Phosphorylation of Williams Syndrome Transcription Factor by MAPK Induces a Switching between Two Distinct Chromatin Remodeling Complexes*

Changes in the environment of a cell precipitate extracellular signals and sequential cascades of protein modification and elicit nuclear transcriptional responses. However, the functional links between intracellular signaling-dependent gene regulation and epigenetic regulation by chromatin-modifying proteins within the nucleus are largely unknown. Here, we describe novel epigenetic regulation by MAPK cascades that modulate formation of an ATP-dependent chromatin remodeling complex, WINAC (WSTF Including Nucleosome Assembly Complex), an SWI/SNF-type complex containing Williams syndrome transcription factor (WSTF). WSTF, a specific component of two chromatin remodeling complexes (SWI/SNF-type WINAC and ISWI-type WICH), was phosphorylated by the stimulation of MAPK cascades in vitro and in vivo. Ser-158 residue in the WAC (WSTF/Acf1/cbpq46) domain, located close to the N terminus of WSTF, was identified as a major phosphorylation target. Using biochemical analysis of a WSTF mutant (WSTF-S158A) stably expressing cell line, the phosphorylation of this residue (Ser-158) was found to be essential for maintaining the association between WSTF and core BAF complex components, thereby maintaining the ATPase activity of WINAC. WINAC-dependent transcriptional regulation of vitamin D receptor was consequently impaired by this WSTF mutation, but the recovery from DNA damage mediated by WICH was not impaired. Our results suggest that WSTF serves as a nuclear sensor of the extracellular signals to fine-tune the chromatin remodeling activity of WINAC. WINAC mediates a previously unknown MAPK-dependent step in epigenetic regulation, and this MAPK-dependent switching mechanism between the two functionally distinct WSTF-containing complexes might underlie the diverse functions of WSTF in various nuclear events.

Chromatin structure is intimately involved in the regulation of gene expression. The dynamics of chromatin structure are tightly regulated through multiple mechanisms such as histone modification, chromatin remodeling, histone variant incorporation, and histone eviction. Chromatin reorganization is performed by several chromatin-modifying complexes to allow effector proteins (transcription factors) access to DNA (4,5). Two major classes of chromatin-modifying complexes have been well characterized (6). One class is a histone-modifying complex, and the other class is an ATP-dependent chromatin-remodeling complex. The latter complex uses ATP hydrolysis to rearrange nucleosomal arrays in a noncovalent manner, thereby facilitating or preventing access of nuclear factors to nucleosomal DNA (7,8). Precise selection of catalytic ATPase subunits in combination with other components determines the role of these complexes in a spatiotemporal manner (6,9). Studies of genetically altered animals deficient in individual complex components suggest that each chromatin remodeling complex contributes to a specific cellular process. However, little is known about the regulatory mechanisms accounting for the specific physiological impact of a complex (6,10,11).
Intracellular signaling impacts gene regulation and thereby modulates cell proliferation, differentiation, migration, and apoptosis. Although several steps of the gene regulation can be modulated by various signaling molecules, the main regulatory mode is through modifications of transcription factors by their downstream effectors (12). Several transcription factors, including nuclear receptors, have already been reported as modification targets (13,14). But a detailed understanding of the regulatory mechanisms controlling transcription together with the reorganization of chromatin structure is lacking.
Vitamin D receptor (VDR) 3 is a member of the steroid/thyroid hormone nuclear receptor superfamily regulating bone metabolism, calcium homeostasis, and cell differentiation by binding with 1␣,25-dihydroxyvitamin D 3 (D3), a physiologi-cally active form of vitamin D (15)(16)(17). Like other nuclear receptors, VDR serves as a ligand-dependent transcription factor that requires distinct classes of co-regulators and multiprotein co-regulator complexes to initiate D3-induced chromatin reorganization (18). These complexes appear to modify chromatin configuration by controlling nucleosomal rearrangement and enzyme-catalyzed modifications of histone tails (19,20). As for VDR, many types of co-regulator complexes have been identified thus far, including p160 family histone-acetylating complexes, DRIP-TRAP complexes, and ATP-dependent chromatin remodeling complexes (21,22).
