Abstract
The epigenomic reader Brd4 is an important drug target for cancers. However, its role in cell differentiation and animal development remains largely unclear. Using two conditional knockout mouse strains and derived cells, we demonstrate that Brd4 controls cell identity gene induction and is essential for adipogenesis and myogenesis. Brd4 co-localizes with lineage-determining transcription factors (LDTFs) on active enhancers during differentiation. LDTFs coordinate with H3K4 mono-methyltransferases MLL3/MLL4 (KMT2C/KMT2D) and H3K27 acetyltransferases CBP/p300 to recruit Brd4 to enhancers activated during differentiation. Brd4 deletion prevents the enrichment of Mediator and RNA polymerase II transcription machinery, but not that of LDTFs, MLL3/MLL4-mediated H3K4me1, and CBP/p300-mediated H3K27ac, on enhancers. Consequently, Brd4 deletion prevents enhancer RNA production, cell identity gene induction and cell differentiation. Interestingly, Brd4 is dispensable for maintaining cell identity genes in differentiated cells. These findings identify Brd4 as an enhancer epigenomic reader that links active enhancers with cell identity gene induction in differentiation.
Similar content being viewed by others
Introduction
Differentiation is controlled by cell-type-specific gene expression, which is under the control of transcriptional enhancers1. Enhancers possess recognition motifs for sequence-specific lineage-determining transcription factors (LDTFs), which bind to and activate enhancers2. LDTFs recruit epigenomic regulators that remodel the chromatin landscape by adding epigenetic modifications (i.e., methylation, acetylation, etc.) to the histone tails of the associated nucleosomes, after which RNA polymerase II (Pol II) is recruited and transcription of enhancer RNA (eRNA) and nearby genes occurs2. Enhancers are enriched with histone H3K4 mono-methylation (H3K4me1), which is mainly deposited by H3K4 methyltransferases MLL3 (KMT2C) and MLL4 (KMT2D)3. H3K4me1 precedes the addition of the active enhancer mark H3K27ac by histone acetyltransferases CBP/p3004,5. CBP/p300 binding identifies active enhancers that control cell-type-specific gene expression6. The acetyl marks on histones act as docking sites for epigenomic readers that have conserved bromodomains7.
The epigenomic reader Brd4 is a member of the bromodomain and extra-terminal domain (BET) family of nuclear proteins that also include Brd2 and Brd38,9. Brd4 is enriched on active enhancers and promoters10,11. Cooperation between LDTFs and CBP/p300 facilitates Brd4 recruitment to its target promoters and enhancers11. Small-molecule competitive BET inhibitors such as JQ1 bind to the acetyl-lysine binding pocket of the BET bromodomains and displaces BET proteins from chromatin12. JQ1-mediated inhibition of Brd4 binding leads to the displacement of the Mediator complex and Pol II from enhancers, which in turn reduces eRNA production and associated gene expression10,13,14.
Adipogenesis and myogenesis are two model systems that are well suited for studying cell differentiation. In adipogenesis, the induction of early adipogenic transcription factors (TFs), including CCAAT/enhancer binding protein-β (C/EBPβ), in turn induces the expression of two master adipogenic TFs, peroxisome proliferator-activated receptor-γ (PPARγ) and C/EBPα15,16. PPARγ and C/EBPα work in cooperation to activate many adipocyte-specific genes. The synchronized nature of adipogenesis in cell culture, wherein the majority of the cells in the confluent population differentiate from preadipocytes to mature adipocytes within 6–8 days, allows for a system conducive to studying gene expression during the process of differentiation17. Myogenesis is another model of synchronized cell differentiation. Myogenic differentiation protein (MyoD) and myogenic factor 5 (Myf5) are required for commitment to the muscle differentiation program, while myogenin plays a necessary role in establishing the terminal muscle phenotype18.
Brd4 is a prominent drug target for cancers but its role in normal cell differentiation and tissue development is largely unexplored. In this study, we use adipogenesis and myogenesis as model systems to explore the role of Brd4 in differentiation and development. Using tissue-specific Brd4 KO mice, we provide in vivo evidence that Brd4 is essential for adipogenesis and myogenesis. During adipogenesis, Brd4 preferentially binds to active enhancers and controls the induction of cell-type-specific genes. Furthermore, we determine the sequential actions of LDTFs, enhancer epigenomic writers MLL3/MLL4 and CBP/p300, Brd4, transcription coactivator complex Mediator and general transcription factor TFIID, Pol II, and transcription elongation factor p-TEFb in enhancer activation and cell identity gene induction during differentiation. We further show that unlike its critical role in cell differentiation, Brd4 is largely dispensable for the maintenance of cell identity gene expression in differentiated cells.
