Orphan Nuclear Receptor Nur77 Induces Zinc Finger Protein GIOT-1 Gene Expression, and GIOT-1 Acts as a Novel Corepressor of Orphan Nuclear Receptor SF-1 via Recruitment of HDAC2*

Kruppel-associated box (KRAB) domain-containing proteins consist of potential transcriptional repression modules. Previously, gonadotropin-inducible ovarian transcription factor-1 (GIOT-1) was identified as a novel KRAB-containing zinc finger protein and shown to have transcriptional repression activity. Here, we demonstrate that orphan nuclear receptor Nur77 regulates GIOT-1 gene expression in testicular Leydig cell lines and that GIOT-1 acts as a novel corepressor of the orphan nuclear receptor steroidogenic factor 1 (SF-1). Mutation analysis of the GIOT-1 promoter and overexpression analysis of dominant-negative Nur77 revealed that luteinizing hormone activates GIOT-1 gene expression through Nur77. Electrophoretic mobility shift and chromatin immunoprecipitation assays showed that Nur77 directly binds to the GIOT-1 promoter. GIOT-1 represses the SF-1 transactivation, and specific interaction between GIOT-1 and SF-1 was observed. We also demonstrate an interaction between GIOT-1 and histone deacetylase 2 (HDAC2). GIOT-1-mediated transrepression was recovered by down-regulation of HDAC2 expression with small interfering RNA of HDAC2. Knock down of the endogenous GIOT-1 results in significant enhancement of CYP17 expression in Leydig cells. In conclusion, this study of cross-talk between GIOT-1 and orphan nuclear receptors will provide new insights into the role of KRAB-containing zinc finger proteins in nuclear receptor action.

droxysteroid dehydrogenase type 2 transcription in adrenal zonation (5). Moreover, Nurr-1 regulates hCYP11B2 in the zona glomerulosa of the adrenal cortex (7). Previously, we reported that Nur77 expression is regulated by LH in testicular Leydig cells (2) and that the atypical orphan nuclear receptor DAX-1 interacts with Nur77 and acts as a novel coregulator of Nur77 (11). Although significant progress has been made in understanding the physiological role of Nur77 in apoptosis, its role as a transcription factor still remains largely unexplored.
Steroidogenic factor-1 (SF-1; NR5A1) is one of the major regulators of endocrine function (12) and is defined as a nuclear receptor based on structural and functional homology with other members of this family (13,14). SF-1 is expressed throughout the zones of the adrenal cortex, testis, ovary, hypothalamus, and anterior pituitary (15), where it regulates a number of genes involved in the biosynthesis of steroid hormones (16 -19) and modulates the genes encoding the insulin/insulin-like growth factor/relaxin family of hormones (20). Recently, it has been reported that NR5A members, SF-1 and human LRH-1, bind phosphatidylinositol second messengers and that an endogenous ligand is required for their maximal activity (21)(22)(23). Unlike many nuclear receptors that harbor an N-terminal ligand-independent activation domain (AF-1) and a C-terminal, ligand-dependent activation domain (AF-2), SF-1 lacks a functional AF-1 domain and depends entirely on the ligand binding domain (LBD) and DNA binding domain (DBD) (24). The transactivations of nuclear receptor are modulated by an activating ligand and interacting protein (25,26). The transcriptional activity of SF-1 is regulated by direct protein interaction with diverse coregulatory proteins. SF-1 transactivation can be either enhanced or repressed by WT-1 (27), , DAX-1 (29), DP103 (30), and ZIP67 (31), which are coexpressed with SF-1 in several tissues, including testis and ovary. The combination of interaction with these proteins might result in temporal and spatial expression of SF-1 target genes. Moreover, some SF-1 coactivators, such as CREB-binding protein and steroid receptor coactivator-1, themselves have histone acetyltransferase activity and DAX-1 recruits the corepressor nuclear receptor corepressor to repress SF-1 transactivation (32). Although SF-1 regulates various gene expression patterns during endocrine and reproductive functioning, the * This work was supported by the National Research