ABSTRACT
Previous studies have shown that CCAAT/enhancer-binding protein α (C/EBPα) plays a very important role during adipocyte terminal differentiation and that AP-2α (activator protein 2α) acts as a repressor to delay the expression of C/EBPα. However, the mechanisms by which AP-2α prevents the expression of C/EBPα are not fully understood. Here, we present evidence that Suv39h1, a histone H3 lysine 9 (H3K9)-specific trimethyltransferase, and G9a, a euchromatic methyltransferase, both interact with AP-2α and enhance AP-2α-mediated transcriptional repression of C/EBPα. Interestingly, we discovered that G9a mediates dimethylation of H3K9, thus providing the substrate, which is methylated by Suv39h1, to H3K9me3 on the C/EBPα promoter. The expression level of AP-2α was consistent with enrichment of H3K9me2 and H3K9me3 on the C/EBPα promoter in 3T3-L1 preadipocytes. Knockdown of Suv39h markedly increased C/EBPα expression and promoted adipogenesis. Conversely, ectopic expression of Suv39h1 delayed C/EBPα expression and impaired the accumulation of triglyceride, while simultaneous knockdown of AP-2α or G9a partially rescued this process. These findings indicate that Suv39h1 enhances AP-2α-mediated transcriptional repression of C/EBPα in an epigenetic manner and further inhibits adipocyte differentiation.
INTRODUCTION
Obesity is the major risk factor for metabolic syndrome, a condition characterized by insulin resistance, type 2 diabetes, hyperlipidemia, and other metabolic disorders (1). Elucidation of the mechanism of adipogenesis may provide a way to treat obesity. The 3T3-L1 preadipocyte cell line has been one of the most well characterized and widely used models to investigate the adipocyte differentiation program. Upon treatment with differentiation inducers, growth-arrested 3T3-L1 preadipocytes express CCAAT enhancer-binding protein β (C/EBPβ), which then activates expression of CCAAT enhancer-binding protein α (C/EBPα) and peroxisome proliferator-activated receptor γ (PPARγ), both of which are required for differentiation (2–4). PPARγ and C/EBPα mutually regulate each other and coordinately induce expression of adipogenic genes, including the 422/aP2, SCD1, Glut4, and obese genes, leading to the adipocyte phenotype (4–6).
C/EBPα not only takes an active part in adipogenesis but also can be detected in a variety of organs, such as liver, lung, kidney, small intestine, brain, and the hematopoietic system, and it acts as a key transcription factor involved in the regulation of cell proliferation and differentiation (7–9). C/EBPα knockout mice die after birth due to defective gluconeogenesis of the liver and subsequent hypoglycemia (10). C/EBPα may also act as a tumor suppressor, and low levels of C/EBPα expression have been found in patients with lung cancer, hepatocellular carcinomas, breast cancer, squamous skin carcinomas, acute myeloid leukemia (AML), or head and neck squamous cell carcinoma (11–16).
Activator protein 2α (AP-2α) is a critical regulator of gene expression during vertebrate development, embryogenesis, and cell differentiation (17–19). Early studies demonstrated that AP-2α is responsible for delaying C/EBPα expression to ensure the progression of mitotic clonal expansion (MCE), which is required for the early stages of adipocyte differentiation and subsequent terminal differentiation (20, 21). Several reports recently found that AP-2α suppresses C/EBPα expression by DNA promoter methylation in head and neck squamous cell carcinoma (22). Other studies have shown that DNA methylation correlates well with histone H3 lysine 9 (H3K9) methylation and that both result in gene inactivation (23–26). Thus, epigenetic histone modifications may underlie AP-2α-mediated inhibition of C/EPBα expression.
Epigenetic mechanisms, in particular histone modifications, play indispensable roles in adipocyte differentiation (27). Recent studies have shown that H3K9 methyltransferases SETDB1 and G9a have a great influence on adipogenesis (28, 29). In particular, G9a-mediated H3K9me2 is selectively enriched on the entire PPARγ locus in preadipocytes and represses PPARγ expression; however, the sequence-specific transcription factors that recruit it to the entire PPARγ locus have not been identified (29). Furthermore, the function of Suv39h1 and Suv39h1-mediated H3K9me3 are poorly understood during adipogenesis. H3K9me3 formation by Suv39h1 has been well characterized in heterochromatic gene silencing (30, 31), but recent reports certified that Suv39h1-mediated H3K9me3 also occurs on euchromatic gene promoters involved in cell lineage commitment, differentiation, proliferation, and inflammation (32–37). In the current study, we show that Suv39h1 enhances AP-2α-dependent repression of C/EBPα by H3K9me3 and further influences adipogenesis.
