Orphan Nuclear Receptor DAX-1 Acts as a Novel Corepressor of Liver X Receptor α and Inhibits Hepatic Lipogenesis*

DAX-1 (dosage-sensitive sex reversal adrenal hypoplasia congenital critical region on X chromosome, gene 1) is a member of the nuclear receptor superfamily that can repress diverse nuclear receptors and has a key role in adreno-gonadal development. Our previous report has demonstrated that DAX-1 can inhibit hepatocyte nuclear factor 4α transactivity and negatively regulate gluconeogenic gene expression (Nedumaran, B., Hong, S., Xie, Y. B., Kim, Y. H., Seo, W. Y., Lee, M. W., Lee, C. H., Koo, S. H., and Choi, H. S. (2009) J. Biol. Chem. 284, 27511–27523). Here, we further expand the role of DAX-1 in hepatic energy metabolism. Transfection assays have demonstrated that DAX-1 can inhibit the transcriptional activity of nuclear receptor liver X receptor α (LXRα). Physical interaction between DAX-1 and LXRα was confirmed Immunofluorescent staining in mouse liver shows that LXRα and DAX-1 are colocalized in the nucleus. Domain mapping analysis shows that the entire region of DAX-1 is involved in the interaction with the ligand binding domain region of LXRα. Competition analyses demonstrate that DAX-1 competes with the coactivator SRC-1 for repressing LXRα transactivity. Chromatin immunoprecipitation assay showed that endogenous DAX-1 recruitment on the SREBP-1c gene promoter was decreased in the presence of LXRα agonist. Overexpression of DAX-1 inhibits T7-induced LXRα target gene expression, whereas knockdown of endogenous DAX-1 significantly increases T7-induced LXRα target gene expression in HepG2 cells. Finally, overexpression of DAX-1 in mouse liver decreases T7-induced LXRα target gene expression, liver triglyceride level, and lipid accumulation. Overall, this study suggests that DAX-1, a novel corepressor of LXRα, functions as a negative regulator of lipogenic enzyme gene expression in liver.

Nuclear receptors are a class of DNA binding transcription factors that regulate gene expression and play important roles in a variety of biological and pathological processes (2). Orphan nuclear receptor DAX-1 (NROB1) is an unusual member of the nuclear receptor superfamily (3). The C-terminal region has the structure that is characteristic of a ligand binding domain (LBD). 5 Unlike other nuclear receptors, the N-terminal region does not have any classical DNA binding domain. However, it has three short repeats (65-67 amino acids) each containing an LXXLL-related motif. Duplication of the DAX-1 gene is associated with male to female reversal in XY individuals, and mutations in DAX-1 are responsible for adrenal hypoplasia congenita, an inherited disorder of adrenal gland development (4). DAX-1 generally acts as a negative regulator to repress the transcriptional activity of receptors such as estrogen receptor, thyroid receptor ␤, steroidogenic factor (SF-1), androgen receptor, estrogen receptor-related receptor ␥, glucocorticoid receptor, nerve growth factor-inducible gene B (Nur77), and peroxisome proliferator-activated receptor ␥ (PPAR␥) (5)(6)(7)(8)(9)(10)(11)(12). We have previously reported that DAX-1 can negatively regulate the expression of gluconeogenic genes by inhibiting the transcriptional activity of hepatocyte nuclear factor 4 ␣ (HNF4␣) (1). DAX-1 is also known to interact with and inhibit the transcriptional activity of transcription factors, including OCT3/4 (13). DAX-1 has been shown to compete with nuclear receptor coactivators such as PGC-1␣ (11), GRIP-1 (10), and SRC-1 (11), and it is also known to recruit corepressors such as NCoR and Alien (9). A recent report showing the three-dimensional structure of DAX-1 reveals that Dax-1 could function as a ligand-independent nuclear receptor as well as a competitive transcriptional corepressor (14). Liver X receptor ␣ (LXR␣) is a member of the nuclear receptor superfamily that heterodimerizes with retinoid X receptor (RXR). LXR␣/RXR heterodimers bind to DR-4-type response elements known as the LXR-response elements (LXRE) in their target genes. LXR␣ is abundantly expressed in tissues with active lipid metabolism, such as white adipose tissue, liver, intestine, and macrophages, whereas the LXR␤ isoform is more ubiquitously expressed (15). Both LXR␣ and LXR␤ are stimulated by several natural and synthetic ligands, including (20S)hydroxycholesterol, (22R)-hydroxycholesterol, 24-hydroxycholesterol, T0314407, T0901317, and GW3965 (16). It has been reported that the p160 coactivator SRC-1 and p300 can bind to the ABCA1 promoter via the oxysterol-induced LXR␣/RXR heterodimer and cause maximal activation of promoter (17). A recent study has demonstrated that SIRT1 deacetylates and activates the nuclear receptor LXR␣ by favoring its ligand-dependent proteasomal degradation, thereby potentially regulating reverse cholesterol transport (18). In the unliganded state, LXR␣ preferentially associates with corepressors such as the nuclear receptor corepressor (NCoR) and silent mediator of retinoic acid receptor and thyroid receptor (SMRT) (19). Furthermore, decreased expression of NCoR has been shown to increase the expression of adipocyte-specific genes (20), and the recruitment of NCoR has also been shown to modulate LXR signaling in liver (21).