Laboratory program (Grant mechanisms that modulate the activity of SF-1 are not well characterized. Kruppel-type zinc finger proteins belong to the largest known family of transcription factors (33,34). These proteins are characterized by Cys 2 -His 2 zinc finger motifs, often repeated in tandem, that fold around a zinc ion (35)(36)(37). About one-third of mammalian Cys 2 -His 2 zinc finger proteins contain a conserved domain of ϳ75 amino acids called KRAB (Kruppel-associated box) (38). The KRAB domain is divided into an A and B box in which the KRAB-A domain itself harbors the transcriptional repression activity and the KRAB-B box seems to be dispensable (39,40). This domain is located at the N-terminal of a Cys 2 -His 2 zinc finger protein and confers strong distance-independent transcriptional repression activity (39 -43). Moreover, the KRAB domain is a proteinprotein interaction motif and possesses transcriptional repression function (39,40). This domain is predicted to form an amphipathic helix that may interact with components of the basal transcriptional factor complex or other cellular proteins (38). Many KRAB domain zinc finger proteins are known to physically associate with a RING domain-containing corepressor protein known as KAP1 (KRAB-associated protein-1)/KRIP-1 (KRAB-A interacting protein)/TIF-1␤ (transcription intermediary factor 1␤) (42,44,45). KAP1 can enhance KRAB-A-mediated repression and silence the transcriptional activity when the KAP1-KRAB-domain protein complex directly tethers to DNA (41)(42)(43). This silencing activity may result from recruitment of histone deacetylase (HDAC) complexes, such as nuclear receptor corepressor (46). Recently, a novel KRAB domain-containing protein (RBaK) has been shown to function as a transcriptional repressor by interacting with retinoblastoma to repress E2F-dependent genes and prevent DNA synthesis (47). This protein has been identified as a novel member of the Cys 2 -His 2 zinc finger protein family named gonadotropin-inducible ovarian transcription factor-1 (GIOT-1) (48). GIOT-1 gene expression is regulated by SF-1 in ovarian granulose cells (49); however, its expression and function in other tissues are still poorly understood.
In this study, we have demonstrated that a KRAB-A domain-containing zinc finger protein, GIOT-1, is a novel target of LH-mediated Nur77 transactivation in testicular Leydig cells and that GIOT-1 can physically interact with SF-1 and repress its transactivation. Moreover, we have determined the interacting region of SF-1 and GIOT-1 and have shown that GIOT-1 can recruit HDAC2 to repress the SF-1 transactivation. Novel cross-talk between GIOT-1 and orphan nuclear receptors will provide new insights into the roles of KRAB-containing zinc finger proteins in nuclear receptor transactivation.

EXPERIMENTAL PROCEDURES
Hormone and Reagents-Ovine luteinizing hormone (LH-s-26; 2,300 IU/mg) was obtained from the National Hormone and Pituitary Distribution Program, NIDDK, National Institutes of Health (Baltimore, MD). Trichostatin A (TSA) was purchased from Sigma.
Cell Culture-HEK 293 and a mouse Leydig cell line, K28, were maintained in Dulbecco's modified Eagle's medium in the presence of 10 and 15% fetal bovine serum (Invitrogen), respectively. The rat Leydig cell line, R2C, was maintained in F10 medium supplemented with 15% horse serum and 2.5% fetal bovine serum. All cells were incubated at 37°C in 5% CO 2 .
Northern Blot Analysis-Northern blot analysis was performed using GIOT-1 cDNA as a probe by following procedures described previously (11).
Transient Transfection and ␤-Galactosidase Assay-Cells were plated in 24-well plates 24 h before transfection and transfections were carried out with SuperFect reagent (Qiagen) or Lipofectamine Plus reagent (Invitrogen), according to the manufacturer's instructions. Total DNA used in each transfection was adjusted by adding appropriate amounts of pcDNA3 vector. Approximately 48 h post-transfection, cells were harvested, and the luciferase activity was measured as described previously (11) and normalized against ␤-galactosidase activity as an internal control.
