [62]. A fourth type of protein, those that are elevated at the Xi in only a small proportion of cells, was also identified. Examination of the nuclear distribution of the multifunctional chromatin protein WSTF [58] in the diploid female hTERT-RPE1 cell line showed that in a small number of cells (Table 1) the protein was enriched in the vicinity of the Barr body ( Figure 1A); a densely staining DNA mass that corresponds to the Xi [6]. An indistinguishable pattern of distribution was obtained when using an antibody raised to an independent epitope of WSTF (Compare the left and right panels in Figure 1A) confirming that the enrichment corresponded to WSTF protein and not an antibody artifact. Extending this analysis to other diploid cells indicated that the enrichment was common to all female, but not male cells examined ( Figure 1B and Table 1).

The WICH complex localizes to the Xi
Prior research indicates that WSTF participates in several different complexes [58]. Thus, we sought to determine what context WSTF can be seen in when enrichment is observed at the Xi. Towards this goal, we performed immunostaining on hTERT-RPE1 cells using antibodies raised against WSTF and subunits of the various WSTF-containing complexes. We used BRM for WSTF Including the Nucleosome Assembly Complex (WINAC) [63] and SNF2H [64] for the WICH and B-WICH chromatin remodeling complexes [56,65].
In hTERT-RPE1 cells, the distribution of WSTF appeared distinct from BRM, and consistent with previous observations, BRM was deficient at the Xi [62] even when WSTF was enriched ( Figure 2, top row). In contrast, WSTF showed substantial overlap with SNF2H at the Xi in 100% of cells that had WSTF enrichment ( Figure 2, bottom row, n = 72).
WSTF participates in two separate SNF2H-containing complexes. The first complex, WICH, consists of WSTF and SNF2H only [56] and is associated with transcription regulation and chromatin remodeling [54,56,66]. The second complex, B-WICH, contains WSTF and SNF2H alongside several other factors [65], and has been linked to ribosomal gene transcription [65,67].
Since both complexes contain WSTF and SNF2H, the colocalization of these two proteins at the Xi raises the question of whether WSTF enrichment at the Xi occurs as part of the WICH or B-WICH complex. The B-WICH complex has been shown to localize to the fibrillar region of nucleoli in order to participate in both RNA polymerase I and RNA polymerase III transcription [65,67,68]. To identify the specific WSTF-complex at the Xi, we performed additional immunofluorescence analyses with antibodies raised against WSTF and the conserved nucleolar protein fibrillarin [69]. Additionally, we performed immuno-DNA FISH using antibodies raised against WSTF and a direct labeled PAC probe, RP5-1174A5, which contains rDNA repeats and highlights the nuclear organizing region (NOR) [60]. Our results indicated that when WSTF enrichment was observed in comparison to fibrillarin, (Figure 3A), 96% of the WSTF enrichments were located on the periphery of nuclei and were not in the vicinity of the fibrillar region of nucleoli ( Figure 3D, n = 93). Similarly, in comparison to the NOR regions ( Figure 3B), 87% of the WSTF enrichments were not located near a nucleoli, but were found on the nuclear periphery ( Figure 3D, n = 46)-evidence that WSTF enrichment at the Xi most likely occurs as part of the WICH complex. To further distinguish WSTF enrichment in the context of the WICH or B-WICH complex, we performed immunofluorescence with antibodies raised against WSTF and   Figure 3C) [65]. Our results indicated that WSTF enrichment appeared at the nuclear periphery and separate from MYBBP1A in 84% of WSTF enriched nuclei ( Figure 3D, n = 45), strongly supporting the conclusion that WSTF enrichment at the Xi is WICH and not B-WICH. From this point forward, where appropriate we will refer to enrichment of SNF2H or WSTF at the Xi as the WICH complex using either as a marker for the chromatin remodeling complex.