In liver, LXR is involved in transcriptional control of Cyp7A1, encoding a critical enzyme in the conversion of cholesterol into bile acids, as well as ABCG5/ABCG, encoding ABC transporters implicated in biliary cholesterol excretion. Induction of intestinal ABCA1, ABCG5, and ABCG8 expression upon LXR activation accelerates fecal cholesterol disposal by reducing the efficiency of cholesterol absorption. LXR has also been reported to control genes that encode proteins involved in de novo lipogenesis. In particular, LXR is known to induce the expression of SREBP-1c, a transcription factor that regulates the expression of various lipogenic genes, including those encoding acetyl-coenzyme A carboxylase and fatty-acid synthase (FAS) (22). LXR is also known to regulate the expression of carbohydrate-response element-binding protein (ChREBP) (23). In addition, LXR is known to directly influence the transcription of genes encoding fatty acid synthase, lipoprotein lipase, cholesterol ester transfer protein, and stearoyl-CoA desaturase 1. Moreover, activation of LXRs by agonistic compounds induces the expression of enzymes involved in the synthesis of fatty acids in liver cells (24).
In this study, we show that DAX-1 inhibits the transcriptional activity of LXR␣ through direct interaction and competition with coactivator SRC-1. Overexpression of DAX-1 decreases the expression of Srebp-1c, FAS, and acetyl-coenzyme A carboxylase, whereas knockdown of endogenous DAX-1 increases the LXR␣ target gene expression. Finally, hepatic overexpression of DAX-1 decreases LXR␣ agonist-induced liver triglyceride level and lipid accumulation in mouse. Collectively, this study demonstrates that DAX-1 represses the transcriptional activity of LXR␣ to control the lipogenesis.
Preparation of Recombinant Adenovirus-For the ectopic expression of the transgene, the adenoviral vector systems were used as described previously (26). Briefly, the cDNA encoding FLAG-DAX-1 was cloned into the pAdTrack shuttle vector. The FLAG-DAX-1 fragment was digested with KpnI/XhoI and was cloned into the KpnI/XhoI site of the pAdTrack-CMV vector. Recombination of AdTrack-CMV-FLAG-DAX-1 with adenoviral gene carrier vector was performed by transformation into pretransformed AdEasy-BJ21-competent cells. Recombinant adenoviruses expressing GFP only or unspecific RNA interference control were described earlier. Oligonucleotides for DAX-1 small hairpin RNA (ctgtaccgctgctgcttctgcggagaa) were synthesized by IDT (Coralville, IA). Adenovirus for small hairpin DAX-1 was generated as described previously (1).
Cell Culture and Transient Transfection Assay-HepG2, 293T, and HeLa cells were maintained in Dulbecco's modified Eagle's medium, and AML12 cells were maintained in Dulbecco's modified Eagle's medium/F-12 (Invitrogen), supplemented with 10% fetal bovine serum (Cambrex Bio Science Walkersville, Inc., Walkersville, MD) and antibiotics (Invitrogen). Cells were split in 24-well plates at densities of 2-8 ϫ 10 4 cells/well the day before transfection. Transient transfections were performed using the SuperFect transfection reagent (Qiagen, Valencia, CA) according to the manufacturer's instruction. Cells were cotransfected with indicated reporter plasmids together with expression vectors encoding various transcription factors. Total DNA used in each transfection was adjusted to 1 g/well by adding appropriate amount of empty vector, and CMV-␤-galactosidase plasmids were cotransfected as an internal control. Cells were harvested ϳ40 -48 h after the transfection for luciferase and ␤-galactosidase assays. The luciferase activity was normalized with ␤-galactosidase activity. Knockdown of hDAX-1 by small interfering RNA and luciferase assays was performed as described previously (1).
Isolation and Culture of Primary Hepatocyte and Animal Experiments-Primary rat hepatocytes were prepared from 200-to 300-g Sprague-Dawley rats by the collagenase perfusion method as described previously (26). After attachment, cells were infected with adenoviruses for 16 h. Subsequently, cells were maintained in the serum-free Medium 199 media (Mediatech) overnight and treated with 100 nM dexamethasone and 10 M forskolin for 2 h with or without 100 nM insulin for 16 h. Male 8-week-old C57BL/6 mice were provided with a standard rodent diet. T0901317 (LXR agonist, 50 mg/kg body weight) or vehicle (1% methylcellulose and 1% Tween 80) were administered by oral gavage each day for 1 week. Adenoviruses GFP or DAX-1 were delivered by tail vein injection on the 4th day of oral gavage. Three days after adenovirus injection, mice were sacrificed. Protein and RNA were extracted from the livers of each group for Western blot and real time quantitative RT-PCR analyses, respectively. All experiments were conducted by the guideline of Sungkyunkwan University School of Medicine Institutional Animal Care and Use Committee (IACUC).