MATERIALS AND METHODS
Cell culture and induction of differentiation.The stromal vascular fraction (SVF) of subcutaneous adipose tissue was obtained from mice and was cultured in F-12–Dulbecco's modified Eagle's medium (DMEM) plus 10% fetal bovine serum (FBS; Gibco). 3T3-L1 preadipocytes were propagated and maintained in DMEM containing 10% (vol/vol) calf serum (Gibco). Two days postconfluence (designated day 0), cells were induced to differentiate with DMEM containing 10% (vol/vol) FBS (Gibco), 1 μg/ml insulin, 1 μM dexamethasone, and 0.5 mM 3-isobutyl-1-methyl-xanthine until day 2. Cells were then fed DMEM supplemented with 10% FBS and 1 μg/ml insulin for 2 days, after which they were fed every other day with DMEM containing 10% FBS. Expression of adipocyte genes and acquisition of adipocyte phenotype began at day 3 and reached a maximum at day 6.
Oil Red O staining.Cells were washed three times with phosphate-buffered saline (PBS) and then fixed for 15 min with 3.7% formaldehyde. Oil Red O (0.5% in isopropanol) was diluted with water (3:2) and incubated with the fixed cells for 2 h at room temperature. Cells were washed with water, and the stained fat droplets in the cells were visualized by light microscopy and photographed.
Plasmid constructs.Murine Suv39h1 and AP-2α cDNA were generated by PCR using the following primers: for Suv39h1, sense, 5′-CCGCTCGAGATGGACTACAAAGACGATGACGACAAGGCGGAAAATTTAAAAGGTTGC-3′, and antisense, 5′-CCGGAATTCCTAGAAGAGGTATTTTCGGCAAGC-3′; for AP-2α, sense, 5′-CCGCTCGAGATGTACCCATACGACGTCCCAGACTACGCTCTTTGGAAACTGACGGATAATATC-3′, and antisense, 5′-CCGGAATTCTCACTTTCTGTGTTTCTCTTCTTTGTC-3′. The PCR product was cloned into a murine stem cell virus (MSCV) retroviral vector with XhoI and EcoRI. Plasmids harboring shRNA against murine Suv39h or AP-2α were designed by using a lentivirus system and the following primers: for the Suv39h sequence, 5′-TACCTCTTTGACCTGGACTAC-3′; for the AP-2α sequence, 5′-GCAGAATTTCTCAACCGACAA-3′. As a control, we used an shRNA vector without hairpin oligonucleotides. All plasmids were confirmed by DNA sequencing. 293T cells cultured in serum-free DMEM were transfected with MSCV or MSCV-Flag-Suv39h1 or MSCV-HA-AP-2α and Ecopac plasmids and pLKO.1-vector or pLKO.1-shSuv39h or pLKO.1-shAP-2α and Δ8.9, and also plasmids carrying the gene for the G protein of vesicular stomatitis virus. Fresh medium containing 10% fetal bovine serum was provided 4 to 6 h after transfection, and the viral medium was collected at 48 to 72 h. 3T3-L1 cells were infected with viruses at 20 to 30% confluence with Polybrene (8 μg/ml).
Western blotting and antibodies.Cells were scraped into lysis buffer containing 2% SDS and 50 mM Tris-HCl (pH 6.8). Lysates were then quantitated, and equal amounts of protein were subjected to SDS-PAGE and immunoblotted with antibodies against Suv39h1, HSP90, PPARγ, C/EBPα, AP-2α, H3K9me2, H3K9me3, and Flag protein. Antibodies against PPARγ (catalog number 2443) and G9a (3306) were from Cell Signaling Technology. Antibodies against H3K9me2 (04-768), H3K9me3 (05-1242), and Suv39h1 (05-615) were from Millipore, and antibodies against HSP90 (sc7947), AP-2α (sc8975 and sc12726), and C/EBPα (sc61) were from Santa Cruz Biotechnology. Antibody against Flag was from Sigma. Antibody against hemagglutinin (HA; 561-5) was obtained from MBL (Nagoya, Japan).
