Ligand-dependent corepressor (LCoR) represses the transcription factor C/EBPβ during early adipocyte differentiation

Nuclear receptors (NRs) regulate gene transcription by recruiting coregulators, involved in chromatin remodeling and assembly of the basal transcription machinery. The NR-associated protein ligand-dependent corepressor (LCoR) has previously been shown to suppress hepatic lipogenesis by decreasing the binding of steroid receptor coactivators to thyroid hormone receptor. However, the role of LCoR in adipogenesis has not been established. Here, we show that LCoR expression is reduced in the early stage of adipogenesis in vitro. LCoR overexpression inhibited 3T3-L1 adipocyte differentiation, whereas LCoR knockdown promoted it. Using an unbiased affinity purification approach, we identified CCAAT/enhancer-binding protein β (C/EBPβ), a key transcriptional regulator in early adipogenesis, and corepressor C-terminal binding proteins as potential components of an LCoR-containing complex in 3T3-L1 adipocytes. We found that LCoR directly interacts with C/EBPβ through its C-terminal helix-turn-helix domain, required for LCoR's inhibitory effects on adipogenesis. LCoR overexpression also inhibited C/EBPβ transcriptional activity, leading to inhibition of mitotic clonal expansion and transcriptional repression of C/EBPα and peroxisome proliferator-activated receptor γ2 (PPARγ2). However, LCoR overexpression did not affect the recruitment of C/EBPβ to the promoters of C/EBPα and PPARγ2 in 3T3-L1 adipocytes. Of note, restoration of PPARγ2 or C/EBPα expression attenuated the inhibitory effect of LCoR on adipogenesis. Mechanistically, LCoR suppressed C/EBPβ-mediated transcription by recruiting C-terminal binding proteins to the C/EBPα and PPARγ2 promoters and by modulating histone modifications. Taken together, our results indicate that LCoR negatively regulates early adipogenesis by repressing C/EBPβ transcriptional activity and add LCoR to the growing list of transcriptional corepressors of adipogenesis.

been shown to interact with transcriptional factor Krüppel-like factor 6 (KLF6) and repress KLF6 target gene transcription (8). These findings suggest that different mechanisms are involved in the regulation of transcription by LCoR.
NR coregulators have been implicated in not only developmental but also metabolic processes and diseases (9). It is now well-accepted that NR coregulators are important metabolic switches and regulate metabolic pathways through their interactions with not only NRs but also other transcription factors. The potential pathological role of LCoR has been studied extensively in breast and prostate cancers using various cell lines (5,8,10). However, the physiological and pathological functions of LCoR in metabolic regulation are largely unknown. In a previous study, we demonstrated that LCoR is able to repress TRmediated expression of lipogenic genes and acts as a negative regulator of hepatic lipogenesis. Down-regulation of hepatic LCoR might contribute to the dysregulation of lipogenesis and the development of fatty livers in obese mouse models and patients with non-alcoholic fatty liver disease. More importantly, hepatic LCoR could serve as a potential target for treating hepatic steatosis (7). Besides liver, adipose tissue is also a key metabolic organ and plays a vital role in the regulation of energy homeostasis. We showed that LCoR is highly expressed in white adipose tissue; however, its role in adipose tissue has never been studied (7).
Understanding the molecular mechanism underlying adipogenesis will provide insights into adipose tissue plasticity and remodeling in adaptive energy homeostasis as well as in metabolic disease. It has been proposed that the development of adipose depots can be divided into two stages, commitment and terminal differentiation, which gives rise to adipocytes from preadipocytes. As a result of intensive studies using cellular models including 3T3-L1 cells and mouse embryonic fibroblasts, it is known that adipogenesis is orchestrated by a complex network of transcriptional cascades (11)(12)(13). As the master transcriptional regulators, peroxisome proliferator-activated receptor ␥ (PPAR␥) and CCAAT/enhancer-binding protein ␣ (C/EBP␣) control almost all of the genes involved in the morphological changes and lipid accumulation during adipocyte differentiation and lipid droplet formation. C/EBP␤ and C/EBP␦, two transcription factors induced in the early stage of adipogenesis, play crucial roles in the initiation of the differentiation program by activating the expression of PPAR␥ and C/EBP␣, which ultimately leads to terminal adipogenic differentiation.
In this study, we showed that LCoR is a negative regulator of adipogenesis. Moreover, our data suggested that LCoR can interact with C/EBP␤ and repress the transcriptional activity of C/EBP␤ by recruiting CtBP in the early stage of adipogenesis. Our results provided novel insights into the molecular mechanism involved in the regulation of early adipogenesis.

The expression of LCoR is dynamically regulated during adipogenesis
LCoR is expressed in a wide variety of fetal and adult human tissues (4), but its expression profile during adipogenesis is unclear. To explore the potential roles of LCoR in adipogenesis, we initiated our study by examining the expression levels of LCoR during 3T3-L1 adipocyte differentiation. The cells were induced to differentiate into mature adipocytes using a standard induction hormone mixture. We observed that the mRNA levels of LCoR were decreased gradually within the first 8 h after induction of differentiation, then increased gradually, and reached a peak at the end of the differentiation (Fig. 1A). Similarly, the protein levels of LCoR declined from the initial stage of differentiation, were maintained at a very low level until day 1, and then increased considerably from day 2 to the end of the differentiation (Fig. 1B). The mRNA expression of C/EBP␤, C/EBP␣, PPAR␥, and adipocyte fatty-acid binding protein 4 (ap2), which are the key players and/or adipogenic markers during adipogenesis, were also examined (supplemental Fig. S1, A-D). We also assessed the expression of LCoR in primary adipocytes using stromal vascular fraction (SVF) cells as described under "Experimental procedures." Similar to the LCoR expression pattern that was observed during 3T3-L1 adipocyte differentiation, the mRNA levels of LCoR were reduced in the early stage and increased gradually afterward during the adipogenesis of SVF cells (Fig. 1C). The dynamic expression of LCoR indicated that LCoR might have a potential role in adipogenesis.