Confocal Microscopy-Confocal microscopy was performed as described previously (1). Briefly, HeLa cells were grown on uncoated glass coverslips and transfected with pEGFP-DAX-1 and pCDNA3/HA-LXR␣ by the Lipofectamine method (Invitrogen). 24 h after transfection, cells were fixed with 3.7% formaldehyde for 40 min, mounted on glass slides, and observed with a laser-scanning confocal microscope (Olympus Corp., Lake Success, NY). For detection of pCDNA3/HA-LXR␣, cells were permeabilized with 2 ml of PBS containing 0.1% Triton X-100 and 0.1 M glycine at room temperature, incubated for 15 min, washed three times with 1ϫ PBS, and blocked with 3% (w/v) bovine serum albumin in PBS for 10 min at room temperature. Cells were directly incubated with primary anti-HA antibody (1) for 1 h at room temperature, washed three times with 1ϫ PBS, and then mounted on the slide and observed under a microscope.
In Vivo Immunofluorescent Staining-Paraffin sections were used for immunofluorescence staining. After deparaffinizing in xylene and rehydration, the sections were treated with sodium citrate buffer (pH 6.5) and microwaved for 15 min. After cooling for 30 min at room temperature, the sections were incubated with 3% bovine serum albumin for 1 h. The tissue sections were incubated overnight at 4°C with rabbit polyclonal anti-DAX-1 (1:75; H-300, Santa Cruz Biotechnology, Santa Cruz, CA) and mouse polyclonal anti-LXR␣ antibody (1:75; H-171, Santa Cruz Biotechnology). Alexa Fluor 488 nm rabbit monoclonal anti-mouse antibody (1:200; Invitrogen) and 546 nm goat monoclonal anti-rabbit antibody (1:200; Invitrogen) were then applied for 1 h at room temperature in the dark. After each step, slides were washed three times for 5 min in PBS. The coverslips were then mounted by ProLong Gold Antifade rea-gent (Invitrogen) and analyzed under a dual fluorescence microscope (LSM 510 META; Carl Zeiss, Jena, Germany). Control slides were processed the same way, except for omission of the primary antibodies.
Western Blot Analysis-Western blot analysis was performed as described previously (1,26). Briefly, HepG2 cells were transfected with the indicated expression vectors or small interfering RNA oligonucleotides. Forty eight hours after transfection, cell lysates were prepared and separated on 12% SDS-polyacrylamide gel. At about 80% confluency, H4IIE and rat primary hepatocytes were treated with insulin. The cells were then harvested at different time points, and proteins were transferred to a nitrocellulose membrane (Amersham Biosciences). The membranes were probed with an anti-HA, FLAG, DAX-1, or ␤-actin antibodies and developed after secondary antibody incubation using the ECL kit (Amersham Biosciences) according to the manufacturer's instruction.
In Vivo Interaction and Coimmunoprecipitation (CoIP) Assays-HepG2 cells grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum were plated in 6-well flat-bottomed microplates at a concentration of 2 ϫ 10 5 cells per well the day before transfection as described previously (1). Briefly, 1 g of each plasmid was transfected into 293T cells with a calcium phosphate precipitation method. Forty eight hours after transfection, cells were solubilized, and the cleared lysates were mixed with 15 l of glutathione-Sepharose beads and rotated for 2 h at 4°C. The bound proteins were eluted by boiling in SDS sample buffer, subjected to SDS-PAGE, and then transferred to polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA). The membranes were probed with anti-HA or anti-GST antibody and then developed using the ECL kit. For coimmunoprecipitation assays, 750 g of total protein extract from fasted (12 h) or refed (12 h) mouse livers were immunoprecipitated using DAX-1 (H-300, Santa Cruz Biotechnology) antibody and anti-rabbit IgG beads (Trueblot, eBioscience). Then Dax-1 and Lxr␣ proteins were detected by Western blot analysis using DAX-1 or LXR␣ (PPZ0412, Abcam) antibodies and with respective secondary antibodies. Signals were detected with ECL-Plus (Amersham Biosciences). MARCH 19, 2010 • VOLUME 285 • NUMBER 12
Statistical Analyses-Results are shown as means Ϯ S.D. The comparison of different groups was carried out using two-tailed unpaired Student's t test, and differences at or under p Ͻ 0.05 were considered statistically significant and reported as in the figure legends.

