TIS7 Regulation of the β-Catenin/Tcf-4 Target Gene Osteopontin (OPN) Is Histone Deacetylase-dependent*

12-O-Tetradecanoylphorbol-13-acetate-induced sequence 7 (TIS7) acts as a transcriptional co-repressor interacting with SIN3, the histone deacetylase-containing complex. The overexpression of TIS7 down-regulates expression of a specific set of genes. Homozygous deletion of this gene in mice delays injury-induced muscle regeneration and inhibits muscle satellite cell differentiation and fusion of myoblasts in vitro. Osteopontin (OPN), a known β-catenin/T cell factor-4 (Tcf-4) downstream target gene, is up-regulated in tumors and in cells with increased motility such as muscle cells. OPN promoter sequence contains binding sites for Sp1, glucocorticoid receptor, E-box-binding factors, octamer motif-binding protein, c-Myc, and other transcription factors. Previously we have shown that TIS7 regulates the OPN expression through the inhibition of the Sp1-activating effects. Here we show that TIS7 has the capacity to inhibit OPN expression also through Lef-1, the second identified OPN regulatory element. TIS7 has the capacity to down-regulate β-catenin/Tcf-4 transcriptional activity. TIS7 homologous deletion in mouse embryonic fibroblasts increased not only the TOPflash reporter gene transcriptional activity but also the expression of c-Myc and OPN. Furthermore, we show that TIS7 overexpression leads to the β-catenin interaction with enzymatically active histone deacetylases. We propose that TIS7 down-regulates the β-catenin/Tcf-4 transcriptional activity via its interaction with histone deacetylase-containing complex thereby inhibiting the expression of β-catenin downstream target genes such as c-Myc and OPN. We hypothesize that TIS7 as a negative regulator of transcriptional activity represses expression of OPN and β-catenin/Tcf-4 target genes, which are involved in myogenesis, muscle maintenance, and regeneration in a histone deacetylase dependent manner.

The mouse tis7 (PC4) gene was identified as an immediate early gene specifically induced by tetradecanoyl phorbol acetate, epidermal growth factor, and fibroblast growth factor in mouse Swiss 3T3 cells and cultured rat astrocytes (1,2). In our previous studies we have shown that TIS7 2 is up-regulated after c-Jun activation in epithelial cells, translo-cates into the nucleus, and acts as a transcriptional co-regulator as documented on the fact that increased levels of TIS7 down-regulate the transcription of a specific set of genes. Among them, one of the most strongly down-regulated in epithelial TIS7-overexpressing cells was OPN (Ϫ2.9-fold; see Table I in Ref. 3). This result was also confirmed by an independent real-time PCR technique (-2.1-fold (3)). We have shown recently that TIS7 is capable of specific transcriptional repression in a HDAC-dependent manner, since it can interact with mSin3B and HDAC1, as well as other members of the HDAC complex (3). Using bioinformatic analysis we have further identified a common binding site for the C/EBP␣-Sp1 transcription factor "module" within the upstream regulatory regions of TIS7-regulated genes, OPN among them (4). Our experimental data showed that TIS7 inhibits the expression of target genes through its inhibitory effect on the Sp1 portion of this module (4). We have recently generated by homologous recombination mice lacking the tis7 gene (5). Disruption of the tis7 gene delayed muscle regeneration and altered isometric contractile properties of skeletal muscles after muscle crush damage in TIS7Ϫ/Ϫ mice. We found reduced expression of myogenic marker genes such as MyoD, myogenin, and laminin-␣2 in the TIS7Ϫ/Ϫ mice. In general, these mice and muscle satellite cells isolated from them display a migration, differentiation, cell cycle, and apoptosis phenotype. Therefore we wanted to know whether the OPN gene expression is up-regulated in TIS7Ϫ/Ϫ animals. Since we have previously shown that TIS7 regulates gene expression via interaction with histone deacetylases, and it was shown recently that ␤-catenin transcriptional activity is activated by the histone acetyl transferase p300 (6), our additional goal was to analyze the possible mechanism of OPN transcriptional regulation by TIS7 regulated acetylation of the ␤-catenin transcriptional complex.