Infection of Adenovirus and RT-PCR-The recombinant adenovirus of Nur77 (AD-Nur77) was prepared as previously described (53). R2C cells were infected with AD-Nur77 or adenovirus alone. Twenty-four hours after infection, cells were extracted for RT-PCR for GIOT-1 (30 cycles), for Nur77 (25 cycles), and for ␤-actin (25 cycles) as a control.
Electrophoretic Mobility Shift Assays-EMSA was performed as described previously (2). Briefly, cultured K28 cells were washed twice with cold phosphate-buffered saline and pelleted by centrifugation at 3,000 rpm for 5 min at 4°C. The pellets were gently resuspended in buffer (20 mM HEPES, 10 mM EDTA, 0.1% Nonidet P-40, 100 mM NaCl, 0.15 M phenylmethylsulfonyl fluoride, leupeptin, and pepstatin) and broken by passing them through a 25-gauge needle 20 times. The cells were kept on ice for 1 h at 4°C and centrifuged at 14,000 rpm at 4°C for 30 min. The supernatants containing the whole cell extracts were aliquoted and stored at Ϫ80°C. Probes used for EMSA experiments were prepared by labeling 10 pmol of double-stranded oligonucleotides with T4 polynucleotide kinase (Promega Corp.) at 37°C for 30 min. The labeled probes were purified by Sephadex G-50 column chromatography. A sample containing 40,000 -50,000 cpm of the purified doublestranded oligonucleotides was used for each reaction. EMSA was performed with 1 g of poly(dI/dC)/sample as a nonspecific competitor. The DNA-protein complexes were separated from the unbound DNA probe via 6% nondenaturing gel electrophoresis at 4°C in Tris base glacial acetic EDTA buffer, and the binding reaction was carried out at 25°C for 30 min. The sequences of oligonucleotides used as probes for GIOT-1 and mutated GIOT-1 were 5Ј-CTAGGACAAGCCTGTGAC-CTTTCT-3Ј and 5Ј-CTAGGCAAGCCTGTGAaaTTTCT-3Ј.
GST Pull-down Assay-[ 35 S]Methionine-labeled proteins were prepared using pcDNA3-HA vectors containing cDNAs encoding for fulllength and deletion constructs of SF-1, DAX-1, SHP, and Nur77, and the TNT-coupled transcriptional translation system with conditions as described by the manufacturer (Promega). GST-fused GIOT-1 and GST alone was expressed in Escherichia coli BL21(DE3) strain and purified using glutathione-Sepharose 4B beads (Amersham Biosciences). In vitro protein-protein interaction assays were carried out as described previously (11,50). For interaction between GIOT-1 and HDACs, GSTfused GIOT-1 and GST alone were expressed in E. coli BL21(DE3) strain and purified using glutathione-Sepharose 4B beads (Amersham Biosciences). Recombinant GST or GST fusion proteins were incubated with pre-washed glutathione-Sepharose 4B beads (Amersham Biosciences) while shaking for 30 min at room temperature and washed three times for 5 min at 4°C with phosphate-buffered saline. To test for GIOT-1 and HDAC interaction, a GST-pull down assay was carried out with 1 ϫ 10 7 K28 Leydig cells that were harvested, washed, and lysed in NP lysis buffer (50 mM Tris, pH 8.0/150 mM NaCl/5 mM EDTA/1% Nonidet P-40/1 mM PMSF/protease inhibitors). K28 cell extract was incubated with GST or GST-GIOT-1 fusion proteins bound to glutathione beads at 4°C overnight. The beads were washed three times for 5 min with 1 ml of Nonidet P-40 lysis buffer, and proteins bound to the beads were eluted into sample buffer. For Western blotting, proteins were separated by 4 -12% SDS-PAGE, transferred to polyvinylidene difluoride (Invitrogen), and incubated with primary polyclonal rabbit HDAC antibodies (HDAC1, -2, -3, -5, -6, and -8, Zymed Laboratories Inc.). After extensive washing, blots were visualized by chemiluminescence.