WINAC is a SWI/SNF-type ATP-dependent chromatin remodeling complex that we recently identified as a VDR-interacting complex (1). WSTF, a component of WINAC, is indispensable for gene regulation by VDR through its chromatin remodeling activity (2,3,21,22). The physiological role of this complex has been clarified in heart development as well as calcium metabolism through analysis of WSTF-deficient mice that lack WINAC-mediated regulation of transcription (3). WSTF constitutes another complex designated WICH (WSTF-ISWI chromatin remodeling complex) (23). WICH serves as an ISWI-type chromatin remodeling complex, which is responsible for recovery from DNA damage at various steps (3,24).
In this study, we identified a novel type of intracellular signaldependent epigenetic regulation mediated by WSTF, which is conceivably required for proper WINAC function but not for WICH function. MAPK-dependent phosphorylation of WSTF is required for maintaining the complex assembly of WINAC essential for its proper function as an ATP-dependent chromatin remodeling complex. WINAC-mediated transcriptional property of VDR was consequently affected by MAPK pathways. However, the recovering process from DNA damage mediated by WICH function was not impaired. Thus, WINAC is a novel nuclear mediator of MAPK signaling cascades for their gene regulations. WSTF serves as a signaling sensor to switch the two chromatin remodeling complexes.

Luciferase Assay
MCF7 cells or MEFs were maintained in phenol red-free Dulbecco's modified Eagle's medium supplemented with 10% charcoal/dextran-treated fetal bovine serum and antibiotics (Invitrogen). At 40 -50% confluence, the cells were transfected with the indicated plasmids by Lipofectamine Plus reagent (Invitrogen) in 24-well trays. The amounts of each DNA transfected to the cells are described in the figure legends and the total amount of DNA was adjusted by supplementing it with empty vector. After times indicated in the figure legends, cells were lysed with 100 l of lysis buffer (Promega, Madison, WI). Luciferase activities from 30 l of lysate were determined using the Dual-Luciferase reporter assay system (Promega). Renilla luciferase was used as a reference to normalize transfection efficiencies in all experiments. All values are means Ϯ S.D. from at least three independent experiments (1-3).

Maintenance of the WSTF-S158A Stably Expressing Cells and WSTF ؊/؊ MEF
For establishment of the MCF7 cells stably expressing WSTF-wild type or WSTF-S158A mutant, MCF7 cells were grown in 10-cm dishes and transfected with 10 g of pcDNA-FLAG-WSTF or pcDNA-FLAG-WSTF-S158A vector by Lipofectamine Plus (Invitrogen). After 48 h, cells were selected with 500 g/ml G418 (Wako, Osaka, Japan) and cloned by cloning rings. From G418-resistant clones, appropriate clones were selected by Western blot as described previously (1,27). MEFs from WSTF Ϫ/Ϫ mice were prepared and handled as described previously (2,3). Briefly, MEF cell lines were obtained from wild type (WT) or WSTF Ϫ/Ϫ 13.5-day-old embryos and used at the 10th generation. The MEF cell lines were replated at a density of 1 ϫ 10 6 cells on gelatin-coated 10-cm dishes and maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37°C in a humidified atmosphere containing 5% CO 2 .

Purification of Phosphoproteins and Partially Purified WINAC
Biochemical purification was done following our standard purification methods (1,(27)(28)(29)(30)(31)(32). The purification of phosphorylated proteins from MCF7 cells was performed with the phosphoprotein purification kit (catalog number 37101, Qiagen) following the manufacturer's protocol (33). Briefly, 10 10-cm culture dishes of MCF7 cells were lysed with 10 ml of lysis buffer containing 0.25% CHAPS. The cell lysates were loaded on the anti-phospho-Ser/Thr column and eluted following the manufacturer's protocol (Qiagen). The eluates were then subjected to Western blot.
The purification of partially purified WINAC was done following our previous paper with some modifications to remove WICH components (1). Briefly, about 10 9 cells of each stable transformant were harvested, and the nuclear extracts (80 mg) were prepared by the method initially described by Dignam et al. (34). One hundred l of hSNF2h antibody was added to the nuclear extracts followed by batch collection with 500 l of protein G-Sepharose (GE Healthcare). After collecting the resin on a 10-ml column, the flow-through fraction was next transferred to an anti-FLAG M2 affinity resin (Sigma) column (400-l bed volume) and eluted from the resin with 400 l of 300 g/ml FLAG peptide (Sigma) for the following assays.