Immunoprecipitation assay.Cells were washed with PBS, scraped off, and collected by centrifugation. Isolation of nuclear proteins was performed with a nuclear extraction kit (Key Gen). The nuclear extract was incubated with the antibodies anti-Flag (Sigma), anti-AP-2α, and anti-G9a at 4°C overnight. The next day, protein A-agarose beads were added. After 2 h of incubation, the beads were washed with Tris-buffered saline with 0.5% Tween 20. The immunoprecipitates were separated by SDS-PAGE and subjected to Western blotting.
Luciferase reporter assay.Promoter regions of mouse Suv39h1 (bp −2021 to +12) and C/EBPα (bp −2044 to +100) were amplified via PCR from genomic DNA of 3T3-L1 cells and cloned into PGL3-basic (Promega, Madison, WI). Specific site mutations were made using a KOD-Plus mutagenesis kit (Toyobo, Japan). Cells were transfected by Lipofectamine 2000 (Invitrogen). Luciferase activity was measured via a dual luciferase reporter assay (Promega), with firefly luciferase activity normalized to renilla activity.
Chromatin immunoprecipitation analysis.For chromatin immunoprecipitation (ChIP), cells were fixed with 1% formaldehyde for 10 min at room temperature with swirling. Glycine was added to a final concentration of 0.125 M, and the incubation was continued for an additional 5 min. Cells were washed twice with ice-cold PBS, harvested by scraping, pelleted, and resuspended in 1 ml of SDS lysis buffer (50 mM Tris-HCl [pH 8.0], 1% SDS, 10 mM EDTA, and protease inhibitors). Samples were sonicated three times for 10 min each with a Bioruptor sonicator (Diagenode, Denville, NJ). Samples were centrifuged at 12,000 × g at 4°C for 5 min. After removal of an input aliquot (whole-cell extract), supernatants were diluted 10-fold in ChIP dilution buffer (20 mM Tris-HCl [pH 8.0], 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, and Complete protease inhibitor tablets). Samples were immunoprecipitated using H3K9me2, H3K9me3, AP-2α, G9a, and Suv39h1 antibodies or a nonspecific IgG control. Immunoprecipitated samples were eluted and reverse cross-linked by incubation overnight at 65°C in elution buffer (50 mM Tris-HCl [pH 8.0], 10 mM EDTA, 1% SDS). Genomic DNA was then extracted by using a PCR purification kit (Qiagen, Valencia, CA). Purified DNA was subjected to real-time quantitative PCR (qPCR) by using the specific primers for the PPARγ2, C/EBPα, and aP2 promoters. Primers used were the following: for PPARγ2, sense, 5′-TTCAGATGTGTGATTAGGAG-3′, and antisense, 5′-AGACTTGGTACATTACAAGG-3′; for C/EBPα, sense, 5′-TCCCTAGTGTTGGCTGGAAG-3′, and antisense, 5′-CAGTAGGATGGTGCCTGCTG-3′; for aP2, sense, 5′-GAAGGTCAAATGTGTCCAAG-3′, and antisense, 5′-GACCCTGTATGTTTTCCTCTG-3′; for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), sense, 5′-CGCCGCCATGTTGCA-3′, and antisense, 5′-GGAAGGCCTAAGCAAGATTTCA-3′. ChIP data were normalized to control IgG or expressed as a percentage of input. Re-ChIPs were done essentially the same as primary ChIPs. Beads from the first ChIP were incubated with 10 mM dithiothreitol at 37°C for 30 min, and the eluted immunocomplex was diluted 1:50 in dilution buffer followed by reimmunoprecipitation with the second antibodies.