Results
Brd4 is required for adipose tissue and muscle development
Two independently developed Brd4 conditional knockout (KO) mouse strains, Brd4 f/f and Brd4 f/f #2, were used in this study. In the Brd4 f/f strain, the exon 3 was flanked by two loxP sites (Fig. 1a), while in the Brd4 f/f #2 strain, the exon 5 was floxed (Supplementary Fig 1a). To study the role of Brd4 in adipose tissue and muscle development, we crossed Brd4 f/f with Myf5-Cre mice to delete Brd4 gene in progenitor cells of brown adipose tissue (BAT) and muscle lineages19,20. The resulting Brd4 f/f;Myf5-Cre mice survived until birth. E18.5 Brd4 f/f;Myf5-Cre embryos were obtained at the expected Mendelian ratio but were unable to breathe and died immediately (Fig. 1b). They showed an abnormal hunched posture due to severe reduction of back muscles (Fig. 1c). Immunohistochemical analysis of cervical regions of E18.5 embryos revealed that the deletion of Brd4 leads to severe reduction of BAT and muscle mass, indicating that Brd4 is essential for BAT and muscle development (Fig. 1d).
Brd4 controls cell identity gene induction
To investigate how Brd4 regulates adipose tissue development, we isolated primary Brd4 f/f preadipocytes from BAT. After immortalization, cells were infected with adenoviruses expressing green fluorescent protein (GFP) or Cre (Fig. 2a, b and Supplementary Fig 2). Deletion of Brd4 by Cre did not affect the growth rate of SV40T-immortalized brown preadipocytes (Fig. 2c), but prevented adipogenesis and the induction of adipocyte marker genes such as Pparg, Cebpa, and Fabp4 (Fig. 2d, e). We confirmed the essential role of Brd4 in adipogenesis in an independent brown preadipocyte cell line derived from Brd4 f/f #2 mice (Supplementary Fig 1b–d). Consistent with the phenotypes observed in both Brd4 knockout cell lines, knockdown of Brd4 in 3T3-L1 white preadipocytes impaired adipogenesis (Supplementary Fig 3). Knockdown of Brd4 in C2C12 myoblast inhibited myogenesis and myocyte gene expression (Supplementary Fig 4a–d).
To find out how Brd4 regulates gene expression during adipogenesis, we performed RNA-Seq analyses before (day 0, D0) and during (D2) adipogenesis of Brd4 f/f preadipocytes infected with adenoviral GFP or Cre. Using a 2.5-fold cutoff for differential expression, we defined up-regulated (816/16,323), down-regulated (1045/16,323), and unchanged (14,462/16,323) genes from D0 to D2 of differentiation (Fig. 2f). Among the 816 up-regulated genes, 351 were induced in a Brd4-dependent manner. Interestingly, only Brd4-dependent up-regulated genes were strongly associated functionally with fat cell differentiation and lipid metabolism (Fig. 2g). During C2C12 myogenesis, Brd4-dependent up-regulated genes were preferentially associated functionally with muscle development and function (Supplementary Fig 4e, f). These results suggest that Brd4 controls cell identity gene induction during adipogenesis and myogenesis.
Brd4 binds to cell identity genes
Next, we performed ChIP-Seq to map the genomic binding of Brd4 before (D0), during (D2), and after (D7) adipogenesis. To exclude false-positive genomic Brd4 binding sites due to off-target effect of the antibody, ChIP-Seq was also done in Brd4 KO cells. To identify high-confidence Brd4 binding sites, we first removed false-positive signals obtained from Brd4 KO cells and then selected overlapping peaks from two biological replicates. In total, we identified 37,111, 13,676, and 27,351 Brd4 binding regions at D0, D2, and D7, respectively (Fig. 3a). Brd4 protein levels decreased only mildly at D2 (Fig. 3b), but the genomic binding of Brd4 redistributed dramatically from D0 to D2 (Fig. 3c). Around 89% of D0 Brd4 binding regions were lost at D2, although some of the lost regions were re-occupied by Brd4 at D7. We also performed Brd4 ChIP-Seq in C2C12 myocytes. The Brd4 binding regions in adipocytes and C2C12 myocytes were largely non-overlapping (Fig. 3d).