OPN is a glyco-phosphoprotein that functions in cell adhesion, chemotaxis, stress-dependent angiogenesis, prevention of apoptosis, and anchorage-independent growth of tumor cells by regulating cell-matrix interactions and cellular signaling through binding with integrin and CD44 receptors. While constitutive expression of OPN exists in several cell types, induced expression has been detected in tumor cells. Substantial evidence has linked OPN with the regulation of metastatic spread by tumor cells (reviewed in Refs. 7 and 8). The molecular mechanisms that regulate the OPN expression are incompletely understood and therefore its transcriptional regulators may be of interest as potential modulators of the OPN-mediated phenotypes such as angiogenesis, apoptosis, inflammation, wound healing, tumor metastasis, adhesion, migration, and cell survival. Deletion and mutagenesis analyses of the OPN promoter region identified a proximal promoter element that is essential for maintaining high levels of OPN expression in the tumor cells. This element contains binding sites for transcription factors Sp1, the glucocorticoid receptor, the E-box-binding factors, and the octamer motif-binding protein (9,10). Furthermore, it was shown that c-Myc and OCT-1 proteins participate in forming DNA-protein complexes with the OPN-binding elements. These elements in the OPN promoter act synergistically to promote up-regulation of OPN synthesis by tumor cells (9). Since activated ␤-catenin and OPN overexpression coincide (11) in migrating cells (12), OPN may be a transcriptional target of the ␤-catenin-T cell factor (Tcf) complex (13). The c-Myc oncogene was identified as a target gene of the ␤-catenin-Tcf complex signaling pathway. Mutations of adenomatous polyposis coli cause aberrant accumulation of ␤-catenin, which then binds T cell factor-4 (Tcf-4), causing increased transcriptional activation of downstream target genes. Expression of c-Myc was shown to be repressed by wild-type adenomatous polyposis coli and activated by ␤-catenin, and these effects were mediated through Tcf-4-binding sites in the c-Myc promoter (14).
Here we identified that OPN transcription is up-regulated in cells and tissues derived from homozygous mice lacking the TIS7 gene. In transactivation assays, using a specific reporter construct and combination of histone acetyl transferase or deacetylase transfection and treatment with histone deacetylase inhibitor we show that increased TIS7 expression negatively regulates the ␤-catenin/Tcf/Lef-mediated transcriptional activation. In co-immunoprecipitation experiments we show that TIS7 possesses the capacity to increase the amount of the histone deacetylase bound to ␤-catenin. Our histone acetylase activity assays document that this ␤-catenin co-immunoprecipitated HDAC is even enzymatically active. Furthermore, in a specific gene expression analysis we have identified c-Myc, a known ␤-catenin downstream target gene to be up-regulated in TIS7Ϫ/Ϫ mouse embryonic fibroblasts. Based on these experimental data we propose that TIS7 regulates the OPN expression most probably through acetylation-dependent changes in ␤-catenin transcriptional activity affecting the c-Myc expression.

EXPERIMENTAL PROCEDURES
Cell Culture-TIS7 mouse embryonic fibroblasts were immortalized according to the NIH 3T3 protocol and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (Invitrogen), 10 mM HEPES, pH 7.3, 1% Penicillin/Streptomycin (Invitrogen), and 1% L-glutamine. The 293 cell line was from the ATCC and was cultured in minimal essential medium-␣ with 10% newborn calf serum. C2C12 cells were cultured in Dulbecco's modified Eagle's medium, supplemented with 10% heat-inactivated fetal bovine serum, 10 mM HEPES, pH 7.3, 1% penicillin/streptomycin, and 1% glutamine. Stocks of C2C12 were kept less than 30% confluent as they tend to fuse when too confluent. C2C12 stocks were grown for about six passages, and then a new vial was thawed as C2C12 cells fuse less with higher passages.
Real-time Polymerase Chain Reaction-Total RNAs were isolated using the TRizol reagent (Invitrogen). Protein samples isolated in parallel were used for Western blot analysis of protein expression. Primer sequences for the amplification of OPN were: GAC CAT GAG ATT GGC AGT GAT TTG (forward primer) and GTT TCG GTC GGA TGT AGT (reverse primer); glyceraldehyde-3-phosphate dehydrogenase (GAPDH), CGG AGT CAA CGG ATT TGG TCG TAT (forward primer) and AGC CTT CTC CAT GGT GGT GAA GAC (reverse primer); and for c-Myc, ATC TGC GAC GAG GAA GAG AA (forward primer) and ATC GCA GAT GAA GCT CTG GT (reverse primer). SYBR-Green-detected PCR was performed with the Opticon MJ Research TM (with software Opticon Monitor Version 1.08) using the reagents and protocol as published previously (19). For the amplification the following program was used: denaturation, 2 min at 95°C; amplification, 30 s at 95°C, 30 s at 55°C, and 30 s at 72°C. This cycle was repeated 39 times. Fluorescence was detected after each elongation step at an optimal temperature to obtain specific signal only. The amplification specificity was determined from the melting curve after each run. Correct amplicon size and identity were confirmed by gel analysis and sequencing. Fluorescence intensity was plotted against cycle number on a logarithmic scale. All samples were analyzed in parallel with standards. Cycle number when fluorescent signal was first detected was plotted against copy number derived from standard curves. Average data for specific PCR products were normalized to values of GAPDH and also on the PCR reaction efficiency using proprietary software from the company Biozym (supplier of the real-time PCR system Opticon).