DAX-1 Inhibits the Transcriptional Activity of LXR␣-We
have recently reported that DAX-1 is expressed in liver and represses the transcriptional activity of LXR␣ to negatively regulate gluconeogenesis (1). To further support the notion that DAX-1 is significantly expressed in liver, we have examined the expression pattern of DAX-1 using both mouse tissue samples and liver cell lines. Our Western blot analysis using mouse tissue samples indicates that DAX-1 is moderately expressed in liver when compared with testis. In addition, we have performed Western blot analysis using two different doses of protein extracts from different cell lines, and this result indicates that DAX-1 is expressed in hepatoma cells such as HepG2, Hepa1c1c7, and H4IIE (Fig. 1A). To confirm whether DAX-1 has any effect on other important liver-specific factors, we have performed transient transfection assays using LXR␣ and its specific reporter or in combination with DAX-1. We found that DAX-1 dose-dependently decreased the reporter activity containing the LXR-binding site (LXRE-Luc) induced by LXR␣ in the presence of its synthetic agonist T0901317 (Fig. 1B). Transfection of DAX-1 also inhibits the basal and LXR␣-mediated reporter activity. We have also confirmed the similar effect in 293T cells indicating that this repressive effect is not cell typespecific (Fig. 1C). Western blot analysis showed that increase in FLAG-DAX-1 protein did not affect the expression level of HA-LXR␣ indicating that this repressive effect was not due the reduction in LXR␣ protein level (Fig. 1D). To investigate the effect of endogenous DAX-1 on LXR␣ transactivity, we knocked down the endogenous DAX-1 using small interfering RNAs. Knockdown of endogenous DAX-1 further significantly increased LXR␣-mediated transactivity in HepG2 cells (Fig.  1E). We have also confirmed that transfection of siDAX-1-2 significantly decreased the endogenous DAX-1 protein level, whereas control siDAX-1-1 did not decrease the expression of DAX-1 (Fig. 1F). Overall, these results indicate that DAX-1 is expressed in liver, and it inhibits the transcriptional activity of LXR␣.
DAX-1 Interacts and Colocalizes with LXR␣-To determine whether the LXR␣ repression by DAX-1 is mediated through a direct physical interaction, we have performed in vitro MBP pulldown assay. MBP alone, MBP-DAX-1, and MBP-SHP were bacterially expressed and incubated with translated in vitro 35 Slabeled LXR␣ WT or LXR␣⌬AF2 proteins. LXR␣ was found to interact with MBP-DAX-1 but not with MBP alone. Interestingly, the interaction was increased significantly in the presence of LXR␣ agonist when compared with vehicle alone. As expected, interaction of LXR␣ with MBP-hSHP was significantly increased in the presence of ligand (supplemental Fig. 1) (28). To further confirm this interaction in vivo, we performed in vivo GST pulldown assay by transfecting GST (pEBG) or GST-DAX-1 (pEBG-DAX-1) with HA-LXR␣. After GST purification, HA-LXR␣ was detected in the coprecipitates only when coexpressed with GST-DAX-1 but not with the negative control GST alone ( Fig. 2A). To examine the interaction between endogenous DAX-1 and LXR␣, we performed coimmunoprecipitation assay using DAX-1 and LXR␣ antibodies. Our results indicated that interaction between DAX-1 and LXR␣ was modestly increased after the treatment of synthetic LXR agonist (Fig. 2B). Next, we performed coimmunoprecipitation assay using normal mouse liver to show the endogenous interaction between DAX-1 and LXR␣ in mouse liver. We immunoprecipitated the liver extract using DAX-1 antibody and performed Western blotting using LXR␣ antibody. We found that LXR␣ was coimmunoprecipitated with DAX-1 but not with IgG alone indicating that DAX-1 interacts with LXR␣ in vivo (Fig. 2C). To investigate whether DAX-1 and LXR␣ are colocalized in the same subcellular compartment, we performed confocal microscopic analysis in HeLa cells. Expression vectors for GFP-DAX-1 and HA-LXR␣ were transfected alone or together in the presence and absence of ligand. As expected, both DAX-1 and LXR␣ were primarily localized in the nucleus when they were transfected alone (1,17). In addition, these two proteins were colocalized in the nucleus when they were transfected together both in the absence or presence of agonist (Fig.  2D). To further confirm the colocalization of DAX-1 and LXR␣ in liver, we performed immunofluorescent staining using normal mouse liver sections. We have observed the nuclear colocalization of DAX-1 and LXR␣ from merged image indicating that these two proteins are colocalized in the nucleus of mouse hepatocytes (Fig. 2E). Taken together, these results demonstrate that DAX-1 physically interacts with LXR␣ in vivo, and they are colocalized in the nucleus.