RNA isolation and RT-qPCR.Total RNAs were extracted with TRIzol (Invitrogen), and the mRNA was reverse transcribed by using a Revert Aid first-strand cDNA synthesis kit (Fermentas). Then, the cDNA were analyzed using the Power SYBR green PCR kit with the ABI Prism 7300 qPCR machine (Applied Biosystems). All qPCR data were normalized to the GAPDH results. Primer sequences for reverse transcription quantitative PCR (RT-qPCR) were as follows: for Suv39h1, sense, 5′-CGCATCGCATTCTTTGCC-3′, and antisense, 5′-AAGCCGTTGTCCCACATTTG-3′; for Suv39h2, sense, 5′-ATCTACGAATGCAACTCAAGGTG-3′, and antisense, 5′-CCACAGCCATTGCTAGTTCTAA-3′; for G9a, sense, 5′-AAAACCATGTCCAAACCTAGCAA-3′, and antisense, 5′-GCGGAAATGCTGGACTTCAG-3′; for SETDB1, sense, 5′-CTTCTGGCTCTGACGGTGATG-3′, and antisense, 5′-GGAAGCCATGTTGGTTGATT3′; for AP-2α, sense, 5′- CACCCCTAATGCCGACTTC-3′, and antisense, 5′- GGTCGTTGACGTGGGAGTA-3′; for C/EBPα, sense, 5′- CAAGAACAGCAACGAGTACCG-3′, and antisense, 5′- GTCACTGGTCAACTCCAGCAC-3′; for PPARγ2, sense, 5′- GTGCCAGTTTCGATCCGTAGA-3′, and antisense, 5′- GGCCAGCATCGTGTAGATGA-3′; for aP2, sense, 5′- ACACCGAGATTTCCTTCAAACTG-3′, and antisense, 5′- CCATCTAGGGTTATGATGCTCTTCA-3′; for GAPDH, sense, 5′-GGCAAATTCAACGGCACAGT-3′, and antisense, 5′-CGCTCCTGGAAGATGGTGAT-3′.
Statistical analysis.Results are expressed as means ± standard errors (SE). Comparisons between groups were made by using an unpaired two-tailed Student's t test, where a P value of <0.05 was considered statistically significant. All experiments were repeated at least three times, and representative data are shown.
RESULTS
The expression of AP-2α and Suv39h1 is inversely correlated with C/EBPα, PPARγ, and aP2 expression during 3T3-L1 preadipocyte differentiation.Previous studies showed that AP-2α has a classic binding site on the C/EBPα promoter and represses C/EBPα expression (38). But it has been unclear whether AP-2α-mediated inhibition of C/EBPα expression can be explained by epigenetic histone silencing. So, we first tested the expression profiles of Suv39h1, AP-2α, and the master transcription factors C/EBPα, PPARγ, and their downstream gene, aP2, during 3T3-L1 preadipocyte differentiation. Interestingly, we found that the expression profiles of Suv39h1 and AP-2α were similar and that protein levels of both were gradually downregulated after adipogenic induction, but the expression of C/EBPα and PPARγ significantly increased from day 2 and their downstream gene, aP2, began to express at day 3 during differentiation (Fig. 1A). We also checked Suv39h1 and -2 mRNA levels and found that they clearly decreased from day 2 (see Fig. S1A in the supplemental material). Actually, the high levels of AP-2α and Suv39h1 were found in SVF from subcutaneous adipose tissue, relative to levels in adipocytes, but Suv39h2 was not detected (see Fig. S1B). Notably, global H3K9me3 also decreased, along with Suv39h1 (Fig. 1A). In order to determine whether H3K9me3 mediated repression of C/EBPα at the beginning of differentiation, ChIP assays were performed, and we found that H3K9me2 and H3K9me3 were both enriched on the promoter of C/EBPα at day 0 but that the phenomenon disappeared at day 3, when C/EBPα was expressed (Fig. 1B). Meanwhile, we also showed the same enrichment on the promoters of PPARγ2 and aP2 at day 0 (see Fig. S1C).