Overexpression of LCoR inhibits 3T3-L1 adipocyte differentiation
To investigate the effect of LCoR on adipocyte differentiation, we stably overexpressed FLAG-tagged LCoR (F-LCoR) in 3T3-L1 preadipocytes using retroviral vectors. Overexpression of LCoR was first confirmed at both mRNA and protein levels ( Fig. 1, D and E). Analysis using the Agilent mouse gene expression microarrays revealed that the overexpression of LCoR down-regulated the expression of those genes implicated in multiple metabolic pathways in differentiated 3T3-L1 adipocytes, including the PPAR signaling pathway, fatty acid metabolism, and fatty acid elongation (supplemental Table S1). Because the PPAR signaling pathway, fatty acid metabolism, and fatty acid elongation are essential for mature adipocyte formation, we hypothesized that LCoR is able to regulate adipogenesis.
To test our hypothesis, the adipogenic abilities of 3T3-L1 preadipocytes overexpressing F-LCoR were accessed by Oil Red O staining after differentiation induction. Compared with control cells, the cells overexpressing F-LCoR showed a severely impaired ability to accumulate lipids (Fig. 1F). We then examined several adipogenic markers to confirm the inhibitory effect of F-LCoR on adipocyte differentiation. The mRNA and protein levels of early adipogenic marker C/EBP␤ were not altered, whereas the mRNA and protein expression levels of late adipogenic markers C/EBP␣, PPAR␥, and their downstream target gene, ap2, were significantly reduced in cells overexpressing F-LCoR (Fig. 1, G-K). These results indicated that overexpression of LCoR is able to suppress adipogenesis.

Knockdown of LCoR facilitates 3T3-L1 adipocyte differentiation
To verify the inhibitory effect of LCoR on adipocyte differentiation, we established a stable LCoR knockdown 3T3-L1 cell line by using retrovirus-mediated short hairpin RNA targeting LCoR (shLCoR). Significant reduction of endogenous LCoR The role of LCoR during early adipogenesis expression by shLCoR was confirmed in 3T3-L1 preadipocytes at both mRNA and protein levels (Fig. 2, A and B). Then these LCoR knockdown 3T3-L1 preadipocytes were subjected to a differentiation assay. Compared with control cells, LCoR knockdown cells accumulated greater amounts of lipids (Fig.  2C). To confirm the effect of LCoR knockdown on adipogenesis, the expression of adipogenic markers was examined (Fig. 2, D-H). Consistent with our findings in 3T3-L1 cells overexpressing LCoR, we found that the knockdown of LCoR did not lead to changes in C/EBP␤ mRNA and protein expression (Fig.   2, D and H). In contrast, the mRNA and protein levels of adipogenic factors C/EBP␣, PPAR␥, and ap2 were relatively higher in LCoR knockdown cells than those in control cells (Fig. 2, E-H).
These results further suggested that inhibition of LCoR expression is able to promote adipogenesis.

LCoR inhibits adipogenesis through directly interacting with C/EBP␤
To reveal the mechanisms underlying the antiadipogenic effect of LCoR, we used an unbiased affinity purification or SVF (C) cells were harvested at the indicated times after adipogenic induction. The mRNA and protein levels of LCoR were measured by qPCR (A and C) or Western blotting (B), respectively. D and E, the overexpression of LCoR in 3T3-L1 preadipocytes was confirmed by qPCR (D) and Western blotting (E). 3T3-L1 preadipocytes were infected with a retroviral vector containing F-LCoR or an empty vector (pMSCV) as a control. F, Oil Red O staining of 3T3-L1 cells stably overexpressing F-LCoR or pMSCV was performed on day (d) 8 of differentiation. G-K, the expression levels of adipogenic genes in 3T3-L1 cells overexpressing LCoR during adipogenesis. 3T3-L1 cells stably overexpressing F-LCoR or pMSCV were harvested at the indicated times after adipogenic induction. The mRNA and protein levels of C/EBP␤, C/EBP␣, PPAR␥, and ap2 were examined by qPCR (G-J) or Western blotting (K), respectively. Immunoblots and Oil Red O staining shown here are representative of at least three independent experiments, which yielded similar results. Data are represented as mean Ϯ S.D. (n ϭ 3). Error bars represent S.D. ***, p Ͻ 0.001.

The role of LCoR during early adipogenesis
approach to identify the binding proteins of LCoR in 3T3-L1 cells overexpressing F-LCoR as described under "Experimental procedures" (Fig. 3A). The mass spectrometric analysis revealed that C/EBP␤, a key regulator of early adipogenesis, was a potential component of LCoR-containing complex ( Fig. 3A and supplemental Table S2). Given that C/EBP␤ controls the transcription of C/EBP␣ and PPAR␥ during adipogenesis and LCoR regulates the expression of C/EBP␣ and PPAR␥ without affecting C/EBP␤, we speculated that LCoR may inhibit adipogenesis through suppressing the transcriptional activity of C/EBP␤ via direct binding. To test this hypothesis, we first confirmed the interaction of LCoR with C/EBP␤ by coimmunoprecipitation (co-IP) assay in HEK 293T cells transfected with F-LCoR and C/EBP␤ constructs (Fig. 3B). Importantly, the interaction of endogenous LCoR with endogenous C/EBP␤ was also observed in 3T3-L1 cells by co-IP assay (Fig. 3C). To identify which region of LCoR is responsible for the interaction with C/EBP␤, we constructed FLAG-tagged N-or C-terminally truncated mutants of LCoR, F-LCoR⌬N (LCoR lacking the N-terminal NR box), and F-LCoR⌬C (LCoR lacking the C-terminal HTH domain) as indicated (Fig. 3D). We found that C/EBP␤ interacted with F-LCoR⌬N but not F-LCoR⌬C, indicating that the HTH domain of LCoR is required for the interaction of LCoR with C/EBP␤ (Fig. 3E). A similar result was obtained from a GST pulldown assay (Fig. 3F). We found that the GST-LCoR fusion protein bound C/EBP␤ in vitro, suggesting that the interaction between C/EBP␤ and LCoR is direct (Fig. 3F). Moreover, GST-LCoR fusion protein lacking HTH domain lost the ability to bind to C/EBP␤, further supporting the notion that the interaction between C/EBP␤ and LCoR requires the C-terminal HTH domain of LCoR (Fig. 3, D and F).
To determine whether the interaction between C/EBP␤ and LCoR is required for the antiadipogenic effect of LCoR, 3T3-L1 preadipocytes stably overexpressing F-LCoR⌬N or F-LCoR⌬C were used for a differentiation assay. We found that overexpression of F-LCoR⌬N inhibited adipogenesis and lipid accumulation to the same extent as overexpression of F-LCoR did. In contrast, overexpression of F-LCoR⌬C, which lost the ability to bind to C/EBP␤, did not affect 3T3-L1 adipocyte differentiation and lipid accumulation (Fig. 3G). Together, these results indicated that LCoR plays an inhibitory role in adipogenesis by physically interacting with C/EBP␤, and the HTH domain of LCoR is required for the interaction with C/EBP␤ and the antiadipogenic effect of LCoR.