ATPase Assay
An ATPase assay was performed following the previous report (35). Briefly, the 5-l reaction mixture containing 10 mM HEPES (pH 7.6), 50 mM KCl, 0.1 mM EDTA, 2 mM MgCl 2 , 0.5 mM dithiothreitol, 7.5% glycerol, 0.5% Nonidet P-40, 30 M cold ATP, 5 Ci of [␣-32 P]ATP, 20 nM plasmid DNA, and purified WINAC complexes from the stable transformant expressing FLAG-WSTF or FLAG-WSTF-S158A mutant (as described in Fig. 5, A and C) was incubated at 37°C for 30 min. Hydrolyzed ATP and ADP were separated by TLC on polyethyleneiminecellulose plates (Sigma). A 1-l aliquot of the reaction mixture was spotted onto the plate, and TLC was carried out in 0.75 M KH 2 PO 4 . Plates were allowed to dry and assessed by autoradiography.

Quantitative RT-PCR
Total RNA was extracted with TRIzol (Invitrogen), and cDNA was synthesized using SuperScriptIII reverse transcriptase (Invitrogen). Reverse transcription of 2 g of total RNA was carried out with 0.2 g of oligo(dT) primer for 50 min at 50°C. Quantitative RT-PCR was performed using SYBR Premix EX Taq (Takara) according to the manufacturer's instructions. Predesigned quantitative RT-PCR primer sets were purchased from Takara. Experimental samples were matched to a standard curve generated by amplifying serially diluted product using the same PCR protocol. To correct for variability in RNA recovery and efficiency of reverse transcription, glyceraldehyde-3-phosphate dehydrogenase cDNA was amplified and quantified in each cDNA preparation (3,31,37).

Cell Survival Assay
All experimental procedures were conducted as described in our previous reports (3,38). MEF cells from WSTF Ϫ/Ϫ and wild type mice were incubated in 60-mm dishes at 40% confluency (4 ϫ 10 5 cells/dish). The indicated expression vectors were transfected with Lipofectamine Plus reagent (Invitrogen). After 24 h, transfected cells were treated with medium containing 0.02% methyl methanesulfonate (MMS) for 1 h, washed with

WSTF Is a MAPK-dependent Phosphoprotein in Vitro and
in Vivo-Our recent analysis of WSTF-deficient animals (3) showed that WSTF contributes to various biological events presumably through the chromatin remodeling activity of the two complexes, WINAC and WICH (1,23). However, the links between the physiological impact of WSTF and the ATP-dependent chromatin remodeling activities of these complexes are still largely unknown. We hypothesized that an unrecognized intracellular signaling pathway might mediate both the physiological functions of WSTF and the chromatin remodeling activities.
To identify intracellular signals affecting the function of these complexes in vivo, we tested inhibitors against several intracellular signaling pathways using our established reporter assay system, Gal4-fused VDR (Gal-VDR) and WSTF in MCF7 (1-3). As shown in Fig. 1A, induction of the activated transforming growth factor-␤ receptor (T␤R-I(TD)) potentiated the 1,25(OH) 2 D 3 (D3)-dependent VDR transcriptional property regardless of the presence or absence of WSTF as reported previously (compare lanes 21-24 with lanes 1-4) (25, 39). Among the tested inhibitors against the signaling pathways, an Akt inhibitor (Akt inhibitor X) did not affect the co-activation function of WSTF (Fig. 1A, compare lanes 17-20 with lanes 1-4). Nevertheless, three inhibitors for MAPK signalings, which interfere with distinct downstream pathways, strongly decreased the co-activation function of WSTF on Gal-VDR (Fig. 1A, compare lanes 5-8, 9 -12, and 13-16 with lanes 1-4,  respectively). MAPK cascades are known to respond to changes in the cellular environment (40,41), and the primary regulatory effectors work through the phosphorylation of transcription factors by downstream kinases such as ERK, JNK, or p38 (12). Although MAPK cascades are known to regulate the expression level of VDR (42), the protein expression levels, as well as the phosphorylation levels (43), were not significantly altered in our system (Fig. 1B). These results suggests that MAPK signalings might modulate the function of WSTF-containing complexes.