Suv39h1 and G9a are recruited by AP-2α to the promoter of C/EBPα in 3T3-L1 preadipocytes. (A) Cell lysates were harvested at the times indicated and subjected to Western blotting with the indicated antibodies. Hsp90 served as a loading control. (B) At the indicated times, cell lysates were prepared and ChIP-qPCR was performed to show the enrichment levels of H3K9me2 and H3K9me3 on the C/EBPα promoter. GAPDH was used as a negative control. (C) ChIP-PCR was performed on postconfluent 3T3-L1 cells with anti-AP-2α, -Suv39h1, and G9a antibodies and nonspecific IgG. GAPDH was used as a negative control. (D) Interaction between AP-2α and Suv39h1 and G9a. Cellular nuclear extracts were prepared and immunoprecipitated with antibody against AP-2α. (E) Soluble chromatin from postconfluent 3T3-L1 preadipocytes was prepared for ChIPs and re-ChIPs with the indicated antibodies. Recovery of the promoter of the C/EBPα gene was analyzed by PCR. Data are presented as means ± standard deviations from 3 independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Suv39h1 and G9a are recruited by AP-2α to the promoter of C/EBPα in 3T3-L1 preadipocytes.Suv39h1 is a H3K9 trimethyltransferase and represses gene activation by adding H3K9me3 to promoter regions (30, 31). G9a has been shown to provide H3K9me2, the substrate that is methylated by Suv39h1 to form H3K9me3 (39, 40). ChIP and qPCR showed that Suv39h1 and G9a are both enriched on the promoter regions of C/EBPα (Fig. 1C). Because Suv39h1 and G9a both lack DNA binding activity, we hypothesized that AP-2α recruits Suv39h1 and G9a to the promoter of C/EBPα, which might silence the C/EBPα gene in an epigenetic manner by sustaining H3K9me3 on the promoter regions. The interaction between AP-2α and Suv39h1 and G9a was confirmed via coimmunoprecipitation experiments. Immunoprecipitation of Flag-Suv39h1-overexpressing 3T3-L1 preadipocytes with Flag antibody pulled down AP-2α, and AP-2α antibody also immunoprecipitated Suv39h1 and G9a (Fig. 1D; see also Fig. S1D in the supplemental material). We further explored the interaction between AP-2α and G9a in 3T3-L1 preadipocytes (see Fig. S1D). We found that AP-2α also bound on the promoter regions of C/EBPα, as with Suv39h1 and G9a, based on ChIP-qPCR (Fig. 1C). Consistently, re-ChIP analysis also showed that AP-2α, Suv39h1, and G9a formed a complex and were all enriched on the promoter of C/EBPα (Fig. 1E). Taken together, these results suggest that Suv39h1 and G9a are present in association with AP-2α on the promoter of C/EBPα.
Suv39h1 and G9a enhance AP-2α-mediated transcriptional repression of C/EBPα.To identify the role of Suv39h1 and G9a in AP-2α-dependent inhibition of C/EBPα expression, the C/EBPα promoter region containing the AP-2α binding site, or the same promoter region containing a mutation that disrupts binding to AP-2α, was subcloned into the luciferase reporter construct pGL3-basic and transfected into 3T3-L1 preadipocytes with or without the AP-2α plasmid. Luciferase activity of the wild-type C/EBPα promoter was obviously decreased when cotransfected with Suv39h1 and G9a expression plasmids; however, this decrease was not observed with the mutant C/EBPα promoter construct. The results demonstrated that Suv39h1 and G9a enhanced the capability of AP-2α to repress C/EBPα transactivation. In order to verify whether the enrichment of H3K9me3 on the C/EBPα promoter depended on AP-2α, we treated cells with AP-2α shRNA to downregulate AP-2α expression and found that not only was the enrichment of Suv39h1, G9a, H3K9me2, and H3K9me3 on the C/EBPα promoter reduced but also adipogenesis was promoted, as indicated by Oil Red O staining and expression of adipogenic genes, such as the C/EBPα, PPARγ, and 422/aP2 genes (Fig. 2B to D). Conversely, ectopic expression of AP-2α resulted in an increased enrichment of Suv39h1, G9a, H3K9me2, and H3K9me3 on the C/EBPα promoter, and the differentiation process was inhibited completely. However, C/EBPα overexpression partially rescued adipogenesis, as indicated by Oil Red O staining and expression of adipogenic genes (Fig. 2E to G). Furthermore, we found a similar phenotype in SVF from subcutaneous adipose tissue (Fig. 2H and I). In order to identify the role of G9a and this complex, we knocked down G9a and found that H3K9me2 and H3K9me3 were both decreased on the C/EBPα promoter (see Fig. S2 in the supplemental material). Collectively, these results verified that Suv39h1 and G9a with AP-2α form a complex on the promoter of C/EBPα and repress its expression in an epigenetic manner.