Antibodies-Raising and purification of the rabbit antiserum to TIS7 was described previously (3). Rabbit polyclonal affinity purified ␣-Myc antibodies were from Gramsch Laboratories (Schwabhausen, Germany). Mouse monoclonal antibodies against Xpress epitope were from Invitrogen, against ␤-catenin from Transduction Laboratories, and against HA epitope tag from Babco, Inc. Mouse monoclonal ␣-HDAC1 antibodies were prepared in the laboratory of C. Seiser. Rabbit polyclonal ␣-HDAC4 antibodies were from Cell Signaling Technology.
HDAC Activity Assay-HDAC activity was determined as described previously (16,17,20) using [ 3 H]acetate-prelabeled chicken reticulocyte histones as substrate. To measure HDAC enzymatic activity equal amounts of immunoprecipitates obtained from 375 l of whole cell lysates were used. Sample aliquots of 50 l were mixed with 10 l of total [ 3 H]acetate-prelabeled chicken reticulocyte histones (1 mg/ml) yielding ϳ35,000 cpm. This mixture was incubated at 37°C for 1 h. The reaction was stopped by addition of 50 l of 1 M HCl, 0.4 M acetyl acetate, and 0.8 ml of ethyl acetate. After centrifugation at 10,000 ϫ g for 5 min an aliquot of 600 l of the upper phase was counted for radioactivity in 3 ml of liquid scintillation mixture.
Immunoprecipitation and Immunoblotting-Cells were lysed on ice in HDAC activity assay buffer as described previously (16,17,20). Lysates were sonicated and centrifuged to separate insoluble proteins. Equal amounts of soluble proteins were immunoprecipitated with 5 g of the respective antibodies followed by 10 l of (1:1) protein G-agarose slurry (Ultralink, Pierce). Following four washes with the lysis buffer, immunoprecipitated proteins were eluted with the sample buffer and boiling. Luciferase assay lysates were used for documentation of protein expression performed by analysis. Western blots for all but HDAC4 antibodies were performed as described previously (21). Western blot with HDAC4 antibodies was performed strictly according to the manufacturer's protocol.

TIS7 Regulates OPN Expression
Transient Transfections and Luciferase Assay-All transient transfections were performed using the Lipofectamine Plus Reagent (Invitrogen). pTOPflash, pFOPflash, or OPN-pGL3 luciferase reporter constructs were used as a reporter constructs. Expression constructs or empty vector DNA used as a control were co-transfected. pCMV-␤-Gal plasmid expressing ␤-galactosidase protein was used to normalize for transfection efficiency. The total amount of transfected DNA (2 g of DNA per well) was equalized by addition of empty vector DNA. Cells were harvested 48 h post-transfection in 0.25 M Tris, pH 7.5, 1% Triton X-100 buffer and assayed for both luciferase and ␤-galactosidase activities. Transfections were performed in triplicates, and all experiments were repeated several times. (3) of the transcriptional regulatory effects of the TIS7 protein we have identified a group of predominantly down-regulated genes following the overexpression of TIS7 in mouse epithelial cells. Among these the expression of OPN (secreted phosprotein I; GenBank TM accession number AA 108928) was significantly down-regulated (Ϫ2.9-fold). Second, we have identified that homozygous TIS7Ϫ/Ϫ mice display a muscle regeneration phenotype and decreased expression of certain genes, important for muscle differentiation. Therefore the OPN gene expression was tested by real-time PCR analysis in RNA samples isolated from quadriceps muscles of TIS7ϩ/ϩ and TIS7Ϫ/Ϫ, respectively. The OPN expression in 3-month-old TIS7Ϫ/Ϫ animals was 29.7-fold higher than in the control group of TIS7ϩ/ϩ animals of the same age (Fig. 1A).

TIS7 Inhibits the OPN Gene Expression-In our previous study
In the next experiment we have measured the OPN gene expression in representative clones of the TIS7ϩ/ϩ and TIS7Ϫ/Ϫ mouse embryonic fibroblasts (MEFs) (22). In Fig. 1B is shown a real-time PCR result of a representative experiment performed in TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9. The OPN gene expression was 2.5-fold increased in the TIS7Ϫ/Ϫ MEFs when compared with the TIS7ϩ/ϩ MEFs.
We were then interested to know whether the increase in OPN RNA levels is caused by up-regulation of RNA synthesis through the binding elements of specific transcription factors. Therefore, we analyzed the OPN transcriptional activity using a luciferase reporter construct driven by an OPN-specific promoter (10) transiently transfected into TIS7 MEFs (TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9). The constitutive levels of OPN transcriptional activity were 6-fold increased in TIS7Ϫ/Ϫ MEFs when compared with the TIS7ϩ/ϩ MEFs (Fig. 1C). OPN was recently identified as one of the ␤-catenin/Tcf-4 transcriptional signaling targets (13,23). To test whether TIS7 regulates OPN expression through the ␤-catenin/Tcf/Lef transcriptional pathway in addition to the previously shown C/EBP␣-Sp1 transcription factor module we tested the activity of an OPN-luciferase-reporter with mutated Tcf-binding sites. Both, mutation of a single as well as two Tcf-binding sites in the OPN promoter effectively abolished Lef-1 and Lef-1/␤-catenin enhancement of the OPN promoterreporter activity (Fig. 1C). These novel data indicated that TIS7 regulates the OPN expression mainly through the cooperative activity of the ␤-catenin/Tcf signaling pathway and only partially through the C/EBP␣-Sp1 transcription factor module (4).