LCoR inhibits the transcriptional activity of C/EBP␤
Because either overexpression or knockdown of LCoR altered the expression of C/EBP␣ and PPAR␥ (Figs. 1, H, I, and K, and 2, D and E), two downstream target genes of C/EBP␤, we speculated that LCoR may inhibit the transcriptional activity of C/EBP␤ through direct binding. To test this hypothesis, we performed luciferase assays using a reporter containing PPAR␥2 promoter region with C/EBP␤-responsive element in HEK 293T cells. We found that overexpression of LCoR repressed the promoter activity of PPAR␥2 in a dose-dependent manner, whereas knockdown of LCoR by specific siRNA enhanced the promoter activity of PPAR␥2 (Fig. 4, A and B). In addition, we found that overexpression of F-LCoR⌬N but not F-LCoR⌬C inhibited the transcriptional activity of C/EBP␤ (Fig. 4C). We also examined the effects of LCoR overexpression and knockdown as well as the effects of its mutants on the promoter activity of PPAR␥2 in 3T3-L1 cells when C/EBP␤ levels were greatly elevated after differentiation induction (Fig.  4, D-F). In agreement with the findings obtained in HEK 293T cells, we found that overexpression of LCoR repressed the promoter activity of PPAR␥2 in 3T3-L1 cells after adipogenic induction, whereas knockdown of LCoR by shLCoR enhanced the promoter activity of PPAR␥2 in 3T3-L1 cells after adipogenic induction (Fig. 4, D and E). Accordingly, we found that LCoR⌬N but not LCoR⌬C had an inhibitory effect on the promoter activity of PPAR␥2 in 3T3-L1 cells after adipogenic induction (Fig. 4F). Notably, the suppressive potencies of these

The role of LCoR during early adipogenesis
LCoR mutants were correlated with their abilities to bind to C/EBP␤ and inhibit adipogenesis (Fig. 3, E-G). These results suggested that the inhibitory effect of LCoR on the transcriptional activity of C/EBP␤ requires direct binding.
The transcriptional activity of C/EBP␤ is important not only for the expression of C/EBP␣ and PPAR␥ but also for the mitotic clonal expansion during 3T3-L1 adipocyte differentiation (14,15). To evaluate the effect of LCoR on mitotic clonal expansion, 3T3-L1 cells stably overexpressing F-LCoR were induced to differentiate for 24 h followed by a BrdU incorporation assay. A significant decrease in the number of BrdU-positive cells was observed in cells overexpressing F-LCoR (Fig. 4, G and H). To further investigate the role of LCoR during mitotic clonal expansion, we measured the expression of cyclin D2 (CCND2) and cyclin E1 (CCNE1), two cyclin genes that are important for cell cycle progression. We found that overexpression of LCoR led to a significant decrease in CCND2 and CCNE1 mRNA levels at both 24 and 48 h after adipogenic induction (Fig. 4, I and J). In addition, the mRNA levels of GINS complex subunit 1 (Gins1) and minichromosome maintenance complex component 3 (Mcm3), two target genes of C/EBP␤ that are required for mitotic clonal expansion (16), were also significantly down-regulated (Fig. 4, K and L). Similar results were obtained when C/EBP␤ expression was knocked down . LCoR inhibits adipogenesis through directly interacting with C/EBP␤. A, 3T3-L1 preadipocytes stably overexpressing F-LCoR or an empty vector (pMSCV) were induced to differentiate for 2 days. LCoR-associated proteins purified from nuclear extracts of indicated cells were separated by SDS-PAGE and visualized by silver staining. The positions of identified proteins are indicated on the right, and molecular mass markers (kDa) are shown on the left. B, HEK 293T cells were transiently transfected with F-LCoR and C/EBP␤ constructs. After 36 h, the cell lysates were immunoprecipitated with antibodies against IgG or FLAG and immunoblotted (IB) with antibodies against C/EBP␤ or LCoR as indicated. C, the lysates of 3T3-L1 cells after differentiation for 48 h were immunoprecipitated with antibodies against IgG or C/EBP␤ and then immunoblotted with antibodies against C/EBP␤ or LCoR as indicated. D, schematic representation of full-length and truncated LCoR proteins with main functional motifs and boundaries indicated. E, HEK 293T cells were transiently transfected with the indicated constructs. After 36 h, the cell lysates were immunoprecipitated with antibodies against FLAG and immunoblotted with antibodies against C/EBP␤ or LCoR. F, GST pulldown analysis was performed using in vitro translated C/EBP␤ with GST control or with full-length or truncated LCoR GST fusion proteins as indicated. G, Oil Red O staining of 3T3-L1 cells stably overexpressing F-LCoR or truncated LCoR was performed on day 8 of differentiation. Blots and Oil Red O staining shown here are representative of at least three independent experiments, which yielded similar results.