To characterize Brd4-associated genes in different stages and cell types, we assigned each Brd4 binding region to the nearest annotated gene. Gene ontology (GO) analysis revealed that at D0 before differentiation, Brd4 binds to genes associated with general biological functions. However, Brd4 moves to adipogenesis-related genes at D2 and D7 of differentiation (Fig. 3e). In C2C12 myocytes, preferred target genes of Brd4 were those involved in muscle cell differentiation. Accordingly, we observed differentiation-stage- and cell-type-specific genomic binding of Brd4 on Pparg and Myod1 loci, which encode the master adipogenic TF PPARγ and myogenic TF MyoD, respectively (Fig. 3f). Together, these results indicate cell-type- and differentiation-stage-specific genomic binding of Brd4 on cell identity genes.
Brd4 co-localizes with LDTFs on active enhancers
Next, we performed motif analysis of the top 3000 Brd4 binding regions at each time point and in different cell types (Fig. 4a). In preadipocytes (D0), Brd4 binding regions were enriched with motifs of AP-1 family of TFs Jun, Jdp2, and JunD. This is consistent with the previous finding that Brd4 interacts directly with c-Jun21. During (D2) and after (D7) adipogenesis, Brd4 binding regions were enriched with motifs of adipogenic TFs such as C/EBPα, C/EBPβ, and PPARγ as well as ATF4, which was recently identified as a novel TF that promotes adipogenesis22. ChIP-Seq analyses of C/EBPα, C/EBPβ, and PPARγ at D2 of adipogenesis confirmed the genomic co-occupancy of Brd4 with these adipogenic TFs (Fig. 4b, c). Particularly, over 50% of the Brd4 binding regions showed co-occupancy with C/EBPα/β. Consistent with previous reports that Brd4 interacts with C/EBPα/β11,21, we observed a physical interaction between Brd4 and C/EBPβ during adipogenesis (Fig. 4d). In C2C12 myocytes, Brd4 binding regions were enriched with motifs of myogenic TF MyoD and its binding partner TCF3 (Supplementary Fig 5a)23. We also confirmed genomic co-localization of Brd4 with MyoD in C2C12 myocytes (Supplementary Fig 5b, c).
Next, we characterized the genomic features of Brd4 binding regions. Based on histone modifications, four types of regulatory elements were defined as described19: active enhancer, primed enhancer (previously described as silent enhancer), active promoter, and silent promoter (Fig. 5a). Interestingly, Brd4 binding sites were mainly located on active enhancers at D0 and D2 but on active promoters at D7 (Fig. 5b). Consistently, Brd4 co-localized with adipogenic TFs on active enhancers at D2 (Supplementary Fig 6). Adipogenic enhancers were defined as active enhancers that are bound by C/EBPs or PPARγ19. Consistent with the genomic distribution, Brd4 was highly enriched on adipogenic enhancers at D2 but enriched on adipogenic promoters, which associate with adipogenic enhancers, at D7 (Fig. 5c). At D2, Brd4 binding was observed on 43.9%, 41.7%, and 78.8% of C/EBP+PPARγ−, C/EBP−PPARγ+, and C/EBP+PPARγ+ adipogenic enhancers, respectively (Fig. 5d). Together, these results indicate that Brd4 co-localizes with LDTFs on active enhancers during adipogenesis.
LDTFs, MLL3/MLL4, and p300 recruit Brd4 to active enhancers
Next, we investigated the mechanisms that recruit Brd4 to active enhancers during adipogenesis. Notably, 73.8% (6129/8309) of Brd4 binding sites within active enhancers were occupied by enhancer epigenomic writers MLL4 and/or p300, indicating a substantial overlap (Fig. 6a). Among the 8309 Brd4+ active enhancers, substantially more of them were co-occupied by both epigenomic writers (p300 or MLL4) and LDTFs (C/EBPs or PPARγ) than by either alone (Fig. 6b, c).