LH and Nur77
Regulate GIOT-1 Promoter Activity-A recent study showed that GIOT-1 gene expression is rapidly and strongly induced by Cells were treated with LH or without LH and assayed for Luciferase activity after 24 h. Transfection efficiency was normalized using ␤-galactosidase activity. Data are the mean Ϯ S.E. values of at least triplicate assays. pregnant mare serum gonadotropin or human chorionic gonadotropin in a testicular Leydig cell line (48), which led us to investigate the mechanism mediating the stimulatory effect of LH on GIOT-1 gene expression in Leydig cells. First, we examined whether GIOT-1 promoter activity is regulated by LH. To map the promoter region of GIOT-1, which is required for LH-mediated activation, a series of 5Ј deletions of GIOT-1 promoter fused to a reporter gene were used in a transient transfection assay with or without LH treatment. As shown in Fig. 1, LH-mediated induction of GIOT-1 promoter activity was retained with deletion of the promoter up to Ϫ161 but lost with the Ϫ122 promoter. These results indicate that the sequences required for the LH response lie in the region between Ϫ161 and Ϫ122, which contains a potential Nur77 binding site (NBRE, AAAGGTCA).
To further investigate a role for the Nur77 protein in the LH-mediated stimulation of the GIOT-1 promoter, we cotransfected a Nur77 expression plasmid with a series of GIOT-1 promoter deletion constructs. We observed a significant stimulation of GIOT-1 promoter activity with deletion up to Ϫ161 following coexpression of Nur77 (Fig.  1C). However, a marked decrease in Nur77-mediated stimulation of the GIOT-1 promoter was observed upon deletion to Ϫ122. The results from Fig. 1 (B and C) suggest that Nur77-mediated transactivation of the GIOT-1 promoter may be directly mediated through the NBRE site.
To address whether the potential NBRE site in the GIOT-1 promoter is responsible for Nur77-mediated activation of the GIOT-1 promoter, an NBRE mutant (AAAGGTCA to AAATTTCA) within the Ϫ282 GIOT-1 promoter construct (MT-282) was used in transient transfection assays. The participation of Nur77 in the induction of GIOT-1 expression by LH was also confirmed by using a dominant-negative mutant of Nur77 (DN-Nur77), which completely inhibits the transactivation activity of not only Nur77, but also other members of the Nur77 family, including Nurr-1 and NOR1 (10). K28 cells were transfected with either a DN-Nur77 expression vector or the corresponding empty vector and were treated with or without LH. As shown in Fig. 1D, the strong induction of the wild-type GIOT-1 promoter (WT-282) activity by LH was inhibited by DN-Nur77 in a dose-dependent manner. However, LH treatment had no significant effect on MT-282 GIOT-1 promoter activity. These results indicate that Nur77 is necessary to induce GIOT-1 expression by LH in testicular Leydig cells.
Activation of GIOT-1 Promoter by Nur77, but Not by SF-1, in the K28 Leydig Cell Line-A consensus Nur77 binding motif (NBRE), 5Ј-AAAG-GTCA-3Ј, was identified in the region spanning the nucleotide Ϫ161 to Ϫ122 of the GIOT-1 promoter. This potential Nur77 binding motif was previously identified as a SF-1 binding sequence in rat granulosa cells (49). To determine whether the putative NBRE sequence element in the GIOT-1 promoter confers the responsiveness to Nur77, Nur77-or SF-1-induced WT-282 and MT-282 reporter activities were compared in K28 cells. Cotransfection with a Nur77 expression vector strongly induced WT-282 promoter activity in a dose-dependent manner ( Fig.  2A). Interestingly, cotransfection of the WT-282 reporter with a SF-1 expression vector produced only a modest stimulation of GIOT-1 promoter activity. However, cotransfection of the MT-282 reporter with Nur77 resulted in a complete loss of activation of the GIOT-1 promoter. Meanwhile, we previously reported that the atypical orphan nuclear receptor DAX-1 interacted with Nur77 and inhibited its transactivation (11). To test whether DAX-1 is also able to inhibit the Nur77-mediated induction of GIOT-1 promoter activity, increasing amounts of DAX-1 expression vector were cotransfected with a fixed amount of Nur77, along with the GIOT-1 reporter construct. Cotransfection of DAX-1 significantly decreased the GIOT-1 promoter activity induced by Nur77 ( Fig. 2A).