Suv39h1 and G9a enhance AP-2α-mediated transcriptional repression of C/EBPα. (A) 3T3-L1 preadipocytes were transfected with reporter constructs with or without the indicated expression plasmid. After 48 h, luciferase activity was analyzed and plotted. (B) 3T3-L1 preadipocytes were treated with control or AP-2α shRNA. Cell lysates were prepared, and ChIP-qPCR was performed to show the enrichment levels of Suv39h1, G9a, H3K9me2, and H3K9me3 on the C/EBPα promoter. (C) Oil Red O staining at day 6 after 1/5 3-isobutyl-1-methylxanthine–dexamethasone–insulin (MDI) induction. (D) Expression levels of AP-2α, C/EBPα, PPARγ, and aP2 were analyzed by Western blotting. (E) 3T3-L1 preadipocytes were treated with MSCV vector or MSCV AP-2α. Cell lysates were prepared, and ChIP-qPCR was performed to show the enrichment levels of Suv39h1, G9a, H3K9me2, and H3K9me3 on the C/EBPα promoter. (F) MSCV-vector or MSCV-AP-2α-overexpressing 3T3-L1 cells were treated with MSCV-vector or MSCV-HA-C/EBPα. Oil Red O stain was applied at day 6. (G) Expression levels of AP-2α, HA-C/EBPα, C/EBPα, PPARγ, and aP2 were analyzed by Western blotting. (H) MSCV-vector or MSCV-HA-C/EBPα was ectopically expressed in SVF cells with MSCV-AP-2α or MSCV-vector. At 8 days after induction, cells were stained with Oil Red O. (I) Expression levels of AP-2α, C/EBPα, PPARγ2, and aP2 were analyzed by RT-qPCR. Data are presented as means ± standard deviations from 3 independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
Suv39h1 modulates adipogenic differentiation by influencing C/EBPα, PPARγ, and aP2 expression in an epigenetic manner.C/EBPα and PPARγ have been shown to play a pivotal role in terminal differentiation (41, 42). To probe whether Suv39h mediates enrichment of H3K9me3 on the promoters of C/EBPα, PPARγ, and aP2, 3T3-L1 preadipocytes were treated with control or Suv39h shRNA. We found that knockdown of Suv39h significantly decreased H3K9me3 but not H3K9me2 compared with the control (Fig. 3B). We also determined that the expression levels of G9a and SETDB1 were not affected by Suv39h shRNA (Fig. 3A). Knockdown of Suv39h resulted in a drastic decrease in the enrichment of H3K9me3 on the promoters of C/EBPα, PPARγ2, and aP2 compared with the control. In contrast, the enrichment of H3K9me2 on the promoters of C/EBPα, PPARγ2, and aP2 slightly increased (Fig. 3C). Furthermore, the protein levels of C/EBPα, PPARγ, and aP2 were significantly increased at corresponding time points during differentiation, and this was consistent with an earlier appearance of lipid droplets accumulation, in contrast with the control (Fig. 3D and E). Additionally, we also found the similar phenotype in the SVF from subcutaneous adipose tissue (Fig. 3F and G).
Knockdown of Suv39h activates C/EBPα expression and promotes adipogenesis. (A) 3T3-L1 cells were treated with control or Suv39h shRNA. Knockdown of Suv39h was confirmed by RT-qPCR and Western blotting at day 0, and the expression levels of SETDB1 and G9a were also determined by RT-qPCR after knockdown of Suv39h. (B) The effects of knockdown of Suv39h on total protein levels of H3K9me2 and H3K9me3 were analyzed by Western blotting. H3 served as a loading control. (C) ChIP-qPCR analysis of the effects of Suv39h knockdown on enrichment levels of H3K9me2 and H3K9me3 on C/EBPα, PPARγ2, and aP2 promoters. (D) Expression levels of C/EBPα, PPARγ, and aP2 were analyzed by Western blotting at the times indicated. Hsp90 served as a loading control. (E) Cells stained or not with Oil Red O were visualized and photographed at the times indicated. (F) SVF cells were treated with control shRNA, AP-2α shRNA, or Suv39h shRNA. Oil Red O staining was performed at day 8 after induction. (G) Expression levels of AP-2α, Suv39h1, C/EBPα, PPARγ2, and aP2 were analyzed by RT-qPCR. Data are presented as means ± standard deviations from 3 independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.