According to our previously published data TIS7 expression affects not only muscle regeneration but also cell cycle progression under differentiation conditions. TIS7Ϫ/Ϫ muscle satellite cells (MSCs) in culture continue to proliferate under low mitogen (differentiation-stimulating) conditions that normally induce cell cycle withdrawal and terminal differentiation of TIS7ϩ/ϩ MSCs (5). Higher number of TIS7Ϫ/Ϫ than of TIS7ϩ/ϩ MSCs stayed in S-phase (5). Since c-Myc is an OPN promoter-binding protein regulating OPN expression (9), a known regulator of the cell cycle progression, and one of the best known ␤-catenin/Tcf-4 transcriptional targets (14), we decided to measure c-Myc expression in TIS7ϩ/ϩ and TIS7Ϫ/Ϫ MEFs. Interestingly, c-Myc gene expression, as measured by the real-time PCR, was 3-fold increased in TIS7Ϫ/Ϫ MEFs when compared with the TIS7ϩ/ϩ MEFs (Fig. 1D).
TIS7 Inhibits ␤-Catenin/Tcf/LEF-mediated Transcriptional Activation -Our previous work identified TIS7 as a transcriptional co-repressor (3). We have shown here that TIS7 inhibits OPN as well as c-Myc gene expression. Since c-Myc is a known downstream target of ␤-catenin and Tcf/Lefmediated transcriptional activation (14), we have measured the transcriptional activity of this complex in TIS7ϩ/ϩ and TIS7Ϫ/Ϫ MEF clones. Several independent MEF clones were transiently co-transfected with a Tcf-4-responsive pTOPflash luciferase reporter. The transcriptional activ- A, real-time PCR analysis of OPN expression in muscle tissue isolated from TIS7ϩ/ϩ or TIS7Ϫ/Ϫ animals, respectively. Total RNAs from the quadriceps muscle of 3-month-old animals were reverse-transcribed (RT) and PCR-amplified using combinations of specific primers. Values for the OPN gene were for each sample normalized on the housekeeping gene GAPDH expression and also on the PCR reaction efficiency using proprietary software from the company Biozym. This measurement was repeated three times, with RNA samples isolated from different animals with the same result. B, real-time PCR analysis of OPN expression in mouse embryonic fibroblasts isolated from TIS7ϩ/ϩ or TIS7Ϫ/Ϫ animals and immortalized according to the NIH 3T3 protocol. The experiment was repeated three times, using different cell clones; representative clones (TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9; both same passage number 10) are shown. C, TIS7 mouse embryonic fibroblasts (TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9) were transiently co-transfected with the wild-type or mutated OPN promoter firefly luciferase reporter plasmids cloned into the pGL3 vector (15) and ␤-galactosidase expression vectors. 48 h after the transfection luciferase and ␤-galactosidase activities were measured. D, real-time PCR analysis of c-Myc expression in TIS7ϩ/ϩ or TIS7Ϫ/Ϫ mouse embryonic fibroblasts; representative clones (TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9) are shown. Real-time PCR was normalized and evaluated as described for A. DECEMBER 2, 2005 • VOLUME 280 • NUMBER 48 ity was almost doubled in all TIS7Ϫ/Ϫ MEF when compared with the TIS7ϩ/ϩ tested clones (numbers of tested independent clones as shown in the figure legend) (Fig. 2A). The TOPflash transcriptional activity was repeatedly measured in several passages of the two representative clones TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9. One representative experiment is shown in Fig. 2B. The TOPflash transcriptional activity was 4.8-fold increased in the TIS7Ϫ/Ϫ clone #9 when compared with the TIS7ϩ/ϩ clone #3 (Fig. 2B). Furthermore we were interested to know whether TIS7 overexpression would revert the increased ␤-catenin/Tcf/Lef transcriptional activity of this clone. The cells were co-transfected with the reporter plasmid as before and additionally with ␤-catenin and TIS7 expression plasmids, respectively, as indicated in the legend of the graph. TIS7 overexpression completely abolished the ␤-catenin-induced transcriptional activity (Fig. 2C). Many features of OPN expression are cell type-specific; therefore we evaluated the transcriptional responses in C2C12 murine skeletal myoblast cell line. C2C12 cells, like the muscle satellite cells, exhibit multiple cellular potentials and readily fuse to form contractile myotubes in culture. We have co-transfected proliferating C2C12 cells with the TOPflash reporter plasmid, ␤-catenin, and TIS7 expression plasmids, respectively, as indicated in the legend of the graph. TIS7 overexpression inhibited the ␤-catenin/Tef/Lef transcriptional activity also in the C2C12 myoblasts documenting that this is not a cell line-specific effect (Fig. 2D). Moreover, additional experiments in human embryonic kidney 293 cells documented a dose-dependent effect of the TIS7-stimulated transcriptional inhibition (supplemental Fig. 1).