It has been reported that the Nur77 subfamily members, Nurr-1 and NOR1, also transactivate the NBRE-containing promoter, and they are induced by LH treatment in the testis (2). To explore the possibility that Nurr-1 or NOR1 is also involved in GIOT-1 gene expression, we cotransfected other Nur77 subfamily members with the GIOT-1 promoter reporter. Although the degree of activation was variable among Nur77 family members, GIOT-1 gene expression was increased by both Nurr-1 and NOR1 (Fig. 2B). These results suggest that the Nur77 family plays a major role in GIOT-1 gene transcription in the testis.
LH and Nur77 Induce GIOT-1 Gene Expression in Testicular Leydig Cell Lines-Our previous study demonstrated that LH rapidly increases the expression of Nur77 mRNA in testicular Leydig cells (2). Moreover, GIOT-1 expression has been well documented in ovarian granulosa cells (48). However, the hormonal regulation of GIOT-1 expression in Leydig cells has not been addressed. Therefore, we analyzed the regulation of GIOT-1 expression by LH in a testicular Leydig cell line, K28. Northern blot analysis revealed that GIOT-1 mRNA was barely detectable basally. However, the GIOT-1 mRNA level was significantly increased within 30 min, peaked at around 1 h, and then returned to basal levels at 3 h after LH treatment.
Based on the observation that overexpression of Nur77 stimulates the activity of the GIOT-1 promoter in transient transfection assays (Fig.  1B), we examined whether adenovirus-mediated Nur77 overexpression could increase the transcription of the GIOT-1 gene. Following infection with or without Nur77-expressing adenovirus (AD-Nur77), RT-PCR was performed to detect GIOT-1 mRNA in rat Leydig R2C cells, which are constitutively steroidogenic and express Nur77 (54). As shown in Fig. 3B, Nur77 overexpression significantly increased the level of GIOT-1 gene expression. These results demonstrate that Nur77 induces GIOT-1 gene expression.
Nur77 Directly Binds and Activates the GIOT-1 Promoter-To investigate whether LH-mediated induction of GIOT-1 gene expression involves DNA binding of Nur77, we performed EMSAs. Nuclear extracts were prepared from mouse testicular Leydig K28 cells with or without LH treatment for 1 h. As shown in Fig. 4A, an intense shifted band was observed in the LH-treated lane. This protein-DNA complex was completely abolished with 50-fold excess of unlabeled specific NBRE probe but not with 50-fold excess of oligomer containing the mutated NBRE motif. To determine whether Nur77 is present in this complex, we preincubated whole cell extracts with an antibody against Nur77. The addition of anti-Nur77 prevented the formation of the LHinduced Nur77-DNA complex, whereas anti-SF-1 had no effect.
To determine whether Nur77 or SF-1 actually binds to the GIOT-1 promoter in Leydig cells in vivo, we performed chromatin immunoprecipitation assays with rat Leydig R2C cells. As shown in Fig. 4B, PCR amplification of a region, from Ϫ206 to ϩ44 of the GIOT-1 promoter, containing a NBRE motif, indicated Nur77 recruitment to the promoter region in vivo. Similarly, SF-1 was also bound to the NBRE-containing region of the GIOT-1 gene promoter but relatively weakly compared with Nur77. No signal was detected from the control PCR for nonspecific immunoprecipitation using anti-HA antibody. In addition, primers to the ␤-actin coding region did not produce any amplified bands, demonstrating the specificity of the immunoprecipitations. These results suggest that Nur77 directly binds to the GIOT-1 promoter, activating its gene expression. (48). To determine whether GIOT-1 plays a regulatory role in the transactivation of orphan nuclear receptors, we performed transient transfection assays with Nur77 or SF-1 along with GIOT-1 expression vectors in K28 cells. Coexpression of increasing amounts of GIOT-1 with a constant amount of SF-1 caused progressive repression of SF-1-mediated transactivation in a dose-dependent manner (Fig. 5A). However, GIOT-1 coexpression showed no significant effect on the Nur77 transactivation (Fig. 5B).