To further validate the role of Suv39h1 during 3T3-L1 differentiation, 3T3-L1 preadipocytes stably expressing Flag-Suv39h1 were established by using a retroviral system. ChIP-qPCR analysis revealed that the enrichment of H3K9me2 on the promoters of C/EBPα, PPARγ2, and aP2 increased markedly at day 3 compared with the control. Strikingly, at day 6 high enrichment of H3K9me3 was obvious on the promoter of aP2 (Fig. 4A and B). Overexpression of Suv39h1 blocked the terminal differentiation of 3T3-L1 preadipocytes, as indicated by expression of the adipocyte marker aP2, as well as by Oil Red O staining (Fig. 4C and D). Ectopic expression of Suv39h1 markedly delayed the expression of C/EBPα and PPARγ until day 5; aP2 expression was not only delayed but also markedly decreased at the protein level (Fig. 4E). This was consistent with Oil Red O staining results (Fig. 4F). Thus, these observations strongly indicate that Suv39h1 represses C/EBPα, PPARγ, and aP2 expression by adding H3K9me3 to their promoters in 3T3-L1 preadipocytes and that Suv39h1 inhibits adipogenesis.
Ectopic expression of Suv39h1 retards C/EBPα expression and represses adipogenesis. (A and B) Suv39h1-overexpressing 3T3-L1 preadipocytes were cultured until 2 days postconfluence and then differentiation was induced. Cells were collected at the times indicated and subjected to ChIP-qPCR analysis to determine the enrichment levels of H3K9me2 and H3K9me3 acting on the C/EBPα, PPARγ2, and aP2 promoters. Data were normalized to results with control cells at day 0. (C) The effect of Suv39h1 overexpression on adipocyte differentiation was assessed at day 6 by Oil Red O staining. (D) Flag-Suv39h1 and 422/aP2 expression was detected by Western blotting. (E) Expression of C/EBPα, PPARγ, and aP2 was analyzed by Western blotting at the times indicated. Hsp90 served as a loading control. (F) Cells were stained with Oil Red O at day 6 and photographed at the times indicated. Data are presented as means ± standard deviations from 3 independent experiments. *, P < 0.05; ***, P < 0.001.
Having uncovered the role of Suv39h1 during adipogenesis, we wished to determine whether the function of Suv39h1 depends on AP-2α and G9a. We used shRNAs to knock down expression of AP-2α or G9a in Flag-Suv39h1-overexpressing 3T3-L1 preadipocytes. As expected, knockdown of AP-2α or G9a could both partially rescue adipogenesis and modestly ease repression of C/EBPα expression, as indicated using Oil Red O staining and Western blotting (Fig. 5A to F). Actually, we performed the same experiment in SVF cells and had similar results (Fig. 5G and H). These data reinforce the idea that G9a provides the substrate for Suv39h1 and that AP-2α represses C/EBPα expression and adipogenesis, partially through Suv39h1, in an epigenetic manner (Fig. 6).
Knockdown of AP-2α or G9a partially rescues adipogenesis in Suv39h1-overexpressing cells. (A, B, D, and E) Flag-Suv39h1-overexpressing 3T3-L1 cells were treated with AP-2α or G9a shRNA. Expression of AP-2α, G9a, and aP2 was analyzed by Western blotting, and Oil Red O staining was performed at day 6. (C and F) Expression of C/EBPα, PPARγ, and aP2 was analyzed by Western blotting at the times indicated. (G) SVF cells with retroviruses expressing Suv39h1 or vector were infected with AP-2α shRNA or G9a shRNA. Oil Red O staining was performed at day 8 after induction. (H) Expression levels of AP-2α, Suv39h1, G9a, C/EBPα, PPARγ2, and aP2 were analyzed by RT-qPCR. Data are presented as means ± standard deviations from 3 independent experiments. **, P < 0.01; ***, P < 0.001.