TIS7 Regulates OPN Expression
TIS7 Inhibits the ␤-Catenin/Tcf/LEF-mediated Transcriptional Activation in a Histone Deacetylase-dependent Manner-Tcf/Lef transcription factors are known to act in complex with histone deacetylases as transcriptional inhibitors (24). Our previous data showed that TIS7 transcriptional co-repressor activity involves similar mechanism of action, namely interaction with SIN3 complex containing the histone deacetylase enzymatic activity (3). Since, the ␤-catenin RNA and protein amounts did not differ between TIS7ϩ/ϩ and TIS7Ϫ/Ϫ muscle lysates and MEFs (data not shown), it was possible that the ␤-catenin transcriptional activity is regulated through a different mechanism. Several authors linked the acetylation of the ␤-catenin with its transcriptional activation (25,26). Therefore, we compared the effects of TIS7 with HDAC1 overexpression on the pTOPflash transcriptional activity. The 293 cells were transiently transfected with pTOPflash luciferase reporter and with the same amounts of either the TIS7 or HDAC1 expression constructs in the absence of ␤-catenin. In addition, 36 h later cells were treated with trichostatin A (TSA), a nonspecific histone deacetylase inhibitor. TIS7 overexpression caused a 4.75-fold, and HDAC1 overexpression a 1.58-fold, decrease in the pTOPflash transcriptional activity (Fig. 3A). Both effects of TIS7 or HDAC1 overexpression were reverted by the TSA treatment. These data suggested that TIS7 may regulate the ␤-catenin and Tcf/Lef-mediated transcriptional activation in a histone deacetylase-dependent manner. Furthermore, we have tested the effects of the TIS7 overexpression on the transcriptional activity induced by the ␤-catenin overexpression. Therefore 293 cells were transiently transfected with pTOPflash luciferase reporter, TIS7, ␤-catenin expression constructs and treated with TSA. TIS7 overexpression inhibited ␤-catenininduced transcriptional activity (Fig. 3B). In addition, this inhibition was overcome by the TSA treatment of the cells (Fig. 3B).
TIS7 Reverts the p300 Stimulatory Effects on the ␤-Catenin/Tcf/LEFmediated Transcriptional Activation-In the next experiment we have tested the effects of TIS7 on the p300-activated ␤-catenin/Tcf/Lef-mediated transcriptional activation. The 293 cells were transiently transfected with pTOPflash luciferase reporter and/or p300 expression construct in the presence or absence of the TIS7 expression construct. The samples were treated with TSA 36 h following the transfection. As shown in Fig. 4A, p300 overexpression significantly induced (7-fold) ␤-catenin/Tcf/Lef transcriptional activity. However, TIS7 overexpression considerably abolished the p300-induced transcriptional activity (2.5-fold instead of 7-fold induction over the control). TSA treatment of the cells further stimulated (up to 21-fold induction) the p300-induced transcriptional activity and overcame also the TIS7-mediated inhibition (12-fold instead of 2.5-fold induction) (Fig. 4A).
The same experiment was also performed in the context of overexpressed ␤-catenin. As in previous experiments, ␤-catenin overexpression significantly induced the ␤-catenin/Tcf/Lef transcriptional activity (up to 20-fold increase). Confirming these published data, p300 overexpression substantially induced (8-fold) the already up-regulated transcriptional activity of ␤-catenin (Fig. 4B). On the other hand, TIS7 overexpression inhibited not only ␤-catenin alone (5.5-fold decrease) but also ␤-catenin/p300-induced transcriptional activity (2.31-fold inhibi-FIGURE 2. TIS7 inhibits ␤-catenin/Tcf-4 transcriptional activity. A, TIS7ϩ/ϩ or TIS7Ϫ/Ϫ mouse embryonic fibroblast cells were transiently co-transfected with the pTOPflash luciferase reporter and ␤-galactosidase expression vector. 48 h after the transfection luciferase and ␤-galactosidase activities were measured. The experiment was repeated with several independent clones as indicated by the legend. Mean and standard deviations are shown. Firefly luciferase activities were normalized on transfection efficiency using the ␤-galactosidase values. B, the experiment was repeated three times with two particular MEF clones (TIS7ϩ/ϩ clone #3 and TIS7Ϫ/Ϫ clone #9); mean and standard deviation are shown. Results were normalized and evaluated as described for A. C, TIS7Ϫ/Ϫ mouse embryonic fibroblast cells (TIS7Ϫ/Ϫ clone #9) were transiently cotransfected with the pTOPflash luciferase reporter, ␤-galactosidase, ␤-catenin, and TIS7 expression vectors (3 g of plasmid DNA per well) as indicated by the legend. Results were normalized and evaluated as described for A. D, C2C12 murine skeletal myoblast cells were transiently co-transfected with the pTOPflash luciferase reporter, ␤-galactosidase, and ␤-catenin and/or TIS7 expression vectors as indicated in the legend. 48 h after the transfection luciferase and ␤-galactosidase activities were measured. Mean and standard deviations are shown. Firefly luciferase activities were normalized on transfection efficiency using the ␤-galactosidase values. tion). TSA treatment restored at least partially (back to 4.15-fold induction) ␤-catenin or ␤-catenin and p300-induced transcriptional activity inhibited by TIS7 overexpression (Fig. 4B).