Next we asked whether WSTF protein could be phosphorylated in vitro by ERK1, a nuclear MAPK effector kinase (40). In our in vitro kinase assay using GST-fused WSTF deletion mutants (2), one mutant containing the WAC domain (GST-WAC) was strongly phosphorylated by recombinant ERK1 protein (Fig. 2B, lane 1). Analyzing the sequence alignment of the phosphorylated region, a consensus MAPK phosphorylation sequence (12,13,40) was found around a serine residue at Ser-158 ( Fig. 2A). To test whether this serine residue was responsible for the phosphorylation, a point mutant of this GST-fused deletion mutant from serine to alanine (GST-WAC-S158A) was constructed and purified (Fig. 2C, upper panel). By the same assay, this point mutant was not phosphorylated by ERK1 (Fig. 2C, 2nd panel). Moreover, this region (WAC domain) was also phosphorylated by recombinant proteins of P38␣ and JNK2␣ but not by AKT1 (Fig. 2C, lower panels; compare with Fig. 1A). Thus this residue proved to be responsible for the ERK1-dependent phosphorylation of this region in vitro. This phosphorylation was specifically performed by MAPK effector kinases. Finally, we asked whether this region was a major phosphorylation site in vivo using MCF7 cells stably expressing WSTF-S158A mutant (S158A). Phosphorylated WSTF was detected in cells stably expressing WSTF-wild type, and phosphorylation was attenuated by the treatment with the MAPK inhibitors (Fig. 3A, compare lane 1 and lanes 2-4). On the other hand, the phosphorylation of WSTF was scarcely detected in WSTF-S158A mutant cells (Fig. 3A, lanes 5-8). Indeed, the phosphorylation level of WSTF-wild type, as well as its expression level, was not altered by D3 stimulation (Fig. 3B), and the co-activation function of a phosphorylation mimic mutant of WSTF (S158E) (44) was comparable with WSTF-wild type (Fig.  3C, compare lane 2 with lanes 6 -9). But the co-activation function of this S158E mutant, as well as S158A mutant, for the VDR transcriptional property was not affected by the indicated MAPK inhibitors in MCF7 cells (Fig. 3C). Taken together, we conclude that WSTF is indeed phosphorylated by the activation of the three MAPK pathways in vivo, and Ser-158 residue is the main target residue of this phosphorylation.
WSTF Phosphorylation by MAPK Signaling Downstream Kinases Is Required for the ATPase Activity of WINAC-To determine whether WSTF was the only target of MAPK-dependent phosphorylation in the WSTF-containing complexes, we asked whether the phosphorylation levels of components other than WSTF were affected by a MAPK inhibitor (ERK inhibitor). In this procedure, whole cell extracts of the MCF7 cells stably expressing WSTF with or without ERK inhibitor (U0126) were added to the phosphoprotein purification column (Qiagen) (33), and then the eluates from the column were subjected to Western blot (Fig. 4A). Among the tested human SWI/SNF components (common core components of human SWI/SNF-type complexes, including WINAC) and hSNF2h (WICH component), only the phosphorylation level of WSTF was altered by the treatment with ERK inhibitor (Fig. 4B). To further understand the ERK pathway-dependent regulatory mode of WINAC function, we then partially purified WINAC from nuclear extracts of the MCF7 cells stably expressing WSTF-wild type and S158A mutant after the depletion of WICH components by hSNF2h antibody (Fig. 5, A and B). Interestingly, when detecting SWI/SNF components recruited by WSTF, several SWI/SNF components (including BRG1/ BRM, the catalytic ATPase components) were recruited to WSTF-wild type, but much less to the S158A mutant (Fig. 5C). Consistent with this observation, the ATPase activity of the immunoprecipitant with anti-FLAG affinity resin from the cells expressing WSTF-S158A mutant was lower than that from the cells expressing WSTF-wild type (Fig. 5D). Thus, it appears that the MAPK-dependent regulation of WINAC function was mediated through the maintenance of complex formation and that the stability of this complex was dependent on the phosphorylation level of the Ser-158 residue in the WSTF protein in vivo.