Mapping of Interaction Domain between DAX-1 and LXR␣-To map the interaction domain of LXR␣ with DAX-1, we performed MBP pulldown assay. As shown in Fig. 3A (upper  panel), we generated four deletion constructs of LXR␣ fused to the N-terminal domain (LXR␣ AB), DNA binding domain (LXR␣ C), hinge and ligand binding domain (LXR␣ DE), and without activation function-2 domain (LXR␣ DE⌬AF-2). In vitro translated and 35 S-labeled HA-DAX-1 was incubated with bacterially expressed MBP alone or MBP-fused LXR␣ deletion constructs. We observed that LXR␣ DE and DE⌬AF-2 showed strong interaction with HA-DAX-1 when compared with LXR␣ WT both in the absence and presence of ligand, whereas weak or no interaction was observed in the cases of LXR␣ C and AB (Fig. 3A, lower panel). These results indicate that DAX-1 interacts mainly with LBD and the hinge region of LXR␣. In addition, we have also performed similar experiment using in vitro translated SRC-1, and the results indicate that the hinge and LBD region of LXR␣ was involved the interaction with SRC-1 in the presence of ligand. Conversely, when we performed reciprocal mapping experiments using deletion con-structs of DAX-1 (Fig. 3B, upper panel), all 35 S-labeled translated DAX-1 proteins interacted with MBP-LXR␣ (Fig. 3B,  lower panel), indicating that the entire DAX-1 protein is involved in the interaction with LXR␣. It has been reported previously that the first LXXLL motif of DAX-1 is more essential for the interaction as well as nuclear accumulation of DAX-1 (30) than other two LXXLL motifs. Therefore, we made the first LXXLL motif mutant DAX-1 (HA-DAX-1 mL1) (Fig.  3C, upper panel). We performed the interaction assay using

Orphan Nuclear Receptor DAX-1 Inhibits LXR␣ Transactivation
bacterially expressed MBP-LXR␣ with HA-DAX-1 mL1 (first LXXLL motif mutant) and DAX-1 WT. No significant change in interaction was observed between LXR␣ and DAX-1 mL1 compared with DAX-1 WT (Fig. 3C, lower panel). Moreover, all these DAX-1 deletion and mutant constructs inhibited the transcription activity of LXR␣ similar to DAX-1 WT (supplemental Fig. 2) indicating that this interaction is independent of the DAX-1 LXXLL motif. Overall, these mapping results indicate that the entire structure of DAX-1 is involved in the interaction with LBD and hinge region of LXR␣.

DAX-1 Competes for and Represses Binding of SRC-1 to
LXR␣-It has been reported that SRC-1 can coactivate the transcriptional activity of LXR␣ (17). To investigate the functional mechanism of LXR␣ repression by DAX-1, we performed competition experiments using the LXR␣ coactivator SRC-1. HepG2 cells were transfected using LXRE-Luc and LXR␣ expression vector with different combinations of expression vectors for DAX-1 and SRC-1. We have found that the transcriptional activity of LXR␣ was coactivated by the transfection of SRC-1, and it was decreased significantly by DAX-1 in a dose-dependent manner. However, DAX-1 repression was significantly released by SRC-1 in a dose-dependent manner (Fig. 4A). Furthermore, we performed in vitro MBP competition assay to confirm competition between DAX-1 and SRC-1 for binding to LXR␣ in vitro. Increasing amounts of unlabeled full-length SRC-1 competed with 35 S-labeled DAX-1 for the binding with MBP-LXR␣. However, the weak interaction between 35 S-labeled DAX-1 and MBP-LXR␣ CT was not altered by the incubation of unlabeled SRC-1 (Fig. 4B). Altogether, these results indicate that DAX-1 competes with SRC-1 for the binding to LXR␣ to repress LXR␣ transactivation.

DAX-1 Decreases Agonist-induced LXR␣ Target Gene Promoter
Activity and Expression-Next, we examined whether DAX-1 can inhibit the activity of natural LXR␣ target gene promoter. HepG2 cells were transfected with the LXR␣ target gene promoter SREBP-1c-Luc and an expression plasmid for DAX-1 and were treated with the LXR␣ agonist. DAX-1 inhibited T7-mediated SREBP-1c-Luc activity in a dose-dependent manner, indicating that DAX-1 can also inhibit LXR␣-target gene promoter activity (Fig. 5A). To examine whether this inhibition was achieved by direct recruitment of DAX-1 on the SREBP-1c promoter, we performed endogenous chromatin immunoprecipitation assay in rat primary hepatocytes in the absence or presence of ligand with or without the infection of Ad-DAX-1. We could find a minimal recruitment of DAX-1 on the LXR␣ binding region of the SREBP-1c promoter in the absence of ligand, whereas this were determined by anti-HA antibody. The same blot was stripped and re-probed with an anti-GST antibody (middle panel) to confirm the expression levels of the GST fusion protein (GST-hDAX-1) and the GST control (GST). WB, Western blot. B, endogenous interaction between LXR␣ and DAX-1 in HepG2 cells. Protein extracts from HepG2 cells treated with vehicle (DMSO) or T7 were coimmunoprecipitated (IP) using DAX-1 antibody or secondary antibody alone (negative control) and Western blotted with LXR␣ antibody. Inputs (10%) for DAX-1 and LXR␣ are shown in the bottom panels. C, coimmunoprecipitation assays with liver extracts (n ϭ 4) demonstrate the functional association between LXR␣ and DAX-1. Protein extracts from livers were immunoprecipitated using DAX-1 antibody or IgG alone (negative control) and Western-blotted with LXR␣ (upper two panels) antibody. Expression of LXR␣ and DAX-1 (lower two panels) from 10% of lysate were analyzed by Western