Since MLL3/MLL4 facilitate CBP/p300 binding to enhancers during cell differentiation24,25 and CBP/p300 facilitate Brd4 binding to active enhancers11, we hypothesized that MLL3/MLL4 are required for Brd4 recruitment to enhancers during differentiation. To test this hypothesis, we performed Brd4 ChIP-Seq in Mll3/Mll4 double KO cells at D2 of adipogenesis. Although deletion of Mll3/Mll4 did not affect Brd4 protein levels (Fig. 6d), it led to the loss of 90.3% of total Brd4 binding sites at D2 (Fig. 6e). Consistently, we observed a marked decrease in Brd4 binding levels on MLL4+ active enhancers in Mll3/Mll4 KO cells (Fig. 6f). Decreased binding of Brd4 in Mll3/Mll4 KO cells could be due to reduced binding of LDTFs to enhancers. Indeed, 60% (3725/6158) of C/EBPβ+ MLL4+ active enhancers showed decreased C/EBPβ binding in Mll3/Mll4 KO cells, which consequently led to decreases in p300 and Brd4 binding (Fig. 6g). However, on the remaining 40% (2433/6158) of C/EBPβ+ MLL4+ active enhancers, deletion of Mll3/Mll4 in preadipocytes did not affect C/EBPβ binding but prevented p300 and Brd4 binding at D2 of adipogenesis. Deletion of Mll3/Mll4 also reduced p300 and Brd4 binding to enhancers on Pparg gene locus during adipogenesis (Fig. 6h). Together, these results suggest that LDTFs and epigenomic writers MLL3/MLL4 and p300 coordinate to recruit Brd4 to active enhancers during adipogenesis.
Brd4 is required for Pol II binding on active enhancers
We next asked how Brd4 regulates cell identity gene induction during adipogenesis. For this purpose, we selected several adipogenic enhancers (e1–e8) on cell identity genes Pparg, Cebpa, and Fabp4, which are bound by Brd4 during (D2) adipogenesis (Fig. 7a). We examined the occupancy of the early adipogenic TF C/EBPβ, MLL4, MLL3/MLL4-mediated H3K4me1, CBP/p300-mediated H3K27ac, Brd4, the MED1 subunit of the Mediator coactivator complex, the TBP subunit of the general transcription factor (GTF) TFIID26, Pol II, and the catalytic subunit CDK9 of the positive transcription elongation factor b (p-TEFb)27 on these enhancers during adipogenesis. We did not observe changes in C/EBPβ and MLL4 binding as well as H3K4me1 and H3K27ac levels on Brd4+ adipogenic enhancers in Brd4 KO cells. However, deletion of Brd4 markedly reduced MED1, TBP, Pol II, and CDK9 binding on Brd4+ adipogenic enhancers but not on enhancers near constitutively active genes Arid1a and Jak1 (n1 and n2) (Fig. 7b). Accordingly, deletion of Brd4 decreased eRNA production from Brd4+ adipogenic enhancers (Fig. 7c). These data suggest a model that sequential binding of LDTFs, epigenomic writers MLL3/MLL4, and CBP/p300 facilitates Brd4 binding on active enhancers, which is required for enhancer binding of Mediator, TFIID, Pol II and p-TEFb, eRNA production, and cell identity gene induction during adipogenesis (Fig. 7d).
Brd4 is largely dispensable for maintaining adipocytes
Terminally differentiated adipocytes express high levels of cell identity genes including master adipogenic TFs Pparg and Cebpa. We asked whether Brd4 is required to maintain adipocyte gene expression. To this end, we crossed Brd4 f/f mice with Adipoq-Cre mice to generate adipocyte-specific Brd4 KO mice28. Deletion of Brd4 was successful in adipose tissues (Fig. 8a, b). However, we did not observe any discernable differences in adipose tissue mass (Fig. 8c, d) or the expression of adipocyte identity genes Pparg, Cebpa, Fabp4, or BAT marker gene Ucp1 (Fig. 8e). RNA-Seq analysis of BAT and epididymal white adipose tissue (eWAT) from Brd4 f/f;Adipoq-Cre mice confirmed that the deletion of Brd4 in adipose tissues does not affect adipocyte and BAT-enriched gene expression (Fig. 8f, g). Our data suggest that while Brd4 is essential for adipose tissue development, it is largely dispensable for the maintenance of adipose tissues and related gene expression.