GIOT-1 Functions as a Corepressor of Orphan Nuclear Receptor SF-1-Previous studies have demonstrated that GIOT-1 has a potent corepressor activity
To determine whether the inhibitory effect of GIOT-1 on SF-1-mediated transcription is due to direct repression of SF-1 activity, we used a SF-1 protein fused to Gal4-DBD (Gal4-SF-1). In HEK293 cells, Gal4-SF-1 increased by ϳ15-fold the activity of the reporter gene driven by Gal4 binding sites. Interestingly, coexpression of GIOT-1 significantly decreased Gal4-SF-1 transactivation in a dose-dependent manner (Fig.  5C). In contrast, GIOT-1 was unable to repress the transcriptional activity of Gal4-Nur77 fusion protein. Taken together, these results demonstrated that GIOT-1 directly inhibits the transcriptional activity of SF-1 but not that of Nur77.  Labeled oligonucleotide was incubated with 10 g of nuclear extracts from LH-treated (LH) or LHuntreated (CTL) K28 cells. Cells treated with LH for 2 h (LH) were incubated with the GIOT-1 probe with a 25-or 50-fold excess of unlabeled GIOT-1 probe competitor (S.C.) and a 25-or 50-fold excess of unlabeled mutated GIOT-1 probe (N.S.C.), and 2 g of Nur77-specific antibody (Nur77-Ab) or SF-1specific antibody (SF-1-Ab) was added to the reaction mixture before the addition of labeled probe. Free indicates running of labeled probe only, and the arrow indicates the Nur77-DNA complex. B, Nur77 was recruited to the endogenous GIOT-1 promoter in the Leydig cells. Chromatin immunoprecipitation assays were performed with R2C cells. Anti-Nur77 and anti-SF-1 or anti-HA antibodies were used for immunoprecipitations, and the immunoprecipitates were analyzed by PCR using a pair of specific primers spanning a region containing the Nur77/SF-1 binding site of the GIOT-1 gene promoters. A control PCR for nonspecific immunoprecipitation was done using primers specific for the ␤-actin coding region. Reproducible results for association with the promoter for Nur77 and SF-1 were shown from three independent assays, respectively.

GIOT-1 Directly Interacts with SF-1-
To determine whether GIOT-1 directly interacts with SF-1, we performed GST pull-down assays. [ 35 S]Methionine-labeled SF-1, DAX-1, SHP, and Nur77 produced by in vitro translation were incubated with the GST fusion protein of GIOT-1 or GST alone. As shown in Fig. 6A, GIOT-1 specifically interacted with SF-1 and not with other nuclear receptors. GST alone did not interact with SF-1. These results are consistent with the data from functional analyses by transient transfection assays. The region of SF-1 responsible for the interaction with GIOT-1 was also assessed by GST pull-down experiments. [ 35 S]Methionine-labeled SF-1 and its deletion mutants (Fig. 6B) produced by in vitro translation were allowed to bind the GSTfused GIOT-1 (GST-GIOT-1) or GST. GST-GIOT-1 interacted with SF-1 LBD as well as full-length SF-1 but not with SF-1DBD or SF-1 ⌬AF2 (Fig. 6C). These results suggest that GIOT-1 directly interacts with SF-1 through the AF-2 domain of SF-1.
It has been reported that the N-terminal repeat region of GIOT-1, which contains a KRAB-A domain, has transcriptional repression (48). Therefore, we tested two parts of GIOT-1, the N-terminal KRAB-A domain (GIOT-1-KRAB-A) and C-terminal zinc finger domain (GIOT-1-zinc finger), for interaction with SF-1 (Fig. 6D). GST pull-down assays revealed that GIOT-1-KRAB-A retained the ability to interact with SF-1 to the same extent as full-length GIOT-1 (Fig. 6E), suggesting that the N-terminal KRAB-A domain of GIOT-1 is responsible for the SF-1 interaction.