Model of Suv39h1-mediated AP-2α-dependent inhibition of C/EBPα expression. AP-2α recruits G9a and Suv39h1 to the C/EBPα promoter and inhibits C/EBPα expression via H3K9me3, further influencing adipogenesis.
DISCUSSION
Adipogenesis is a complicated process that is tightly regulated by an elaborate network of transcription factors that control the expression of hundreds of genes responsible for establishing the mature adipocyte phenotype (43). C/EBPα is one of the master regulators during adipogenesis, and its expression has been shown to be necessary for initiating terminal differentiation both in vivo and in vitro (44, 45). Histone epigenetic modifications have a pivotal role in controlling transcriptional regulation (27). To better understand the differentiation program, it is necessary to further elucidate the mechanism of adipocyte differentiation from an epigenetic perspective.
Previous studies have shown that H3K9me2 is enriched on the promoters of C/EBPα and PPARγ in preadipocytes and decreases with the increase of H3K4me3 during differentiation (29, 46). But it is poorly understood whether H3K9me3 contributes to the repression of C/EBPα in the early stages of differentiation and whether Suv39h1 has any function during adipogenesis. Our data suggest that Suv39h1 is responsible for the enrichment of H3K9me3 on C/EBPα, PPARγ2, and aP2 promoters in 3T3-L1 preadipocytes (Fig. 1B; see also Fig. S1C in the supplemental material). Knockdown of Suv39h abrogated the enrichment of H3K9me3 on C/EBPα, PPARγ2, and aP2 promoters and was associated with a slight increase in H3K9me2 enrichment on these same promoters (Fig. 3C). This is consistent with the many reports showing that Suv39h can specifically act on the substrate of H3K9me2, resulting in the production of H3K9me3 (39, 40).
The balance of histone methyltransferase and demethylase activities regulates the changing status of histone methylation on promoters (47). Our lab recently showed that histone demethylase Kdm4b functions as a cofactor of C/EBPβ to promote mitotic clonal expansion during differentiation (48). When inducer was added, histone demethylase began to remove H3K9me3 from C/EBPα and PPARγ promoters until C/EBPα and PPARγ were expressed completely at day 3 (Fig. 1A). Inversely, H3K9me2 provided the substrate for Suv39h1 to maintain C/EBPα and PPARγ silence. So, the mutual antagonistic functions of methytranferase and demeythlase resulted in an increase of H3K9me2 when forced expression of Suv39h1 was forced at day 3 (Fig. 4A). Further study is needed to determine which demethylases are responsible for removing H3K9me3 to regulate C/EBPα and PPARγ expression.
C/EBPα and PPARγ mutually regulate and form a feed forward regulatory loop to facilitate terminal differentiation (49). We found that ectopic expression of Suv39h1 delayed C/EBPα and PPARγ expression until day 5. This raises the question of why terminal differentiation remains blocked when C/EBPα expression recovers at day 5. Expression of aP2, an important adipogenetic marker, may be the key. ChIP results showed that H3K9me3 is remarkably enriched on the promoter of the aP2 gene at day 6. This means that C/EBPα and PPARγ cannot regulate aP2 expression, and terminal differentiation remains blocked.
In conclusion, we have identified a novel mechanism for the epigenetic control of C/EBPα expression via H3K9me3.
ACKNOWLEDGMENTS
This work was supported by National Key Basic Research Project grants (2011CB910201 to Q. Q. Tang and 2013CB530601 and 2011CBA01103 to X. Li), National Natural Science Foundation grants (81270954 and 30870510 to X. Li), the National Natural Science Foundation of China (31030048 and 81390350 to Q. Q. Tang), the Shanghai Rising Star Program (13QH1400800 to X. Li), and the Shanghai New Excellent Medicine Talents Program (XYQ2011037 to X. Li). The Department of Biochemistry and Molecular Biology at Fudan University Shanghai Medical College is supported by Shanghai Leading Academic Discipline Project B110 and 985 Project 985III-YFX0302.
FOOTNOTES
- Received 14 January 2014.
- Returned for modification 31 January 2014.
- Accepted 5 April 2014.
- Accepted manuscript posted online 14 April 2014.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/MCB.00070-14.
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