TIS7 Induces the Interaction of ␤-Catenin with Enzymatically Active
Histone Deacetylases-It was shown previously that detectable levels of HDAC1 protein may be co-immunoprecipitated with ␤-catenin from cells expressing ␤-catenin, HDAC1, and LEF1 (24). However, this histone deacetylase, although present in relatively high amounts, was not enzymatically active (24). Therefore, we were wondering whether TIS7 possesses the ability to bring together an enzymatically active histone deacetylase with the ␤-catenin/Tcf/Lef complex. To test this hypothesis we co-transfected 293 cells with ␤-catenin, Lef1-HA, and HDAC1 in the absence or presence of TIS7 expression constructs. As shown in Fig. 5A, the complex co-immunopreciptated with ␤-catenin from cells, which were not co-transfected with TIS7, contained histone deacetylase enzymatic activity, which was not higher than the background signal. However, there was a significant increase in measurable amounts of histone deacetylase enzymatic activity in cell lysates from cells co-transfected with TIS7 (Fig. 5A). All proteins co-immunoprecipitated with ␤-catenin were detected by Western blot analysis using appropriate specific antibodies (Fig. 5B). These data confirmed our previous findings that TIS7 forms an enzymatically active complex containing HDAC1 (3) and second showed that ␤-catenin can interact with enzymatically active FIGURE 3. TIS7 inhibits ␤-catenin/Tcf-4 transcriptional activity. A, 293 cells were transiently co-transfected with the pTOPflash or pFOPflash luciferase reporter, ␤-galactosidase, and TIS7 expression vectors as indicated in the legend. 36 h after the transfection cells were treated with 10 ng/ml TSA or left untreated, and 12 h later luciferase and ␤-galactosidase activities were measured. Expression of exogenous TIS7 protein is documented by the Western blot (WB) analysis using antibodies detecting expression of the indicated proteins in same cell lysates as used for the luciferase assay. B, 293 cells were transiently co-transfected with the pTOPflash luciferase reporter, ␤-galactosidase, TIS7, and ␤-catenin expression vectors as indicated in the legend. 36 h after the transfection cells were treated with 10 ng/ml TSA or left untreated, and 12 h later luciferase and ␤-galactosidase activities were measured. Expression of exogenous proteins is documented by the Western blot analysis detecting expression of indicated proteins in the same cell lysate as used for the luciferase assay. . TIS7 inhibits ␤-catenin/Tcf-4 transcriptional activity in an HDAC-dependent manner. A, 293 cells were transiently co-transfected with the pTOPflash luciferase reporter, ␤-galactosidase, TIS7, and p300 expression vectors as indicated in the legend. 36 h after the transfection cells were treated with 10 ng/ml TSA or left untreated, and 12 h later luciferase and ␤-galactosidase activities were measured. The experiment was repeated at least three times; results were calculated as described in the legend to Fig. 2. B, 293 cells were transiently co-transfected with same vectors as in A and with the ␤-catenin expression vector, respectively, as indicated in the legend. TSA treatment and experiment evaluation were performed as described for A. FIGURE 5. HDAC co-immunoprecipitated with ␤-catenin is also enzymatically active. A, 293 cells were transiently co-transfected with the ␤-catenin, HDAC1, Lef-1, and TIS7 expression vectors as indicated in the legend. Cell lysates prepared 48 h later were immunoprecipitated (IP) with monoclonal anti-␤-catenin antibodies. HDAC enzymatic activity was measured following the immunoprecipitation as described under "Experimental Procedures." B, all proteins co-immunoprecipitated with ␤-catenin were detected by Western blot (WB) analysis using appropriate specific antibodies. This experiment was repeated several times with similar results; one representative experiment is shown. C, muscle cell lysates were prepared from the quadriceps muscles of TIS7ϩ/ϩ and TIS7Ϫ/Ϫ mice. Equal amounts of tissue were homogenized, and immunoprecipitations were performed as described under "Experimental Procedures" for HDAC activity assay. HDAC4 Western blot was performed according to the antibody manufacturer's protocol.