Because the GIOT-1-KRAB-A domain is required for the interaction with SF-1, we investigated whether GIOT-1-KRAB-A alone could repress SF-1-mediated transactivation. Transient transfection assays were performed with SF-1 in the presence of GIOT-1-KRAB-A or GIOT-1-zinc finger (Fig. 6F). As expected from interaction assays, coexpression of the GIOT-1-zinc finger failed to reduce SF-1-mediated transactivation, whereas GIOT-1-KRAB-A caused progressive suppression of SF-1-mediated transactivation. These results suggest that GIOT-1-KRAB-A is involved in both the interaction with SF-1 and the repression of its transactivation.
HDAC2 Is Required for the Repressive Activity of GIOT-1-It has been reported that TSA, a specific inhibitor of HDACs (55), interferes with GIOT-1 repression of TIF1␤ function (48). To investigate the detailed mechanism of GIOT-1 action on SF-1 function, we first assessed the involvement of HDACs on the repressive activity of GIOT-1 using TSA and mammalian one-hybrid assays. Cotransfection of Gal4-GIOT-1 with Gal4-tk-Luc reporter resulted in a decrease in luciferase activity compared with Gal4 alone. However, TSA treatment alleviated the GIOT-1-mediated repression (Fig. 7A). We also examined the TSA interference of GIOT-1 repression activity on SF-1 transactivation by transient transfection analyses using a reporter gene. SF-1 transactivation was markedly increased in the cells treated with TSA (Fig. 7B). The magnitude of the transactivation observed in the presence of TSA (100 nM) was more than two times of that observed in the absence of TSA. As expected, the amelioration of GIOT-1-mediated repression of SF-1 transactivation was observed with TSA treatment at all attempted levels of GIOT-1 expression. These results suggest that the repressive activity of GIOT-1 involves the recruitment of HDAC.
To further investigate the role of GIOT-1 in SF-1-mediated gene activation, we examined the mRNA expression of CYP17, a well known SF-1 target gene, in testicular Leydig cells after knock down of endogenous GIOT-1 with GIOT-1 siRNAs (Fig. 7F). Of four tested siRNAs, each targeted to a different region of GIOT-1, siGIOT-1-II, and siGIOT-1-III were found to knock down the expression of GIOT-1 in R2C cells, as monitored by quantitative real-time RT-PCR. The repressive effects of siGIOT-1-II and siGIOT-1-III occurred in a dose-dependent manner. Interestingly, reduction of GIOT-1 levels with siGIOT-1-II and siGIOT-1-III enhanced the expression of CYP17 in R2C cells. These results provide strong evidence that endogenous GIOT-1 plays a role in the modulation of SF-1 transcriptional activity in testicular Leydig cells. Taken together, these results suggest that GIOT-1 represses SF-1 transactivation in vivo by recruiting HDAC2.

DISCUSSION
The stimulatory effect of FSH on the gene expression of GIOT-1 in the ovarian granulosa cells has been previously characterized (49). However, the regulation of GIOT-1 expression in testicular Leydig cells has not been studied yet. In addition, the cellular function of GIOT-1 remains uninvestigated. In this report, we have identified the GIOT-1 gene as a novel target of the orphan nuclear receptor Nur77. We have demonstrated that LH stimulated GIOT-1 expression in testicular Leydig cells via Nur77, which binds to the specific response element in the GIOT-1 promoter. Furthermore, we demonstrate that GIOT-1 functions as a corepressor of SF-1 by recruiting HDAC2.
In contrast to other transcription factors that are constitutively expressed and activated by post-translational modification, the gene expression of Nur77 is tightly regulated by extracellular signals, whereas its post-translational modification is also important for its function (56). Because Nur77 has been reported to have high selectivity for its response element compared with other nuclear receptors, our results strongly suggest that Nur77 is a major regulator of GIOT-1 gene expression in testicular Leydig cells, although SF-1 induces GIOT-1 gene expression in ovarian granulosa cells. In contrast to SF-1, which is constitutively active in many steroidogenic tissues, including testicular Leydig cells and binds to an element similar to, but distinct from, Nur77 (57), Nur77 is present at low levels under basal conditions and becomes rapidly induced in response to external signals. This suggests that LHmediated GIOT-1 gene induction most likely involves Nur77 rather than SF-1. Moreover, our observations of Nur77 recruitment to the GIOT-1 promoter and the enhancement of GIOT-1 gene expression upon adenovirus-mediated overexpression of Nur77 support the hypothesis that Nur77 plays an important role in GIOT-1 gene transcription.