The role of LCoR during early adipogenesis
using retrovirus-mediated short hairpin RNA targeting C/EBP␤ (shC/EBP␤) in 3T3-L1 cells after adipogenic induction (Fig. 4, M-Q). Together, these data suggested that overexpression of LCoR is able to inhibit mitotic clonal expansion by suppressing the transcriptional activity of C/EBP␤, which also contributes to the repression of adipogenesis.
Because C/EBP␤ is a transcriptional factor and stimulates the expression of C/EBP␣ and PPAR␥ by directly binding to their promoters, we examined whether LCoR inhibits the transcriptional activity of C/EBP␤ by affecting its DNA binding ability. 3T3-L1 cells stably overexpressing LCoR or its mutants were induced to differentiate for 24 h. Afterward, a chromatin The role of LCoR during early adipogenesis immunoprecipitation (ChIP) assay was performed to test whether overexpression of LCoR or its mutants could affect the recruitment of C/EBP␤ to the promoters of its target genes. In agreement with the current notion that C/EBP␤ is recruited to the promoters of C/EBP␣ and PPAR␥2 after adipogenic induction, we found that C/EBP␤ was enriched in the promoter region of C/EBP␣ and PPAR␥2 at 24 h after adipogenic induction (Fig. 5, A and B). In addition, we found that overexpression of LCoR as well as of its mutants did not affect the recruitment of C/EBP␤ to the promoters of C/EBP␣ and PPAR␥2 in 3T3-L1 cells either before or after adipogenic induction (Fig. 5, A and  B). Because LCoR is able to interact with C/EBP␤, as expected, we observed an increased recruitment of LCoR to the promoters of C/EBP␣ and PPAR␥2 after the induction of adipogenesis (Fig. 5, C and D). Consistent with our finding that F-LCoR⌬N but not F-LCoR⌬C is able to interact with C/EBP␤, we only found the enrichment of LCoR and F-LCoR⌬N but not F-LCoR⌬C on the promoters of C/EBP␣ and PPAR␥2 after adipogenic induction (Fig. 5, C and D). These results suggested that LCoR is recruited to the promoters of C/EBP␣ and PPAR␥2 together with C/EBP␤ after adipogenic induction through direct binding.
It is known that once C/EBP␤ acquires DNA binding ability it is localized to centromeres and forms a punctate pattern in the nuclei (17). To further confirm that the DNA binding ability of C/EBP␤ is not affected by LCoR overexpression, we examined the centromeric localization of C/EBP␤ during adipogenesis. The localization of C/EBP␤ and LCoR was analyzed by immunofluorescence in 3T3-L1 cells stably overexpressing LCoR or it mutants after adipogenic induction for 24 h. We found that C/EBP␤ formed a punctate pattern in all groups, indicating that the overexpression of LCoR or its mutants is not able to affect the centromeric localization of C/EBP␤ (Fig. 5E). As expected, we found that either F-LCoR or F-LCoR⌬N formed a punctate pattern and was colocalized with C/EBP␤ (Fig. 5E). In contrast, because F-LCoR⌬C could not bind to C/EBP␤, F-LCoR⌬C was distributed diffusely within whole nuclei and was not colocalized with C/EBP␤ (Fig. 5E). These results further suggested that the interaction of LCoR with C/EBP␤ during early adipogenesis is required for the inhibitory effect of LCoR.

Overexpression of PPAR␥2 or C/EBP␣ attenuates the inhibitory effect of LCoR on adipogenesis
Given that LCoR is able to repress the transcriptional activity of C/EBP␤, we examined whether reintroducing a target gene of C/EBP␤ (C/EBP␣ or PPAR␥) could attenuate the inhibitory effect of LCoR on adipogenesis. The expression of PPAR␥2 or C/EBP␣in3T3-L1preadipocytesoverexpressingLCoRwasmeasured at both mRNA and protein levels (Fig. 6, A-D). Cells were

The role of LCoR during early adipogenesis
induced to undergo adipogenesis in the presence of the synthetic PPAR␥ ligand rosiglitazone (Rosi) or vehicle (DMSO) alone. The adipogenic abilities were accessed by Oil Red O staining (Fig. 6E). In the absence of Rosi, C/EBP␣ induced about 5-10% and PPAR␥2 induced about 40 -50% of cells overexpressing LCoR to resume adipogenesis (Fig. 6E). Treatment with Rosi further enhanced the C/EBP␣-induced differentiation, whereas PPAR␥2 totally restored the adipogenic abilities of the cells overexpressing LCoR in the presence of Rosi (Fig.  6E). The effect of PPAR␥2 or C/EBP␣ overexpression on the expression of mature adipocytes makers, including ap2, perilipin 2 (plin2), and adiponectin, in the cells overexpressing LCoR was also analyzed (Fig. 6, F-H), and similar results were obtained. Of note, overexpression of PPAR␥2 totally restored the expression of ap2, plin2, and adiponectin in those cells overexpressing LCoR in the presence of Rosi. These results suggested that LCoR might suppress adipogenesis via C/EBP␤mediated transcription of C/EBP␣ and PPAR␥2.