TIS7 Regulates OPN Expression
HDAC1. HDAC4, a representative of the class II HDACs, is unique in several aspects of function and regulation and is highly expressed in muscle (27). Class II HDACs interact with MEF2 transcription factors and specifically suppress myoblast differentiation. In muscle cell lysates derived from the quadriceps muscles of TIS7ϩ/ϩ and TIS7Ϫ/Ϫ mice we could co-immunoprecipitate HDAC4 with ␤-catenin (Fig. 5C). Interestingly enough, in TIS7Ϫ/Ϫ muscles less HDAC4 co-immunoprecipitated with ␤-catenin (Fig. 5C), supporting the idea that TIS7 inhibits the transcriptional activity of ␤-catenin by its interaction with specific histone deacetylases.
Our data presented here show that TIS7 negatively regulates the ␤-catenin/Tcf-4 transcriptional activity. The expression of known ␤-catenin/Tcf-4 target genes c-Myc and OPN is up-regulated in tissues and cells derived from homozygous TIS7Ϫ/Ϫ mice. The transcriptional activity of the ␤-catenin/Tcf-4 reporter gene pTOPflash is down-regulated by TIS7 overexpression, and this is reverted by the TSA treatment, which confirms the histone deacetylase-dependent manner of TIS7 inhibitory activity. Furthermore, TIS7 positively affects co-immunoprecipitation of ␤-catenin with enzymatically active histone deacetylases. Therefore we propose that TIS7 acts as a specific histone deacetylasedependent repressor of transcriptional activity inhibiting the expression of ␤-catenin/Tcf-4 target genes.

DISCUSSION
TIS7 is an immediate early gene product that upon activation translocates into the nucleus and acts as a transcriptional co-repressor. The TIS7 protein associates with a protein complex containing histone deacetylase enzymatic activity and thereby affects the expression of specific genes (3). Among these one of the most strongly regulated genes was OPN (Ref. 3 and therein supplemental Fig. 1E). Using bioinformatic analysis we identified a common binding site for the C/EBP␣-Sp1 transcription factor module within the upstream regulatory regions of TIS7-regulated genes, including OPN, and experimentally confirmed that TIS7 specifically interfered with the Sp1 transcriptional activity (4). Studies on the mouse or human OPN promoter have revealed a highly modular structure (28) involving, among others, regions responsive to Tcf components of the ␤-catenin/Tcf/Lef signaling pathway (13). Mutagenesis of the Tcf domain effectively blocks its binding to Lef-1 (10). Ectopic expression of activated ␤-catenin cooperating with the activity of growth factors induces expression of OPN (12) and thereby cell migration (11,29). ␤-Catenin interacts with the Tcf/Lef family of transcription factors and as a key component of the Wnt signaling pathway activates transcription of Wnt target genes, e.g. cyclin D and c-Myc (14). The c-Myc protein was detected to participate in formation of DNAprotein complexes on one of two newly identified regulatory elements of the OPN promoter (9). The expression of OPN, a versatile regulator of inflammation and tissue repair, was shown to be up-regulated in regenerating muscle (30). Since Wnt proteins activate expression of Myf5 and MyoD myogenic proteins (31), Wnt signaling appears to be necessary and, in some instances, sufficient to induce and maintain the myogenic program from adult stem cells (32). Therefore, modulation of the Wnt pathway might represent a promising therapeutic opportunity for the treatment of neuromuscular degenerative diseases (32). Osteopontin may represent one of important molecular targets for Wnts.
Skeletal muscle has the ability to regenerate in response to various injuries, including disease processes such as inherited muscular dystrophies. In muscle regeneration, the MSCs play the leading role. When a muscle is damaged, MSCs get activated, leave the G 0 phase of the cell cycle, proliferate, and fuse to form multinucleated myotubes. The resulting myotubes subsequently replace the damaged muscle fibers (reviewed in Refs. 33 and 34). Our previously published data showed a transient increase in TIS7 protein expression and nuclear localization in muscles during the regeneration following experimental muscle crush damage. After muscle crush damage, the force of isolated soleus muscles was significantly decreased in TIS7Ϫ/Ϫ when compared with TIS7ϩ/ϩ mice (5). Consistently, cultured primary MSCs from TIS7Ϫ/Ϫ mice displayed a marked reduction in the differentiation potential and fusion index (5). Further experiments suggested that cultured TIS7Ϫ/Ϫ MSCs continue to proliferate under low mitogen (differentiation-stimulating) conditions that normally induce cell cycle withdrawal and terminal differentiation of TIS7ϩ/ϩ MSCs (5). This raises the possibility that TIS7 via its transcriptional co-repressor function might affect ␤-catenin signaling activity and thereby muscle differentiation.