blotting with specific antibodies. recruitment was totally abolished in the presence of the LXR␣ agonist. As expected, the recruitment of LXR␣, SRC-1, and acetylated lysine 9 of histone H3 was significantly increased after the treatment of LXR␣ agonist. However, the recruitment of these proteins was significantly decreased by the infection of adenovirus DAX-1. These results suggest the existence of competition mechanism between DAX-1 and SRC-1 for repression of the LXR␣ transactivation (Fig. 5B). To confirm whether DAX-1 can also decrease the expression of LXR␣ target genes, we infected adenovirus DAX-1 (Ad-DAX-1) in HepG2 and rat primary hepatocytes. We found that infection of Ad-DAX-1 dose-dependently decreased T7-induced expression of LXR␣ target genes, SREBP-1c and FAS (Fig. 5, C and D). We further examined the effect of DAX-1 on down-regulating the LXR␣ target gene expression by performing adenovirus-mediated knockdown experiments. Infection of adenovirusexpressing short hairpin RNA for DAX-1 (Ad-sh-DAX-1) significantly abolished the endogenous DAX-1 expression compared with a control virus (Ad-US). Knockdown of endogenous DAX-1 significantly increased the expression of T7-mediated SREBP-1c and FAS gene expression (Fig. 5E). Overall, these results indicate that DAX-1 can down-regulate the LXR␣ target gene expression in both the liver cell line and rat primary hepatocytes.

DAX-1 Decreases Triglyceride Level and Lipid Accumulation in
Liver-To assess the functional consequences of the down-regulation of T7-mediated lipogenic genes by DAX-1, we first examined the expression of DAX-1 and LXR␣ target genes in normal and obese-diabetic (db/db) mice. As reported previously (31), Srebp-1c, FAS, and acetyl-coenzyme A carboxylase were significantly higher in db/db mice compared with normal mice, whereas the expression of DAX-1 was significantly low in db/db mice compared with normal mice (Fig.  6A). These results indicate that there was an inverse correlation between DAX-1 and LXR␣ target gene expression in the livers of normal and db/db mice. Next, we treated the normal mice with LXR␣ agonist for 7 days and performed a tail vein injection of Ad-DAX-1. We found that LXR␣ target genes were significantly increased after the treatment of the LXR agonist, and it was drastically decreased by the infection of Ad-DAX-1. We also examined other LXR␣ target genes such as Cyp7a1 and ChREBP. We observed that agonist-induced both Cyp7a1 and ChREBP were significantly decreased by the infection of Ad-DAX-1 (Fig. 6B). We had also observed that T7-mediated induction of nuclear SREBP-1c and acetyl-coenzyme A carboxylase protein levels were decreased (supplemental Fig. 3) in the liver extracts infected with Ad-DAX-1. Consistent with the decrease in lipogenic gene expression, T7-induced liver triglyceride levels were also decreased after DAX-1 expression (Fig.  6C). However, plasma cholesterol and triglyceride levels were not significantly changed by adenovirus tail vein injection (supplemental Fig. 4). Finally, we tried to find out whether DAX-1 can decrease the lipid accumulation in liver. We per-FIGURE 5. DAX-1 decreases LXR␣ agonist-mediated target gene promoter activity and expression. A, HepG2 cells were transfected with SREBP-1c-Luc and with LXR␣ and DAX-1, and cell lysates were utilized for luciferase and ␤-galactosidase assays. The results shown are means of ␤-galactosidase values from three independent experiments. B, chromatin immunoprecipitation analysis using antibodies for LXR␣, DAX-1, SRC-1, and acetyl-H3. PCR amplification of immunoprecipitated chromatin fragments was conducted using primer pairs specific for the proximal, regulatory region (1) and a distal, nonregulatory region (2) of the SREBP-1C gene promoter (left panel). Cell extracts from primary hepatocytes treated with vehicle (DMSO) or LXR␣ agonist (T7) with or without infection of Ad-DAX-1 were immunoprecipitated with LXR␣, DAX-1, SRC-1, and acetyl-H3 antibodies (right panel). After reverse cross-linking, DNA was extracted, and PCR was performed using primers for LXR-RE containing proximal and nonspecific distal region of SREBP-1c promoter. As a negative control, cell extracts were incubated with IgG without any preincubation of primary antibody. Inputs (5%) for both proximal and distal SREBP promoter are shown. C and D, quantitative PCR analysis was performed using total RNA extracted from HepG2 (C), and RT-PCR analysis was performed using total RNA from rat primary hepatocytes (D) after the treatment of LXR␣ agonist with or without adenovirus DAX-1 infection. DAX-1, SREBP-1C, FAS, and ␤-actin genes amplified using specific primers for DAX-1, SREBP-1C, FAS, and ␤-actin were used for PCR. E, Student's t test. Western blot analysis showing protein level of DAX-1 from HepG2 cells infected with mock and Ad-shDAX-1 (right panel).  MARCH 19, 2010 • VOLUME 285 • NUMBER 12

JOURNAL OF BIOLOGICAL CHEMISTRY 9229
formed Oil Red O staining using a mouse liver sample treated with vehicle alone or T7-treated mice with either Ad-GFP or Ad-DAX-1 infection. A significant decrease in T7-mediated lipid accumulation was observed in livers of mice with Ad-DAX-1 infection (Fig. 6D) indicating that DAX-1 can decrease T7-mediated lipid accumulation in liver. Overall, these results suggest that DAX-1 can control the LXR agonistmediated lipogenesis in liver.