Phosphorylation of WSTF Is Required for WINAC Function but Not for WICH Function in Vivo-Finally, we analyzed the contribution of WSTF phosphorylation to the physiological impact of WSTF. In our previous reports, we showed that the chromatin remodeling activity of WINAC contributed to both ligand-dependent repression as well as activation by VDR (1,2,36,45). To test whether MAPK-dependent phosphorylation affected these two transcriptional activities, quantitative PCR analysis was performed comparing the two stable lines. We chose two representative VDR target genes 25(OH)24-hydroxylase as a positively regulated gene and 25(OH)1␣-hydroxylase as a negatively regulated gene (1,36). Comparing the expression profiles of the genes after D3 stimulation, both the D3-depen- dent trans-activation property and the trans-repression property of WINAC were attenuated in WSTF-S158A mutant cells (Fig. 6A). Consistent with our previous findings (1,36), the known D3-dependent change of the histone modifications, which follow WINAC-mediated chromatin reorganization on the promoter of the indicated genes, was seen in the WSTFwild type stably expressing cells (Fig. 6B, lanes 1 and 2 and 5 and  6). When compared, the levels of histone modifications in the indicated gene promoters, ligand-dependent change of histone methylation (H3K9me3 and H3K4me3) as well as a histone acetylation (AcH3), were attenuated in both positively and negatively regulated VDR target gene promoters in WSTF-S158A mutant cells (Fig. 6B, compare lane 2 with 4 and lane 6 with 8). The contribution of WSTF phosphorylation to the transcriptional property of VDR was confirmed in WSTF Ϫ/Ϫ MEF cells by testing the stimulation of the transcriptional activity of Gal-VDR by overexpression of WSTF-wild type and S158A mutant. As shown in Fig. 7A, WSTF-wild type could recover the transcriptional property of Gal-VDR efficiently, but S158A mutant did not (compare lanes 2-4). The ERK inhibitor blocked only the co-activation function of WSTF-wild type (Fig. 7A, compare lanes 2-4 with 5-7). These results suggest that MAPK-dependent phosphorylation of WSTF indeed contributes to the properties of WINAC in vivo and consequently is also required forthefullactivityofVDRasaliganddependent transcription factor.
Next we determined whether the phosphorylation of WSTF contributed to WICH function. From our previous analysis, the function of WICH appears obvious in the recovery from DNA damage in MEF cells from WSTF Ϫ/Ϫ animals (3). Thus, we tested the cell survival rate after DNA damage with overexpression of WSTF-wild type and WSTF-S158A mutant in the MEFs from WSTF Ϫ/Ϫ mice as performed previously (3). As expected, clear contribution of WICH to the recovery from DNA damage was seen (Fig.  7B, compare lanes 2 and 3 and 4 and  5). However, to our surprise, the cells transfected with WSTF-S158A mutant also recovered from DNA damage to an extent comparable with the cells transfected with WSTF-wild type (Fig.  7B, compare lanes 4 and 5 with 6  and 7, respectively). This result was supported by the analysis of complex formation between Snf2h and WSTF-S158A mutant in the wild type MEF cells as compared with that between Brg1 and WSTF (Fig. 7C). These results suggest that MAPK-dependent phosphorylation of WSTF is required for WINAC function but dispensable for WICH function in vivo (Fig. 7D).

DISCUSSION
Nuclear events such as transcription, DNA replication, and DNA repair are now believed to be orchestrated by strict epigenetic controls through reorganization of chromatin structure. Some of the protein complexes regulating changes in chromatin structure have been shown to link with intracellular signaling cascades (26,31,32,46,47). However, the underlying mechanisms of signal-dependent epigenetic changes are not fully understood. For example, in the case of chromatin remodelers, although some components are known to be recruited by specific transcription factors in a signal-dependent manner (48,49), the manner in which the specific combinations of components assemble on DNA has remained elusive (21,22). In this study, signal-dependent complex stabilization was observed by the phosphorylation of a specific component, WSTF. This result implies novel signal-dependent regulation of complex assembly by a protein modification downstream of MAPK signaling cascades (50).