JQ1 blocks PPARγ-stimulated adipogenesis
Since Brd4 is essential for the induction of PPARγ in the early stage of adipogenesis, we tested whether forced expression of PPARγ could rescue adipogenesis in Brd4 KO cells. Indeed, adipogenesis defects in Brd4 KO cells could be rescued by retroviral vector-mediated expression of ectopic PPARγ (Fig. 9a, b). Interestingly, the expression levels of Brd2 and Brd3 increased in Brd4 KO cells, suggesting a functional redundancy among BET family proteins (Fig. 9b). Inhibiting BET proteins by JQ1 treatment completely blocked PPARγ-stimulated adipogenesis (Fig. 9a, b). These results indicate that Brd4 is the major functional BET protein before the induction of PPARγ and suggest that Brd4 is functionally redundant with Brd2/Brd3 in promoting adipocyte gene expression downstream of PPARγ. Consistent with these results, inhibiting BET proteins by JQ1 treatment inhibited the synthetic PPARγ ligand rosiglitazone (Rosi)-induced expression of PPARγ target genes Cebpa, Adipoq, and Fabp4 in undifferentiated preadipocytes expressing ectopic PPARγ (Fig. 7c), indicating that BET family proteins are required for PPARγ target gene expression.
Discussion
Using two independently developed Brd4 conditional KO mice and derived cell lines, we demonstrate that Brd4 is essential for adipogenesis and myogenesis in culture and in vivo. Using RNA-Seq, we show Brd4 promotes cell-type-specific gene expression during cell differentiation. Using ChIP-Seq, we show Brd4 predominantly binds to active enhancers during adipogenesis but preferentially on active promoters after adipogenesis. Further, enhancer epigenomic writers MLL3/MLL4 are required for Brd4 binding to active enhancers during adipogenesis. Finally, Brd4 facilitates enhancer binding of Mediator, TFIID, Pol II, and p-TEFb, and eRNA transcription during adipogenesis. Our findings identify Brd4 as an enhancer epigenomic reader that connects active enhancers with cell identity gene induction during differentiation. Together with previous findings in the literature, our data suggest a model in which sequential actions of LDTFs, H3K4 mono-methyltransferases MLL3/MLL4, H3K27 acetyltransferases CBP/p300, epigenomic reader Brd4, transcription coactivator Mediator, and Pol II transcription machinery on enhancers control cell identity gene expression during differentiation.
Brd4 inhibitors are promising drug candidates for treating cancers and other diseases29,30. However, few studies have looked at the role of Brd4 in cell differentiation and animal development, except that Brd4 is required for the differentiation of erythroid and osteoblast in cell culture31,32, that inducible knockdown of Brd4 in mice results in skin hyperplasia and loss of cell diversity in intestine33, and that myeloid lineage-specific deletion of Brd4 leads to the compromised innate immune response34. By crossing Brd4 f/f with Myf5-Cre mice, we demonstrate that Brd4 is required for adipose tissue and muscle development in vivo. By knocking down Brd4 in 3T3-L1 white preadipocytes and C2C12 myoblasts and knocking out Brd4 in brown preadipocytes, we confirmed that the essential role of Brd4 in adipogenesis and myogenesis is cell-autonomous. Our finding on Brd4 in adipogenesis is consistent with a previous report that JQ1 inhibits differentiation of mesenchymal C3H10T1/2 cells toward adipocytes in culture35. JQ1 treatment cannot distinguish functional roles of BET proteins in adipogenesis. Our genetic study distinguishes the roles of BET family members in adipogenesis and indicates that Brd4 is the major BET protein controlling PPARγ induction in the early phase of adipogenesis while the functionally redundant Brd2, Brd3, and Brd4 control the induction of PPARγ downstream adipocyte genes.
Brd4 localizes on both active promoters and active enhancers in human and mouse tumor cells10,11. In this study, we show dynamic changes of Brd4 binding regions at various stages of differentiation. In our adipogenesis model system, Brd4 mainly localizes on active enhancers that are associated with cell identity genes induced during differentiation. After differentiation, Brd4 binding on active enhancers largely remains but with reduced binding intensity. Interestingly, Brd4 binds to ~75% of all active promoters in differentiated cells, while only ~8% of all active promoters are occupied by Brd4 during differentiation. These results suggest distinct functions of Brd4 during and after cell differentiation. In differentiating cells, Brd4 localization on active enhancers controls cell identity gene induction necessary for terminal differentiation. The role of Brd4 on active promoters in terminally differentiated cells remains to be understood.