LCoR inhibits C/EBP␤-mediated transcription and adipogenesis by recruiting CtBP
It has been shown that CtBPs play a regulatory role in adipogenesis (18 -20). We noticed that both CtBP1 and CtBP2 were also identified as potential components of LCoR-containing complex in differentiated 3T3-L1 adipocytes (Fig. 3A). Based on these findings, we hypothesized that CtBPs may mediate the suppressive effect of LCoR on the transcriptional activity of C/EBP␤ and early adipogenesis. Given the fact that CtBP1 and CtBP2 are highly related and functionally redundant (20 -22), we made expression constructs carrying CtBP1 gene for further studies. The interaction between CtBP1 and LCoR was first analyzed by using a GST pulldown assay. We found that LCoR interacted with CtBP1 directly (Fig. 7A). However, we could not detect the binding of CtBP1 to C/EBP␤ in a GST pulldown assay, suggesting that CtBP1 does not interact with C/EBP␤ directly (Fig. 7A). We also examined the localization of CtBP1 in 3T3-L1 cells at 24 h after differentiation induction. Immu-

The role of LCoR during early adipogenesis
nofluorescence analysis revealed that CtBP1 was distributed diffusely in the nuclei, whereas C/EBP␤ formed a punctate pattern after differentiation induction (Fig. 7B). The different localization of CtBP1 and C/EBP␤ was much more obvious in 3T3-L1 cells undergoing mitosis (Fig. 7C). These results indicated that CtBP1 normally does not colocalize with C/EBP␤ in Figure 7. LCoR inhibits adipogenesis through recruiting CtBP. A, GST pulldown analysis was performed using in vitro translated LCoR or C/EBP␤ with GST or GST-CtBP1 fusion proteins as indicated. B, confocal microscope images of endogenous C/EBP␤ (green) and CtBP1 (red) in 3T3-L1 cells stably expressing F-LCoR or an empty vector (pMSCV). Cells were induced to differentiate for 24 h before they were fixed. C, confocal microscope images of C/EBP␤ (green) and CtBP1 (red) in 3T3-L1 cells. Cells were induced to differentiate for 24 h before they were fixed. The different localization of CtBP1 and C/EBP␤ was much more obvious in 3T3-L1 cells undergoing mitosis. D, schematic representation of mutant LCoRs (LCoR D1 and LCoR D2) with CtBP-binding sites deleted. E, HEK 293T cells were transiently transfected with the indicated constructs. After 36 h, the cell lysates were immunoprecipitated with antibodies against C/EBP␤ and immunoblotted with antibodies against CtBP1, C/EBP␤, or FLAG. F, the amount of CtBP1 that was coimmunoprecipitated with C/EBP␤ was normalized to the amount of CtBP1 input (top panel) or further normalized by the amount of coimmunoprecipitated LCoR or LCoR mutants (bottom panel). G, HEK 293T cells were transiently transfected with a luciferase reporter plasmid containing PPAR␥2 promoter (PPAR␥2-luc) with or without a C/EBP␤ construct in the absence or presence of LCoR, LCoR D1, or LCoR D2 as indicated. The luciferase activity was measured and is shown as relative luciferase units (RLU). H, Oil Red O staining of 3T3-L1 cells stably overexpressing F-LCoR, F-LCoR D1, or F-LCoR D2 was performed on day 8 of differentiation. I, the efficiency of the knockdown of CtBP1 and CtBP2 by shCtBP1 and shCtBP2, respectively, in 3T3-L1 preadipocytes was determined by Western blotting. J, Oil Red O staining of 3T3-L1 cells infected with F-LCoR or both F-LCoR and shCtBPs (shCtBP1 ϩ shCtBP2) was performed on day 8 of differentiation. K-M, the mRNA levels of C/EBP␣, PPAR␥, and ap2 were examined by qPCR analysis in these differentiated 3T3-L1 cells infected with F-LCoR or both F-LCoR and shCtBPs. Immunoblots, confocal microscope images, and Oil Red O staining shown here are representative of at least three independent experiments. Luciferase assay and qPCR data are represented as mean Ϯ S.D. (n ϭ 3-4). Error bars represent S.D. **, p Ͻ 0.01; ***, p Ͻ 0.001; ns, not significant. aa, amino acids.

The role of LCoR during early adipogenesis
3T3-L1 cells after adipogenic induction, especially when the cells are dividing. In contrast, we observed that CtBP1 colocalized with C/EBP␤ and formed a punctate pattern in 3T3-L1 cells overexpressing LCoR after induction of adipogenesis, suggesting that LCoR serves as a bridge to tether CtBP1 to C/EBP␤, and overexpression of LCoR could lead to the colocalization of CtBPs with C/EBP␤ (Fig. 7B). These results also indicated that the down-regulation of LCoR during early adipogenesis could keep most of the CtBPs away from C/EBP␤, thereby facilitating mitotic clonal expansion and promoting adipogenesis.
It has been shown that LCoR can interact with CtBPs through two CtBP-binding sites, PLDLPVR (site 1) and VLDL-STK (site 2), within its N-terminal region (4). To determine whether the recruitment of CtBPs by LCoR is required for the inhibitory effect of LCoR on the transcriptional activity of C/EBP␤ and adipogenesis, we constructed FLAG-tagged LCoR mutants, F-LCoR D1 (lacking CtBP-binding site 1) and F-LCoR D2 (lacking two CtBP-binding sites) (Fig. 7D). Then we examined whether these LCoR mutants could affect the recruitment of CtBP1 to LCoR-C/EBP␤ complex by performing a co-IP assay in HEK 293T cells. We found that both F-LCoR D1 and F-LCoR D2 were coprecipitated with C/EBP␤, suggesting that these two mutations did not affect the binding of LCoR to C/EBP␤ (Fig. 7E). As expected, we observed that, in the presence of wild-type LCoR, CtBP1 could be coprecipitated with C/EBP␤ (Fig. 7E). However, in the presence of LCoR mutants that lost one or both CtBP-binding sites, the coprecipitation of CtBP1 with the LCoR-C/EBP␤ complex was inhibited accordingly (Fig. 7, E and F). Consistently, we found that LCoR D1 had less inhibitory effect on PPAR␥2 promoter activity compared with wild-type LCoR, whereas LCoR D2 has almost no inhibitory effect as revealed by luciferase assay (Fig. 7G). Moreover, Oil Red O staining results showed that F-LCoR D1 had less antiadipogenic ability, whereas F-LCoR D2 had nearly lost the antiadipogenic effect on 3T3-L1 adipocyte differentiation (Fig.  7H). We also tested whether knockdown of CtBPs could attenuate the effect of LCoR on adipogenesis. In agreement with our previous findings, we found that knockdown of CtBPs using retrovirus-mediated short hairpin RNA targeting CtBP1 and CtBP2 (shCtBPs) restored the adipogenic capacity of LCoRoverexpressing 3T3-L1 cells (Fig. 7, I and J). Consistently, the mRNA expression of C/EBP␣, PPAR␥, and ap2 was derepressed in LCoR-overexpressing 3T3-L1 cells after knockdown of CtBPs by shCtBPs (Fig. 7, K-M). Together, these results indicated that CtBPs mediate the inhibitory effect of LCoR on C/EBP␤-mediated transcription and adipogenesis.