WICH interacts with the Xi during its replication in late S phase
The transient nature of the WICH enrichment at the Xi ( Table 1) suggests association in a cell-cycle dependent manner. Prior research has revealed that WSTF associates with pericentric heterochromatin [56] and is recruited to replication foci through interaction with proliferating cell nuclear antigen (PCNA) [54]. Therefore, we sought to determine if elevated levels of WICH at the Xi correlated with late S-phase, the period in which the Xi DNA is replicated [21,22]. Cells show characteristic patterns of BrdU incorporation for early, mid, and late S-phase [70], and the Xi can be readily detected in late S-phase [71]. Cells were briefly pulse labeled with BrdU and WSTF distribution was examined. WSTF showed substantial, but not complete overlap with BrdU at the Xi ( Figure 4A), and all cells with WSTF enrichment at the Xi were in late S-phase (n = 53). Not all late S cells, however, were enriched for WSTF (22.0% of late S-phase cells did not exhibit WSTF enrichment, n = 59) indicating that WICH association with the Xi occurs during a narrow period of late S-phase, accounting for the low number of cells displaying enrichment (Table 1).
To further characterize the relationship of WICH with replicating DNA, we made comparisons to the distribution of the p150 subunit of chromatin assembly factor (CAF1), which plays a critical role in assembly of nucleosomes on newly replicated DNA [72], and the DNA polymerase auxiliary protein PCNA, which coordinates DNA synthesis on the leading and lagging strands [73]. PCNA, along with CAF1, has a central role in the inheritance of chromatin state during S-phase [74] and can be used to highlight the replicating Xi during this point in the cell cycle [38]. These two proteins largely overlapped at the Xi, as can be seen with the example shown in Figure 4B. A comparison of SNF2H or WSTF with CAF1 or PCNA, respectively, showed that while the various protein complexes were apparent at the replicating Xi at the same point in time ( Figure 4B), they occupied largely separate territories with relatively small regions of overlap ( Figure 4B right panel), and, therefore, some WICH was independently occupying regions of the Xi as it was replicated. Prior studies have shown that WSTF interacts with PCNA [54] and this is believed to be the mechanism used to recruit WICH to replication foci [66]. Our data are mostly consistent with these observations, although sites of WSTF association with the Xi that were not co-occupied by PCNA suggest that WICH can be retained at replicating chromatin in a PCNA-independent manner [54].
WICH associates with the Xi before BRCA1 and c-H2A.X We have previously shown that BRCA1 and c-H2A.X transiently associate with the Xi during late S-phase [38]. Therefore, we sought to determine the spatiotemporal relationship of WICH with BRCA1 and c-H2A.X at the Xi by comparing patterns of each with relation to one another as well as to a marker of G2. Although WICH and BRCA1 can be enriched at the Xi at the same time ( Figure 5A, top row), the two showed no obvious overlap and instead occupied closely abutting regions of the Xi territory. In 20% of nuclei (n = 65), WSTF enrichment was seen without BRCA1 ( Figure 5A, middle row). Likewise, 59.1% of nuclei (n = 44) showed BRCA1 enrichment without WSTF ( Figure 5A, bottom row), suggesting a sequential association of the two proteins at the Xi. In contrast, WSTF partially overlapped c-H2A.X distribution ( Figure 5B, top row) in some nuclei, but like BRCA1, nuclei were observed that were enriched only for WSTF ( Figure 5B, middle row) or only for c-H2A.X ( Figure 5B, bottom row). Given that BRCA1 also overlaps c-H2A.X [38] suggests that c-H2A.X at the Xi bridges the periods of WSTF and BRCA1 association. It remains unknown, however, in what order these proteins associate with the Xi (WICH, c-H2A.X, then BRCA1 or the opposite order). Phosphorylation of histone H3 at serine 10 (H3S10ph) initiates at heterochromatic regions during G2 and continues to accumulate as cells proceed toward mitosis [75]. However, in hTERT-RPE1 cells, H3S10ph is detected several hours prior to mitosis [76] with 23.19% of H3S10ph-positive nuclei also exhibiting BrdU incorporation, marking cells in either very late S-phase or G2 (Table 2). Using indirect immunofluorescence, we looked for cells showing elevated levels of SNF2H, BRCA1, and c-H2A.X at the Xi relative to detection of H3S10ph ( Figure 6). SNF2H enrichment never coincided with the appearance of H3S10ph (n = 50). However, H3S10ph was present in 72.4% of BRCA1 enriched nuclei and (n = 76) and 41.5% of c-H2A.X enriched nuclei (n = 82). We interpret these findings as the WICH complex associating with the Xi prior to both c-H2A.X and BRCA1, with BRCA1 associating last.