LDTFs bind to cell-type-specific enhancers and recruit H3K4 mono-methyltransferases MLL3/MLL419. MLL3/MLL4 are required for enhancer binding of H3K27 acetyltransferases CBP/p300 and enhancer activation during adipogenesis and embryonic stem cell differentiation24,25. It has been shown that hematopoietic LDTFs and CBP/p300 facilitate Brd4 recruitment to active enhancers in leukemia cells11. Similarly, our data show that adipogenic TFs and p300 cooperatively recruit Brd4 to active enhancers during adipogenesis. Furthermore, we demonstrate that MLL3/MLL4 are critical for Brd4 binding on enhancers. Consistent with our data, a recent study showed that Brd4 is required for binding of Mediator and CDK9 on enhancers including the Myc super-enhancer in mouse acute myeloid leukemia (AML cells)13. Brd4 may regulate Mediator binding to enhancers through its physical interaction with Mediator complex27. On the other hand, since PPARγ physically associates with Mediator complex as well36, it is also possible that decreased Mediator binding in Brd4 KO cells is due to the decreased expression of PPARγ, a direct target of Brd4. Physical and functional association between Brd4 and CDK9, a component of p-TEFb, has been well documented27,37. Our data suggest that Brd4 is a molecular bridge between cell-type-specific enhancers and general transcription machinery. Taken together, we propose a model that adipocyte gene expression is induced by sequential actions of LDTFs C/EBPα, C/EBPβ, and PPARγ, enhancer epigenomic writers MLL3/MLL4 and CBP/p300, enhancer epigenomic reader Brd4, Mediator, and Pol II transcription machinery on enhancers (Fig. 7d).
Methods
Plasmids and antibodies
The lentiviral shRNA plasmids pLKO.1 targeting mouse Brd4 (clone IDs TRCN0000311976, TRCN0000088480, and TRCN0000088481) were purchased from Sigma. Anti-RbBP5 (A300–109A), anti-Brd4 (A301–985A100), and anti-MED1 (A300–793A) were from Bethyl Laboratories. Anti-C/EBPα (sc-61X), anti-C/EBPβ (sc-150X), anti-PPARγ (sc-7196X), and anti-p300 (sc-585X) were from Santa Cruz Biotechnology. Anti-H3K4me1 (ab8895), anti-H3K4me2 (ab7766), and anti-H3K27ac (ab4729) were from Abcam. Anti-Pol II (17-672) was from Millipore. For western blot analysis, all antibodies were diluted to 1 μg ml−1. Uncropped blots are available in the Supplementary Figure 7.
Generation of mouse strains
Two Brd4 conditional KO mouse lines were used in this study. In the first Brd4 conditional KO mouse line (Brd4 f/f), the first coding exon of Brd4 gene, exon 3, is flanked by loxP sites. Genotyping the Brd4 alleles, PCR was done using the following primers: 5′-GCCTAGATCAGTGCCTCCATTG-3′ and 5′-ACTGGAACTACATGGCAGCCTG-3′. PCR amplified 244 bp from the wild-type and 344 bp from the floxed allele. Brd4 f/f mice were crossed with Myf5-Cre (Jackson no. 007893, C57BL/6 J and 129S4/SvJaeSor mixed background) or Adipoq-Cre (C57BL/6 background)28 to generate Brd4 f/f;Myf5-Cre or Brd4 f/f;Adipoq-Cre mice. For characterization of Brd4 f/f;Adipoq-Cre mice, we used six 12-week-old male mice per genotype. Animals were not randomized and the researchers were not blinded during the experiment and when assessing the outcome. No animals were excluded from the analysis.
To establish the second Brd4 conditional KO mouse line (Brd4 f/f #2), the heterozygous conditional Brd4 gene trap mice (Brd4 tm1a(EUCOMM)Wtsi) were obtained from the KOMP2 program at Baylor College of Medicine and crossed with FLP1 mice (Jackson no. 003946) to delete the neomycin cassette. In the resulting Brd4 f/f #2 mice, the exon 5 of Brd4 gene is flanked by loxP sites. Genotyping of the Brd4 alleles was done using the following primers: 5'-GGACATGGTGACAGAGTGG-3' and 5'-TCAAATGAATTCACTAGAACTAC-3'. PCR amplified 168 bp from the wild-type and 284 bp from the floxed allele. Brd4 f/f #2 mice were crossed with Cre-ER (Jackson no. 008463) to generate Brd4 f/f #2;Cre-ER mice.