Lef-1 and ␤-catenin proteins functionally synergize in OPN promoter activity with the AP-1 activator protein, c-Jun, and the Ets transcription family proteins (10,(35)(36)(37). c-Jun N-terminal kinase-dependent phosphorylation of c-Jun results in binding of an AP-1 homodimer to the OPN promoter (38). The AP-1 (Ϫ76)-binding site is also involved in UTP-induced OPN transcription (39). ␤-Catenin synergy with PEA3/ c-Jun/Lef-1 on the activity of OPN promoter is mediated through the identified Tcf sites (10). TIS7 is increased following the induction of the c-Jun activity (3), and therefore the TIS7 expression may be regulated in a form of a negative feedback loop affecting directly or indirectly the expression of downstream target genes, such as OPN.
We have previously shown that OPN expression is down-regulated in TIS7-transfected cells. Here we show an increased OPN gene expression in muscle lysates and in MEFs derived from TIS7Ϫ/Ϫ mice, however, without transfection of any component of the regulatory pathway. These data confirm our earlier hypothesis (3) that TIS7 possesses the capacity to repress the transcription of specific genes. The OPN promoter contains among multiple regulatory regions with binding elements for various transcription factors (glucocorticoid receptor, E-box, c-Myc, OCT-1) and also a C/EBP␣-Sp1 transcription factor modulebinding site (4). In our previous study (4) we have identified that TIS7 partially, but specifically, inhibits the Sp1 stimulatory activity on the OPN promoter.
Here we have analyzed the TIS7 regulatory effects on the OPN expression using an OPN reporter construct with regulatory elements responsive to the Lef-1 and AP-1 binding (10). The activity of this reporter construct was significantly up-regulated in TIS7Ϫ/Ϫ MEFs

TIS7 Regulates OPN Expression
suggesting that TIS7 regulates OPN expression through one or several binding sites present in the region of the OPN promoter used in this reporter construct. Mutation of Lef-1-binding sites in the OPN promoter completely abolished the reporter transcriptional activity indicating that TIS7 regulates OPN expression mainly through the ␤-catenin/Tcf/Lef regulatory component of this reporter construct. These data extend our former findings showing the Sp1-dependent OPN transcriptional regulation by TIS7 (4).
␤-Catenin/Tcf-4 transcriptional complex and its downstream target c-Myc are both regulating the OPN expression (9 -13, 23). OPN expression is increased in lactating mammary gland and during tumorigenesis. In both cases OPN up-regulation goes hand in hand with the increased c-Myc expression (40,41).
Data presented here show that TIS7 inhibits, in a histone deacetylasedependent manner, the ␤-catenin/Tcf-4 transcriptional activity and OPN expression as well. Based on our recent data and results published previously by others and us, we propose a model mechanism by which TIS7 might inhibit the ␤-catenin/Tcf-4-dependent OPN expression summarized in the Fig. 6. In a response to a stimulus such as regeneration following the muscle damage TIS7 RNA and protein comcentration increases. This induces the formation of an enzymatically active histone deacetylase-containing complex that interacts with ␤-catenin and decreases its transcriptional activity (Fig. 6, ①). Several recently published papers showed that the transcriptional activity of ␤-catenin itself is acetylation-regulated (6,25,42). Our data presented here show that the complex co-immunoprecipitated with ␤-catenin from TIS7overexpressing cells contained histone deacetylase activity. Furthermore, in TIS7Ϫ/Ϫ muscles co-immunoprecipitated with ␤-catenin less HDAC4 than in wild-type mouse muscles. This finding suggests that TIS7 might regulate myogenesis by affecting the interaction between ␤-catenin and HDACs, mainly class II. Several groups have identified that ␤-catenin transcriptional activity is acetylation regulated by p300 (6,25,43,44). Our current data confirm these results, and furthermore we show that TIS7 has the capacity to inhibit ␤-catenin/Tcf-4 transcriptional activity even after strong p300 stimulation. We propose that TIS7 stabilizes or induces the formation of a ␤-catenin-histone deacetylase complex with an inhibitory activity on particular promoters, e.g. OPN.
Since TIS7 inhibits ␤-catenin/Tcf-4-regulated c-Myc expression and consequently inhibits OPN expression, this mechanism represents one branch of TIS7 regulation of OPN expression (Fig. 6, ②). Yet another mechanism of the TIS7 transcriptional inhibitory activity might be interference with the direct binding of the ␤-catenin/Tcf-4 protein complex to the Tcf-4-binding site in the OPN promoter (Fig. 6, ③). As a result of decreased ␤-catenin transcriptional activity and OPN expression TIS7 induces muscle regeneration through induced myoblast differentiation.
The fact that TIS7 co-represses expression of OPN and ␤-catenin/Tcf-4 target genes, which are involved in myogenesis, muscle maintenance, and regeneration in a histone deacetylase-dependent manner, shows the usage of histone deacetylase inhibitors as potential therapeutics for the treatment of muscle degenerative diseases from a new perspective.