WICH shows minimal overlap with some but not all chromatin features of the Xi
Heterochromatin of the human Xi is characterized by the cis association of XIST RNA [8], the presence of histone variant macroH2A1 [15], binding of the human homologues of drosophila heterochromatin protein 1 (HP1) [62], alpha, beta, and gamma [77,78], and trimethylation of histone H3 lysine residues 9 (H3K9me3) [79] and 27 (H3K27me3) [19,20], that are nonrandomly arranged into two non-overlapping territories at the Xi (Figure 7) [79]. Previously we reported that BRCA1 and c-H2A.X do not appear to overlap with chromatin features of the Xi [38]. Similarly, we found that WICH did not overlap with features of the facultative-like heterochromatin on the Xi (macroH2A1, H3K27me3, and XIST RNA; Figure 7, red) [79], but instead Figure 8. WICH enrichment at the Xi does not overlap with H3K27me3, macroH2A1 or XIST, but does show limited overlap with the HP1 and H3K9me3 territories. A WSTF or SNF2H enrichment relative to facultative-like heterochromatin. Images show examples for the distribution of SNF2H (Red) relative to macroH2A.1 (Green, top row), and WSTF (Green) relative to H3K27me3 (Red, middle row) in hTERT-RPE1 nuclei by indirect immunofluorescence. Bottom panels show WSTF distribution by indirect immunofluorescence (Red) relative to hybridization pattern of a direct-labeled XIST probe (Green) in a hTERT-RPE1 nucleus. White arrowheads point to the Xi. Overlapping signals appear yellow. Nuclei counterstained with DAPI (Blue). The last panel in the series shows an enlarged view of the area indicated by white arrowheads. B WSTF or SNF2H enrichment relative to constitutive-like heterochromatin marks. Images show examples for the distribution of SNF2H (Red) relative to H3K9me3 (Green, top row) and WSTF (Green) relative to HP1a (Red, second row), HP1b (Red, third row) and HP1c (Red, bottom row) in hTERT-RPE1 nuclei by indirect immunofluorescence. Overlapping signals appear yellow. Panels to the right show nuclei counterstained with DAPI (Blue). White arrowheads point to the Xi. The last panel in the series shows an enlarged view of the area indicated by white arrowheads. doi:10.1371/journal.pone.0050023.g008 occupied comparable sized territories immediately adjacent ( Figure 8A). Next we examined the constitutive-like heterochromatin features H3K9me3 and HP1a, HP1b, and HP1c, thatin hTERT-RPE1 cells occupy a substantial proportion of the Xi [79] (Figure 7, green). In most instances these features also appeared distinct from WICH. However, unlike the facultativelike heterochromatin markers that usually occupied territories that were side-by-side with WICH, the constitutive-like heterochromatin markers frequently appeared intertwined with WICH ( Figure 8B, right panels). Furthermore, close examination revealed small, but not insignificant regions of overlap with HP1 and H3K9me3-something not observed with macroH2A1, H3K27me3, or XIST.