LCoR alters histone modifications on adipogenic promoters
Histone modification plays an important role in regulating gene transcription. Generally, histone H3 acetylation, for example acetylation of histone 3 lysines 9 and 14 (H3K9acK14ac), is correlated with transcriptional activation, whereas dimethylation of histone 3 lysine 9 (H3K9me2) is correlated with repression (23,24). It has been reported that CtBPs can form a functional complex with histone deacetylases 1/2 and histone methyltransferase G9a and result in transcriptional repression by converting active Lys-9-acetylated histone H3 to repressive Lys-9-methylated H3 (25). Based on our findings, we hypothesized that LCoR may alter the histone modifications on the promoters of C/EBP␣ and PPAR␥ through the recruitment of CtBPs, thereby repressing transcription. To test our hypothesis, we first examined whether CtBP1 could be recruited to the promoters of C/EBP␣ and PPAR␥2 by performing a ChIP assay. As expected, we found that overexpression of LCoR significantly increased the occupancy of CtBP1 on the promoter of PPAR␥2 (Fig. 8A). Overexpression of LCoR also significantly reduced the levels of H3K9acK14ac and increased the levels of H3K9me2 on the promoter of PPAR␥2 (Fig. 8, B and C). We also tested whether knockdown of CtBPs could attenuate the effect of LCoR on the histone modification on the PPAR␥2 promoter. As expected, we found that knockdown of CtBPs by shCtBPs not only reduced the occupancy of CtBP1 on the promoter of PPAR␥2 in LCoR-overexpressing 3T3-L1 cells but also abolished the effect of LCoR overexpression on the enrichment of H3K9acK14ac and H3K9me2 on the promoter of PPAR␥2 in 3T3-L1 cells (Fig. 8, D-F). These results suggested that CtBPs might mediate the suppressive effect of LCoR on C/EBP␤-controlled transcription through modulating histone modifications on adipogenic promoters.
Taken together, our study here demonstrated that LCoR is a negative regulator of adipogenesis. LCoR is able to interact with C/EBP␤ and inhibit its activity by recruiting CtBP corepressor complex during early adipogenesis, which might result in chromatin remodeling and transcription repression (Fig. 8G).

Discussion
Uncovering the molecular basis of adipogenesis is crucial for a better understanding of adipose tissue plasticity and remodeling in metabolic homeostasis and disease (26). It is known that the regulation of adipogenesis is orchestrated by regulatory circuits controlled to a great extent by transcriptional mechanisms (11,12). A growing body of evidence suggests that transcriptional coregulators play pivotal roles in transcriptional regulation during adipogenesis (11,27,28). For example, NCoR, SMRT, and SRC-3 were found to be key regulators of adipocyte differentiation (29 -32). However, the role and importance of LCoR in adipogenesis have not been reported before. Here, we demonstrated that LCOR is a negative regulator of adipogenesis. By repressing the transcriptional activity of C/EBP␤, LCoR not only suppresses the C/EBP␤-mediated expression of C/EBP␣ and PPAR␥ but also inhibits mitotic clonal expansion. Taken together, our study reveals a physiological role of LCoR in the control of mitotic clonal expansion and adipocyte differentiation, adding LCoR to the growing list of key transcriptional coregulators of energy homeostasis.
Growing evidence has demonstrated that nuclear receptor coregulators are able to modulate transcription controlled by other types of transcriptional factors. For example, SMRT is able to control adipogenesis through C/EBP␤ and KAISO (31), whereas SRC-3 is able to regulate PPAR␥ expression by targeting C/EBP␣ and C/EBP␦ (33). In our study, we showed that LCoR is able to interact with both C/EBP␤ and CtBP1, whereas C/EBP␤ and CtBP1 did not interact with each other directly. Given that CtBP was colocalized with C/EBP␤ and enriched on the promoter regions of C/EBP␤ target genes upon LCoR overexpression, we speculated that LCoR serves as a bridge to tether

The role of LCoR during early adipogenesis
CtBP1 to C/EBP␤, which results in transcription repression and adipogenesis inhibition. It has been reported that CtBP can regulate adipogenesis through multiple transcription factors, including KLF3 and GATA (18,19). Therefore, our result revealed a new regulatory mechanism for CtBP in the regulation of adipogenesis.
LCoR, originally identified as a corepressor of estrogen receptor ␣, was shown to interact with estrogen receptor ␣ and some other nuclear receptors only in the presence of ligands (4). Previously, we demonstrated that LCoR interacts with TR and represses the transcriptional activity of TR in a hormone-independent manner. Furthermore, we found that LCoR acts as a competitor for coactivators SRC-1/3 in binding to TR in liver (7). Because both TR and SRCs are involved in the regulation of adipogenesis (28,32,34), whether LCoR also could modulate adipogenesis through TR and/or SRC transcription requires further study. In contrast, our study suggested that the downregulation of LCoR expression in the early stage might facilitate mitotic clonal expansion and promote C/EBP␤-mediated C/EBP␣ and PPAR␥ expression. However, it is still not clear why the expression of LCoR increases gradually at the end of adipogenesis. SMRT has been shown to control adipogenesis through either C/EBP␤ or PPAR␥, suggesting that this corepressor may play distinct roles at different stages. Therefore, further studies using an inducible system or adenoviral system to study the effect of LCoR in the late stage are required to address these questions.
In summary, our study indicates that LCoR acts as a negative regulator of early adipogenesis. LCoR suppresses adipogenesis by targeting C/EBP␤-mediated gene expression and mitotic clonal expansion. Further mechanistic studies revealed that LCoR is able to interact with C/EBP␤ and recruit another core-

The role of LCoR during early adipogenesis
pressor, CtBP, to LCoR-C/EBP␤ complex, which led to chromatin remodeling and transcription repression (Fig. 8G). Our studies and others suggest the existence of a transcription coregulator network that modulates adipogenesis. Further study of this network will help us understand adipose biology and identify potential drug targets for metabolic disease.