Chromatin remodeling complexes work at various situations to facilitate access of the biological effectors to the target regions of the genome through altering the adjacent chromatin A, schematic diagram of the partial purification of the WINAC complex. 80 mg of nuclear extracts prepared from MCF7 cells stably expressing FLAG-WSTF or FLAG-WSTF-S158A mutant were subjected to immunoprecipitation with ␣hSNF2 antibody to remove WICH complex. The immunocomplexes fused to protein G-Sepharose beads were collected in the 10-ml empty column, and the flow-through (FT) fraction was loaded onto the ␣FLAG M2-agarose column. The FLAG-WSTF containing WINAC complex bounded to the ␣FLAG M2-agarose was eluted with FLAG peptides. B, confirmation of WICH complex depletion from nuclear extracts. The charged nuclear extracts (NE) and flow-through of the ␣hSNF2 antibody column were subjected to Western blot using ␣hSNF2 antibody. The fraction binding to the hSNF2h antibody column was also subjected to Western blot using ␣FLAG antibody. hSNF2h was depleted in the flow-through fractions from nuclear extracts from both WSTF-wild type and WST-S158A cells, and roughly equal amounts of WSTF-wild type and WSTF-S158A mutant could be detected in the immunocomplex of the ␣hSNF2 antibody. C, recruitment of the core BAF complex components to WSTF was abrogated by WSTF-S158A mutation. The partially purified WINAC complex components were analyzed by Western blot using indicated antibodies. Indicated BAF components (BRG1, BRM, BAF155, BAF170, BAF250, BAF60a, and INI1) are the known components of WINAC complex (1). D, ATPase activity of WINAC is impaired by the WSTF-S158A mutation. Partially purified WINAC prepared as described in A was incubated with radiolabeled [␣-32 P]ATP, and reaction mixtures were subjected to thin layer chromatography. Hydrolyzed [␣-32 P]ADP and unhydrolyzed [␣-32 P]ATP are indicated on the right. status (8,51). Considering the known roles of the specific components of each chromatin remodeling complex (52)(53)(54)(55), we believe that WSTF, as a specific component of WINAC, specifically works as a sensor of the various intracellular signalings for turning on the chromatin remodeling activity when required. Indeed, in our first screening, several inhibitors against various intracellular signaling cascades affected the co-activation function of WSTF (data not shown). Further mechanical analysis seems essential to understand the biological impacts of WSTF as an epigenetic determinant under various extracellular stresses when distinct intracellular signaling cascades are activated.
The physiological impact of MAPK-dependent modification of WSTF can be appreciated when WSTF-deficient mice are considered (3). WSTF is a shared component of two chromatin remodeling complexes, WINAC and WICH (1,23). WSTF-deficient mice have cardiovascular abnormalities that are presumably due to WINAC-dependent malfunction of cardiac transcription factors. On the other hand, DNA damage repair was also impaired probably due to the dysfunction of WICH (3). From our present analysis, MAPK-dependent phosphorylation of WSTF was found indispensable for proper WINAC function but not for WICH function. Thus, we surmise that MAPK-de-pendent regulation of WINAC function must have contributed to the abnormal cardiac development characteristic of WSTFdeficient mice.
It is also well known that the activation of distinct MAPK pathways has a different biological impact in heart tissues (56,57). ERK pathway is essential for the heart development and is consequently related to certain hereditary diseases (58,59), whereas p38 and JNK pathways, as well as ERK pathway, are for adaptational myocyte growth after birth (60). Thus, we speculate that the phosphorylation of WSTF by each MAPK downstream effector kinase has a distinct biological impact at various phases in the heart tissues. Combined with our histological analysis of the heart tissues from the WSTF-deficient mice embryos (3), it is conceivable that their phenotypes are attributed to the lack of biological impacts of MAPK signalings (presumably of ERK pathway) during the heart development, at least in part. As the WSTF-deficient mice die soon after birth (3), it seems impossible to test the role of WSTF-mediated chromatin remodeling activity at the adaptational myocyte growth in the situations such as the formation of myocardial hypertrophy or cardiac ischemia (56). For the better understanding of the contribution of each MAPK pathway to the FIGURE 6. Phosphorylation of WSTF by MAPK downstream kinases is required for the transcriptional regulation of VDR by WINAC. A, D3-dependent transcriptional regulation mediated by VDR is impaired in MCF7 cells stably expressing WSTF-S158A mutant. mRNA levels of VDR-targeted genes were evaluated by quantitative RT-PCR. MCF7 cells stably expressing WSTF-wild type or WSTF-S158A mutant were treated with 10 Ϫ7 M D3 for the indicated time. The figures show the relative expression level of the indicated genes determined by the expression levels in the cells treated with vehicle control. The expression levels of the genes were normalized by the endogenous expression of the glyceraldehyde-3-phosphate dehydrogenase gene. The error bars indicate standard deviations. All measurements were done in triplicate. B, effect of WSTF-S158A mutation on histone tail modification as compared with WSTF-wild type. MCF7 cells stably expressing WSTF-wild type or WSTF-S158A mutant were treated with 10 Ϫ7 M D3. 24 h after treatment, ChIP analysis was performed with rabbit IgG (IgG), ␣-FLAG antibody (FLAG), ␣-acetyl histone H3 (AcH3), ␣-trimethylated histone H3 lysine 9 (H3K9me3), and ␣-trimethylated histone H3 lysine 4 (H3K4me3) antibodies. Immunoprecipitated chromatins were then subjected to PCR using indicated primer pairs for 25(OH)24-hydroxylase (24(OH)ase) or 25(OH)1␣-hydroxylase (1␣(OH)ase), respectively.