DISCUSSION
We have recently demonstrated that DAX-1 is expressed in liver and controls the gluconeogenesis by inhibiting the transcriptional activity of HNF4␣ (1). In this study, our expression analysis in different tissues and cell lines has further supported our previous evidence that DAX-1 is significantly expressed in liver. We have also found that DAX-1 can inhibit the transcriptional activity of another liver-enriched nuclear receptor LXR␣. This has been achieved by direct physical interaction as well as the competition with LXR␣ coactivator SRC-1. Overexpression of DAX-l significantly decreases the LXR agonist-mediated induction of target gene expression, whereas knockdown of endogenous DAX-1 significantly increases the expression of LXR␣ target genes such as SREBP-1c and FAS. Finally, we observed that adenovirus-mediated expression of DAX-1 significantly decreased T7-mediated induction of lipogenic genes and lipid accumulation in liver. Altogether, this study suggests that DAX-1 can control the lipogenesis by inhibiting the transcriptional activity of LXR␣ in liver.
Although DAX-1 has been shown to be primarily expressed in gonads and adrenal glands, there are some reports showing the existence of DAX-1 in the liver (32)(33)(34). Our expression analyses in different tissues samples of mice and in cell lines have also provided more evidence that DAX-1 is expressed in liver. Our previous report has also demonstrated that DAX-1 can control hepatic gluconeogenesis by inhibiting the transcriptional activity of HNF␣ (1). Consistent with our previous report, we have found that DAX-1 can also inhibit the transcriptional activity of another liverspecific receptor LXR␣. Similar to this DAX-1-mediated repressive effect, the closely related family member SHP has also been to shown to inhibit the transcriptional activity of LXR␣ (28). We have found the repressive effect of DAX-1 was not cell type-specific, reminiscent of our previous reports that the DAX-1-mediated transcriptional repression of HNF4␣ and PPAR␥ was not cell typespecific (1, 12).
Next, we have shown the physical interaction between DAX-1 and LXR␣ both in vitro and in vivo. This direct interaction was increased in the presence of the LXR agonist, which is also consistent with the previous study that the interaction between SHP and LXR␣ was stronger in the presence of LXR agonist (28). This increase in interaction might be due to the increased expression and stability of the LXR␣ protein in the presence of agonist (35,36). Consistent with these reports, we have also observed an increase in LXR␣ protein level in the input panel after the treatment of synthetic agonist. Moreover, coactivators and corepressors were known to recognize overlapping surfaces of liganded and unliganded nuclear receptors, respectively. At a sufficiently high concentration, the NCoR has been previously shown to influence the activity of the LXR even in the presence of a potent full agonist that destabilizes NCoR binding (19). In accordance with previous reports (1,17), DAX-1 and LXR␣ were predominantly localized in the nucleus when they were expressed alone both in the absence or presence of agonist. These two proteins were colocalized in the nucleus when they were coexpressed both in the absence or presence of agonist. This result strengthens the previous evidence that RIP140 and LXR␣ were colocalized in the nucleus (37). Our domain mapping results depicted that LBD and the hinge region of LXR␣ is involved in the interaction with DAX-1 and is consistent with previous reports that SHP and RIP140 interacted with the LBD region of LXR␣ (28,38) or DAX-1 interacted with DNA binding domain and hinge region of PPAR␥ (12). On the other side, the entire DAX-1 protein is involved in the interaction with LXR␣, which is also consistent with our previous observation that all the domains of DAX-1 interacted with PPAR␥ and HNF4␣ (1,12). However, the LBD region of DAX-1 was involved in the interaction with Nur77 (11). However, SHP utilized its C-terminal domain for the interaction with LXR␣ (28). It has been previously reported that DAX-1 has three LXXLL motifs in its N-terminal repetitive region and the first LXXLL motif was more essential for the interaction as well as nuclear accumulation of DAX-1 than other two motifs (30). In this study, the first LXXLL motif mutant did not show any significant change in interaction with LXR␣ (Fig. 3C), and it could still inhibit the LXR␣ transactivation like wild type DAX-1 (supplemental Fig. 2). We have found that DAX-1 was competing with LXR␣ coactivator SRC-1 for binding to LXR␣ in the presence of agonist. Similarly, competition between DAX-1 was shown to be competed with SRC-1 for repressing Nur77 (11). However, DAX-1 competed with coactivators such as PGC-1␣ and GRIP-1 for repressing the transcriptional activity of PPAR␥ and glucocorticoid receptor, respectively (10,12). It has previously been reported that DAX-1 can recruit NCoR to suppress the transcriptional activity of SF-1 (9), and NCoR was also known to repress the transcriptional activity of LXR␣ (19). Therefore, in addition to the coactivator competition mechanism, DAX-1 may also recruit NCoR to repress LXR␣ transactivation.