The limited overlap for some Xi features with WICH and complete lack for others could be interpreted as WICH enrichment reflecting a chromatin territory close to the Xi, but not the Xi itself. To address this, we took two complementary approaches. First, we compared the enrichment of WICH to the distribution of histone H3 dimethylated at lysine 4 (H3K4me2) and H3 acetylated at lysine 9 (H3K9Ac). Nucleosomes of the Xi are hypoacetylated [17,80] and underrepresented for H3K4me2 [14], which is consistent with the association of both modifications with euchromatin [81]. As such, by immunofluorescence the Xi appears as a distinct hole in the distribution, demarcating the territory of the chromosome [14,62]. Elevated levels of WICH at the Xi clearly reside within this hypo-H3K9Ac and hypo-H3K4me2 region ( Figure 9). Next, we performed WSTF immunofluorescence combined with fluorescence in situ hybridization (FISH) with X-linked bacterial artificial chromosomes (BACs). Several BAC clones were selected that reside within bands of macroH2A.1/H3K27me3 or H3K9me3 (Figure 7, blue; Table S1) [82]. While most BACs appeared close to sites of WICH Xi-enrichment (data not shown), one BAC (RP11-423G05) containing the DOCK11 gene overlapped with WICH enrichment in 74.6% of cells ( Figure 10, n = 59; Figure 7, black arrowhead).
This suggests that WICH is not associating uniformly with all regions of the Xi, but perhaps its activity is restricted to specific regions.

Discussion
WSTF has been implicated in heterochromatin replication, in particular HP1-containing heterochromatin [54,56], and is a candidate for the maintenance of chromatin states in somatic cell division [66]. In a mouse mutagenesis screen, both components of the WICH complex (Snf2h and Wstf) [56] were independently identified as enhancers of transgene variegation [83] supporting a role for this complex in epigenetic inheritance. Here we report that the WICH complex associates with the territory of the Xi as the chromosome is replicated in late S-phase prior to the appearance of c-H2A.X and BRCA1 at the Xi [38]. While elevated levels of WICH at the Xi appear to precede serine 139 phosphorylation of H2A.X, the two do appear to overlap in some cells. The same is not true for BRCA1 that always appears spatially distinct from WICH and that, with c-H2A.X, remains associated with the Xi territory into a later point in late S-phase or, possibly, early G2 than WICH enrichment. One interpretation of these data is that phosphorylation of serine 139 of H2A.X occurs at the Xi immediately before BRCA1, which is supported by c-H2A.X recruiting BRCA1 [84].
Chromatin of the human Xi is composed of two spatially distinct types of heterochromatin for which the underlying DNA replicates during distinct periods of late S-phase [79,82]. DNA characterized by facultative-like heterochromatin (macroH2A1, H3K27me3, and XIST) replicates in the first part of late S-phase, whereas DNA packaged into constitutive-like heterochromatin (H3K9me3 and HP1) replicates immediately after this period in very late S-phase [79]. Our previous analysis of BRCA1 and c-H2A.X at the Xi in late S-phase found that neither showed extensive overlap with either heterochromatin type, but instead appeared to reside in a territory immediately adjacent [38,48]. What role BRCA1 has at the Xi in this capacity remains unclear [28,47,48]. Consistent with these observations, we do not see extensive overlap between the WICH complex and features of the Xi. However, close examination of deconvolved images of WSTF with the distribution of HP1 or H3K9me3 does show that at least some regions are overlapping. However, it is important to note that HP1 has been shown to be recruited to sites of DNA damage to assist in repair [85,86]. It is possible therefore, that when WICH is seen overlapping with HP1 at the Xi, that this HP1 is not necessarily associating with constitutive-like heterochromatin, but is, in fact, localizing to regions of DNA damage, which could be found at either constitutive-like or facultative-like heterochromatin.