RNA extraction and quantitative PCR
Total RNA of cells was extracted using TRIzol reagent (Invitrogen 15596018) according to the manufacturer's instructions. Reverse transcription was performed using a PrimeScript RT reagent kit (TaKaRa). cDNAs were quantified in an ABI Real-Time System (Applied Biosystems). Relative gene expression levels were calculated using comparative C T (TATA-binding protein was used as the reference gene) and normalized as indicated. Primer sequences used for qPCR are provided in supplemental Table S3.

Isolation of SVF cells
Inguinal fat pads from C57BL/6J mice were minced and digested in Dulbecco's modified Eagle's medium (DMEM) containing 2 mg/ml type I collagenase (Invitrogen 17100017) and 10 mM HEPES (Invitrogen 15630080) for 40 min at 37°C with shaking, and an equal volume of DMEM containing 10% fetal bovine serum (FBS) was added to terminate the digestion. The digested fat was filtered through a sterile 100-m cell strainer (BD Falcon) to remove undigested fragments. The filtrate was centrifuged at 250 ϫ g for 10 min. The supernatant containing mature adipocytes was collected and washed twice with PBS before harvesting. The cell pellet was resuspended in PBS and further filtered through a sterile 40-m cell strainer (BD Falcon) to obtain SVF cells. The SVF cells were either harvested for RNA and protein analyses or cultured for adipogenic differentiation.

Retroviral infection
Retroviral packaging was achieved by cotransfection of pMSCV constructs or pSIREN-RetroQ with pCL-Eco vectors at a ratio of 1:1 into HEK 293T cells using Lipofectamine 2000 transfection reagent (Invitrogen 11668019). 36 and 48 h after transfection, the viral supernatants were collected, supplemented with 8 g/ml Polybrene, filtered through 0.45-m filter The role of LCoR during early adipogenesis (Millipore SLHV033RB), and added to cells. Stable cell lines were selected in the presence of 2.5 g/ml puromycin (Sigma P8833).

GST pulldown
LCoR-or CtBP1-GST fusion proteins were expressed and purified using glutathione-agarose (Santa Cruz Biotechnology sc-2009). In vitro transcription/translation was performed using the TNT T7 Quick Coupled Transcription/Translation System (Promega L1170) according to the manufacturer's instructions. Purified GST fusion proteins and in vitro translated proteins were mixed and incubated in BII buffer (20 mM Tris-Cl (pH 7.5), 100 mM NaCl, 2 mM EDTA, 0.1% Nonidet P-40, 2 mM DTT, 0.05% BSA, 5% glycerol) supplemented with phosphatase inhibitor at 4°C overnight. After washing five times with PBS, GST-precipitated complexes were separated by SDS-PAGE and detected by Western blotting.

LCoR protein complex purification
3T3-L1 preadipocytes stably expressing either pMSCVpuro-F-LCoR or pMSCVpuro empty vector were induced to differentiate as described under "Cell culture." After 24-h induction, the cells were harvested, and the nuclear proteins were extracted as described under "Cytoplasmic and nuclear protein extraction." The nuclear proteins were further diluted in IP lysis buffer and incubated with anti-FLAG M2 affinity gel (mouse; Sigma A2220) at 4°C overnight. The purified nuclear extracts were eluted by 0.1 M glycine HCl (pH 3.5). The eluted proteins were separated by SDS-PAGE and visualized by silver staining. Several prominent bands between 55 and 34 kDa were excised and analyzed by mass spectrometry at the research center for proteome analysis in Shanghai Applied Protein Technology Co., Ltd. Briefly, the immunoprecipitated proteins were digested by trypsin and subjected to reverse-phase liquid chromatography with tandem mass spectrometry (LC-MS/MS) using a high-resolution hybrid mass spectrometer (LTQ-Orbitrap, Thermo Scientific). MS/MS spectra were automatically searched against ipi.HUMAN.v3.53 using Bioworks-Browser rev. 3.1 (Thermo Electron, San Jose, CA). Protein identification results were extracted from SEQUEST.out files with BuildSummary.

Chromatin immunoprecipitation
ChIP was performed using a Magna ChIP G kit (Millipore 17-611) according to the manufacturer's instructions. Briefly, 3T3-L1 preadipocytes stably expressing pMSCVpuro-F-LCoR, pMSCVpuro-F-LCoR⌬N, pMSCVpuro-F-LCoR⌬C, or pMSCVpuro empty vector were grown in 10-cm dishes and induced to differentiate as described under "Cell Culture." The cells were cross-linked at the indicated time with 1% formaldehyde at room temperature for 10 min. Unreacted formaldehyde was quenched by adding glycine to a final concentration of 0.125 M at room temperature for 5 min. After washing twice with ice-cold PBS, cells were harvested by scraping and pelleted. Then the cells were resuspended in 0.5 ml of cell lysis buffer containing protease inhibitor and incubated on ice for 15 min. The cell suspension was centrifuged at 800 ϫ g at 4°C for 5 min. The pellet was resuspended in 0.5 ml of nuclear lysis buffer and sonicated to shear cross-linked DNA to 200 -1000 base pairs. After centrifugation at 12,000 ϫ g at 4°C for 10 min, protein-DNA complexes were diluted at a ratio of 1:10 with ChIP dilution buffer, and 1% of the mixture was removed as "input." The chromatin fractions were incubated at 4°C overnight with 20 l of Protein G magnetic beads and 2 g of one of the following antibodies: anti-FLAG M2 (mouse monoclonal), anti-C/EBP␤ (mouse monoclonal), anti-CtBP1, anti-H3K9acK14ac (Millipore 06-599), anti-H3K9me2 (Millipore 17-648). After washing in low-salt buffer, high-salt buffer, LiCl buffer, and Tris-EDTA buffer, proteins were eliminated using Proteinase K at 62°C for 2 h with shaking. DNA was purified using the kit column and analyzed by qPCR. Primer sequences used in the ChIP assays are provided in supplemental Table S3.