LXRs can directly promote SREBP-1c gene transcription through two LXRE-binding sites in the SREBP-1c promoter (37), and synthetic LXR agonist can up-regulate SREBP-1c gene expression (22). We have found that LXR agonist-mediated SREBP-1c promoter activity was decreased by the transfection of DAX-1. Similarly, it has been previously reported that agonist-induced LXR target gene promoter ABCA1 and SREBP-1c were significantly repressed by NCoR and SMRT (19). It is well known that the expression of FAS and acetyl-coenzyme A carboxylase can be up-regulated by LXR␣ synthetic agonist (22). In addition, we have found that LXR␣, DAX-1, and SRC-1 were recruited on the LXR binding region of the SREBP-1c promoter in the absence of ligand. It is consistent with a previous report that LXR agonist T0901317 significantly increased the binding of LXR␣ on the SREBP-1c promoter (28). In contrast to the recruitment of LXR␣, DAX-1, which was recruited to the SREBP-1c promoter, was significantly dissociated from the promoter after treatment of the LXR␣ agonist. This phenomenon is similar to the previous observation that the recruitment of NCoR was significantly decreased after the treatment of the LXR␣ agonist (22). In addition, protein kinase A phosphorylation of LXR␣ has been shown to impair the DNA binding activity by preventing LXR␣/RXR dimerization and decreases its transcription activity by inhibiting the recruitment of coactivator SRC-1 and enhancing recruitment of corepressor NCoR (22,39). We observed the decrease in recruitment of acetylated histone 3 on chromatin structure by the overexpression of DAX-1 indicating that DAX-1 could affect the chromatin structure like SHP (40). Although interaction between DAX-1 and LXR␣ was increased in the presence of ligand, their recruitments on target gene promoter were inversely correlated.
We have found a significant decrease in LXR␣ agonist-mediated lipogenic gene expression, triglyceride levels, and lipid accumulation in the liver by DAX-1 expression. Similarly, some reports demonstrated that activation of protein kinase A and protein kinase C decreased LXR␣ target gene expressions (39,41), and our previous report suggested that SIK1 can regulate hepatic lipogenesis by controlling SREBP-1c phosphorylation (26). However, the histone deacetylase Sirt1 has been shown to positively regulate LXR␣ target gene expression (18). Our previous report has suggested that salt inducible kinase 1 (SIK1) can induce DAX-1 gene expression in liver (1). Thus, we are currently investigating whether SIK1-mediated induction of DAX-1 in liver can control the expression of lipogenic genes, including Srebp-1c, FAS, and acetyl-coenzyme A carboxylase. LXR is known to regulate the expression of ChREBP, which is also known to directly promote the lipogenic gene transcription (23,42). Therefore, we are currently exploring whether DAX-1 has any direct effect on ChREBP transactivation. Recently, the crystal structure of DAX-1 has also been elucidated (14), and it has been demonstrated that DAX-1 and SHP can form homodimers as well as heterodimers (43). Similar to DAX-1 regulatory effect, SHP was also shown to decrease the lipogenic gene expression and triglycerides in liver (44). In addition, DAX-1 and SHP had similar regulatory effects to control the cAMP-mediated induction of gluconeogenic genes such as phosphoenolpyruvate carboxykinase and Glc-6-Pase. Overall, DAX-1 behaves like SHP in the liver to negatively regulate gluconeogenesis and lipogenesis by repressing HNF4␣ and LXR␣, respectively (1,29). Therefore, target-specific combinatorial expression or double knock-out of these two receptors in liver will be useful to understand the molecular mechanisms to control both glucose and lipid metabolism. A search for a potent inducer of DAX-1 will also be necessary to understand the physiological importance of DAX-1 in liver.
In summary, DAX-1 represses the transcriptional activity of LXR␣ by competing with its coactivator SRC-1, and it has been achieved through the direct physical interaction with LXR␣. In addition, DAX-1 decreases LXR agonist-induced expression of lipogenic genes. Overall, this study suggests that DAX-1 acts as a novel corepressor of LXR␣ and plays a key role in controlling the lipid metabolism in liver.