Overlap was never seen between WSTF enrichment and the facultative-like heterochromatin type. One plausible explanation for this could be that this type of heterochromatin dissipates as the chromatin is replicated [71] and, as a consequence, the WICH complex could actually be acting on this territory when seen adjacent to HP1 and H3K9me3 chromatin. Alternatively, WICH may not be associating with facultative-like chromatin as suggested by our inability to detect an overlap between WICH and FISH with probes from this territory [82]. Only a probe from the H3K9me3/HP1 region was coincident with WICH enrichment suggesting that its role on the Xi is restricted to the HP1containing chromatin. Previous WSTF knock-down experiments demonstrated that primarily HP1b-containing regions were impacted by WSTF deficiency [54]. Although we concluded that WSTF enrichment at the human Xi represented the WICH complex and not the nucleolar-associated B-WICH complex, it is interesting that HP1 subtypes have been linked to the transcription of ribosomal genes [87,88]. Therefore, WSTF may act at HP1b in general, regardless of whether as WICH or B-WICH.
In mouse, replication of the Xi coincides with relocalization of the chromosome to a perinucleolar position, in the vicinity of a region of Snf2h enrichment [89]. It is conceivable that the perinucleolar Snf2h actually represents the WICH complex suggesting a common mechanism of Xi replication between man and mouse. Failure to localize to this region compromised Xi heterochromatin and gene silencing [89], supporting an important role for Snf2h containing complexes in maintaining the Xi. In identifying the WSTF containing complex that is associating with the human Xi, we concluded that it is the WICH complex since WSTF enrichment did not colocalize with a known marker of the nucleoli-associated B-WICH complex [58].
This analysis highlights, however, a difference between human and mouse. Our data indicate that the replicating human Xi with WICH enrichment were found more often at the nuclear periphery than in a perinucleolar region. Nevertheless, the presence of Snf2h/SNF2H is common suggesting that while the subnuclear localization may not be conserved in man, the chromatin-remodeling complex is, and, therefore, is likely important for the maintenance of human Xi chromatin.
WSTF has also been identified as the kinase responsible for the constitutive phosphorylation of H2A.X at tyrosine 142 [55]. Interestingly, the tyrosine 142 phosphorylation is depleted in response to DNA damage by the phosphatase eyes absent 1-3 (EYA1-3) [90]. Removal of this phosphate from tyrosine triggers the recruitment of DNA damage response proteins, such as MDC1 [55,90], whereas failure to do so activates apoptosis [90]. It is conceivable that elevated levels of WICH at the Xi during late Sphase corresponds to a wave of H2A.X tyrosine 142 phosphorylation, reestablishing the modification in newly formed chromatin. If WICH functions to prime the chromatin for a future DNA damage response, its association at the Xi in this capacity would be unlinked to the presence of c-H2A.X and recruitment of BRCA1, which would be present after EYA1-3 removed the tyrosine phosphate from H2A.X in response to damage.
Precisely what function is being performed as WICH, c-H2A.X, and BRCA1 sequentially associate with the Xi is debatable. Given that BRCA1 and WICH are both implicated in the maintenance of heterochromatin [43,66] and that mouse Xi chromatin is compromised when the kinases responsible for H2A.X serine 139 phosphorylation are inhibited [29] suggests, however, a direct or indirect role in the maintenance of Xi heterochromatin. Additionally, the role of WSTF in reestablishing the ground state for future DNA damage responses [55] and the fact that WSTF associates with newly replicated DNA provides further evidence that the WICH complex is an integral component in chromatin maturation. Further investigation of the function of WICH is required to reveal how this dynamic complex contributes to the many layers of chromatin maturation and maintenance.

Supporting Information
Table S1 List of BAC clone pools used for FISH with WSTF indirect-immunofluorescence. BAC clone members of each pool (1-6 as indicated in Figure 7) are highlighted in different colors and their coordinates on the X (where known) are given based on hg18, as are genes or genomic features of interest in the corresponding BAC. The BAC clone RP11-423G05 that is shown overlapping WSTF in Figure 10 is highlighted in black. (PDF)