It has been reported that SF-1 binds to the GIOT-1 promoter and that phosphorylation of SF-1 by protein kinase A increases its ability to enhance GIOT-1 gene expression (49). Thus, it is likely that activated protein kinase A phosphorylates SF-1 leading to enhanced transcription of GIOT-1. We previously showed that, in testicular Leydig cells, LH treatment rapidly and dramatically increased the levels of Nur77 mRNA and protein, and protein kinase A increased Nur77 expression (2). In addition, both Nur77 and SF-1 have been implicated in the regulation of 21-hydroxylase (CYP21) transcription (58). Furthermore, Nur77 protein level is significantly elevated in SF-1 Ϯ adrenal (59). Together, these findings suggest that a balanced function of two proteins may be important for GIOT-1 gene expression. Nevertheless, Nur77 and SF-1 are not functionally equivalent, because subtle changes of the consensus DNAbinding sequence affect the transactivation of Nur77 and SF-1 differentially. Therefore, we propose that the effects of LH on GIOT-1 tran-scription occur through two possible pathways; increasing Nur77 gene expression and activating SF-1 through phosphorylation, although SF-1 alone did not significantly stimulate GIOT-1 promoter activity ( Fig. 2A).
In this report, we demonstrate that GIOT-1 is a novel corepressor of SF-1. The KRAB-A domain of GIOT-1 exhibits transcriptional repression activity, which is consistent with the function of the KRAB-A domain described previously (39,40). We also demonstrate that GOIT-1 interacts with SF-1 via the AF-2 domain, repressing its transactivation. These results indicate that the direct interaction between SF-1 and GIOT-1 is independent of additional factors. Although the possibility that other domains of SF-1 also participate in the binding of GIOT-1 cannot be excluded, our results clearly showed that the SF-1 AF-2 domain is an absolute requirement for the interaction with GIOT-1.
A functional connection between GIOT-1-mediated repression and HDAC activity was implied when the HDAC inhibitor, TSA, blocked GIOT-1 repression. In this study, we demonstrate that GIOT-1 can recruit HDAC2 and that the transcriptional repression of GIOT-1 is dependent on HDAC2 activity. This is comparable with the report that the transcriptional repression mediated by KRAB-A-KAP1 may result from association with members of the HP1 family (60 -63), a family of non-histone heterochromatin-associated proteins with an established gene-silencing function (64).
Although several other C2H2 zinc finger proteins were previously found to directly interact with SF-1 and modulate its transcriptional activity (3,31), we provide the first evidence that the KRAB-A containing zinc finger protein, GIOT-1, acts as corepressor of SF-1. Moreover, our results elucidate a molecular mechanism for GIOT-1-mediated repression of SF-1 transactivation, which is linked to the recruitment of the HDAC2 component. Based on our current observations, we propose that GIOT-1 recruits HDAC2 to SF-1 by interacting with the AF-2 domain of SF-1, and that HDAC2 is subsequently able to repress the transactivation activity of SF-1 (Fig. 7). However, the possible participation of components of other corepressor complexes in the GIOT-1mediated SF-1 repression remains to be determined. In addition, further study will be required to assess the generality of the inhibitory mechanism of GIOT-1 and determine whether other mechanisms may contribute to the inhibitory effect of GIOT-1.
In conclusion, here we provide direct experimental evidence that orphan nuclear receptor Nur77 is involved in LH-mediated GIOT-1 gene expression in testicular Leydig cells and that GIOT-1 acts as a novel corepressor of the orphan nuclear receptor SF-1. To our knowledge, GIOT-1 is the first member of the KRAB-containing protein reported to interact with SF-1. The identification of HDAC2 involvement also provides new insight into the molecular mechanism by which the corepressor GIOT-1 mediates its repression. GIOT-1-mediated repression of SF-1 transactivation may play an important role in the regulation of SF-1 target genes during development and function of reproductive and steroid-producing tissues. Studies are currently underway to assess these possibilities.