WSTF function at various conditions, target gene analysis might be helpful when mice selectively ablated of WSTF in hearts are available.
Considering the selective regulation of the function of two WSTF-containing complexes, MAPK-dependent phosphorylation might represent a regulatory switch for the two complexes to work properly under specific conditions (see Fig. 7D). Moreover, it is possible that combined protein modifications by other kinases or certain signaling effectors with this MAPK-dependent phosphorylation might fine-tune the two complexes to work separately at a distinct situation. A recent report suggests that WSTF acts as a tyrosine kinase in the WICH complex during DNA repair process (24). Together with our findings, intracellular signaling-induced protein modification of WSTF can modulate the enzymatic activity of WSTF itself. It is also conceivable that the modification state of WSTF defines the spe- FIGURE 7. Phosphorylation of WSTF is required for WINAC function, but not for WICH function in vivo. A, contribution of WSTF phosphorylation to co-activation function of WSTF for VDR. MEF cells from WSTF Ϫ/Ϫ mice incubated in 24-well trays were transfected with pM-VDR-DEF (15 ng), pcDNA3-WSTF, -WSTF (S158A), or empty pcDNA3 vector (45 ng), pGL3-17m2g (150 ng), and pRL-CMV (1 ng). 3 h after transfection, medium was removed and changed to fresh medium containing 1% fetal bovine serum. Treatment with D3 and ERK inhibitor (10 Ϫ7 M) was performed as described in Fig. 1A. 16 h after the treatments, cells were lysed with 100 l of lysis buffer (Promega), and luciferase activities were analyzed. The error bars indicate standard deviations. All measurements were done in a triplicate manner. B, function of WICH is not impaired by the WSTF-S158A mutant. Cell survival assays were performed as described under "Experimental Procedures." MEF cells from WSTF Ϫ/Ϫ mice were transfected with 5 g of indicated vectors, and cells were treated with 0.02% of MMS for 1 h. Surviving cells were counted 4 days after MMS treatment. Results are expressed as the means Ϯ S.D. of six independent experiments. % survival rate indicates the percentage of the cell number compared with that of the non-MMS-treated MEF cells as a control. p value was calculated by Student's t test (n ϭ 6). Single asterisk indicates p Ͻ 0.05; NS means not significant; KO, knock-out. C, formation of WICH complex is not impaired by the WSTF-S158A mutation in MEF cells. Three 10-cm cultures of wild type MEF cells were transfected with pcDNA-FLAG-WSTF or -WSTF-S158A vectors (10 g each). 24 h after transfection, cells were lysed with 1 ml of 1% Nonidet P-40 buffer and subjected to immunoprecipitation (IP) with FLAG M2-agarose (10-l bead volume), followed by Western blot (WB) using ␣-FLAG, ␣-Brg1, and ␣-Snf2h. D, schematic representation of MAPK-dependent regulation of WSTF-containing complexes. Stimulation of MAPK pathways (ERK, JNK, and p38) might serve as a switch for turning on WINAC function (SWI/SNF-type) under specific conditions. WICH (ISWI-type) can function even in the absence of the extracellular stresses stimulating MAPK signalings.