Cytoplasmic and nuclear protein extraction
Cytoplasmic and nuclear proteins were extracted using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Pierce 78833) according to the manufacturer's instructions. Briefly, 3T3-L1 preadipocytes were induced to differentiate for 2 days as described under Cell culture." The cells were harvested with trypsin-EDTA at the indicated time and centrifuged at 500 ϫ g for 5 min. After washing twice with PBS, cells were resuspended in Cytoplasmic Extraction Reagent (CER) I and incubated on ice for 10 min. CER II was then added and vortexed for 5 s. The cells were centrifuged at 16,000 ϫ g for 5 min, and supernatants were collected as the cytoplasmic extract. The pellets were The role of LCoR during early adipogenesis resuspended in ice-cold Nuclear Extraction Reagent (NER) and incubated on ice for 40 min; during this incubation, lysates were vortexed for 15 s every 10 min. Finally, cells were centrifuged at 16,000 ϫ g for 10 min to obtain the nuclear extract.

BrdU incorporation assay
3T3-L1 cells were induced to differentiate as described under "Cell culture." 18 h after induction, cells were labeled for 2 h with 10 M BrdU and washed three times with PBS before fixation with 4% formaldehyde. After denaturation (2 N HCl for 20 min) and neutralization (0.1 M sodium borate, pH 8.5), cells were incubated with anti-BrdU monoclonal antibody (Santa Cruz Biotechnology sc-32323) at 4°C overnight and then incubated with Alexa Fluor 488-conjugated secondary antibody for 1 h at room temperature. Cells were also stained with DAPI for 15 min at room temperature to enable nuclear detection. Staining was evaluated and photographed by high-content screening instruments (Thermo Fisher Scientific Cellomics).

Immunofluorescence
3T3-L1 cells cultured on coverslips were induced to differentiate for 24 h before fixation in a solution containing 4% formaldehyde (Polysciences Inc.) and 4% sucrose (Sigma-Aldrich) in PBS for 30 min. After washing three times with PBS, the cells were blocked with blocking buffer containing 0.5% Triton X-100 and 5% serum in PBS for 30 min and then coincubated overnight at 4°C with primary antibodies against FLAG (mouse monoclonal) and C/EBP␤ (rabbit polyclonal) or coincubated with primary antibodies against C/EBP␤ (mouse monoclonal) and CtBP1 as indicated. Primary antibodies were diluted with blocking buffer. After washed three times again with PBS, the cells were incubated with Alexa Fluor 488-and 555-conjugated secondary antibodies (Invitrogen A21202 and A31572) for 1 h. Secondary antibodies were diluted with 0.5% Triton X-100 and 2% serum in PBS. Nuclei were stained with DAPI in PBS for 5 min. Confocal images were obtained with a confocal microscope (Olympus FW1000).

Luciferase reporter assay
Transient transfection assay was performed in 293T cells to analyze C/EBP␤ activity. 293T cells were seeded in 24-well plates and grown to 80 -90% confluence, and then cells in each well were transfected with 125 ng of pGL3-PPAR␥2 promoterreporter construct (Ϫ602 to ϩ52) alone or in combination with 25 ng of pMSCVpuro-C/EBP␤ in the absence or presence of increasing quantities of pMSCVpuro-LCoR (5, 10, and 20 ng), 25 nM siRNAs targeting human LCoR, or 20 ng of full-length or truncated LCoR as indicated in the legends with Lipofectamine 2000 transfection reagent. The quantity of transfected DNA was kept constant in all conditions by adding a corresponding amount of empty vector or scrambled siRNAs. Sequences of siRNAs targeting human LCoR are provided in supplemental Table S3. 5 ng of pRL-TK (Promega) carrying Renilla luciferase was cotransfected as an internal control. 36 h after transfection, 293T cells were harvested, and the luciferase activities were analyzed using the Dual-Luciferase assay kit (Promega) according to the manufacturer's instructions. To further analyze the effect of LCoR on C/EBP␤ activity in 3T3-L1 cells, 3T3-L1 cells were infected with retrovirus expressing a PPAR␥2 promoter (Ϫ602 to ϩ52)-driven luciferase reporter with or without retrovirus expressing LCoR, shLCoR, LCoR⌬N, or LCoR⌬C as indicated in the legends. The relative luciferase activity was normalized against the protein concentration of each cell lysate sample. 48 h after induction, the 3T3-L1 cells were harvested, and the luciferase activities were analyzed using a luciferase assay kit (Promega).

Statistical analysis
All experiments were performed at least three times. Data are presented as means Ϯ S.D. Error bars represent S.D. Student's t test was performed to assess whether the means of two groups are statistically significant from each other (p Ͻ 0.05). GraphPad Prism 5.0 software was applied to all statistical analyses. The density of specific bands from immunoblots shown here was measured by ImageJ (1.29v, National Institutes of Health) and normalized to either protein input in GST pulldown and immunoprecipitation assays as indicated or protein levels of GAPDH, tubulin, or ␤-actin because we did not see any evidence showing that our experimental manipulations would affect their expression (supplemental Table S4).