All mouse experiments were performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals and approved by the Animal Care and Use Committee of NIDDK, NIH.
Histology and immunohistochemistry
E18.5 embryos were isolated by Cesarean section, fixed in 4% paraformaldehyde, dehydrated in a methanol series, and embedded in paraffin for sectioning. Paraffin sections were stained with routine H&E or subjected to immunohistochemistry using anti-Myosin (MF20; Developmental Studies Hybridoma Bank, 1:20 dilution) and anti-UCP1 (ab10983; Abcam, 1:400 dilution) antibodies19.
Cell culture and differentiation assays
Primary brown preadipocytes were isolated from interscapular BAT of newborn Brd4 f/f, Brd4 f/f #2;Cre-ER and immortalized by infecting retroviruses expressing SV40T38. Adipogenesis of immortalized brown preadipocytes was induced with DMEM supplemented with 10% fetal bovine serum (FBS), 0.02 μM insulin, 1 nM T3, 0.5 mM IBMX, 2 μg ml−1 DEX, and 0.125 mM indomethacin for 2 days39. After this period, the culture medium was supplemented with FBS, insulin, and T3 only. 3T3-L1 cells were from Daniel Lane.
C2C12 myoblasts were purchased from ATCC and cultured in growth medium of DMEM containing 15% FBS. Myogenesis was induced by replacing growth medium to DMEM containing 2% horse serum when cells were ~70% confluent. Before changing the medium, cells were washed with plain DMEM twice.
qRT-PCR
Total RNA was extracted using TRIzol (Invitrogen) and reverse transcribed using ProtoScript II first-strand cDNA synthesis kit (NEB), following manufacturer’s instructions. qRT-PCR of Brd4 exon 3 was done using SYBR green primers: forward 5′-CCCAGAGACCTCCAACCCTAA-3′ and reverse 5′-AACTGGTGTTTCCATAGTGTCTTGAG-3′. qRT-PCR of Brd4 exon 5 was done using the primers: forward 5′-TGACATCGTCTTAATGGCAGAAG-3′ and reverse 5′-CCTTTTGCCTGGACTATCATGAT-3′.
ChIP-Seq and RNA-Seq
For ChIP-Seq analysis, formaldehyde was added directly to cell culture medium to a final concentration of 2%. After 10 min of incubation at room temperature, glycine was added to a concentration of 125 mM to quench crosslinking reaction. Approximately 2 × 107 cells were washed with 20 ml cold PBS in culture dish twice and scraped off in 10 ml Farnham lysis buffer (5 mM PIPES, pH 8.0, 85 mM KCl, 0.5% NP-40, supplemented with protease inhibitors), then pelleted by centrifugation at 3000×g for 5 min at 4 °C. Cell pellet was resuspended in 10 ml lysis buffer and pelleted again to remove cytosolic proteins. Resulting nuclear pellet was sonicated in 2 ml TE buffer (10 mM Tris-Cl, pH 8.0, 1 mM EDTA, supplemented with protease inhibitors) to achieve DNA fragments of 200–500 bp. Detergents were added to digested chromatin fractions to make 1× RIPA buffer (10 mM Tris-Cl, pH 7.6, 1 mM EDTA, 0.1% SDS, 0.1% sodium deoxycholate, 1% triton X-100). After centrifugation, supernatant was collected in a new tube. For each ChIP, 8–10 μg antibodies were pre-incubated with 50 μl Dynabeads Protein A (Life Technologies) in 1 ml PBS overnight at 4 °C under gentle rotation. Next day, 1 ml chromatin from 1 × 107 cells was mixed with antibody-beads complex and incubated overnight at 4 °C with gentle rotation. Chromatin immunoprecipitates were washed twice with 1 ml RIPA buffer, twice with 1 ml RIPA containing 300 mM NaCl, twice with 1 ml LiCl buffer (50 mM Tris-Cl, pH 7.5, 250 mM LiCl, 0.5% NP-40, 0.5% sodium deoxycholate), and twice with 1 ml TE buffer. Samples were reverse-crosslinked in 100 μl elution buffer (1% SDS, 0.1 M NaHCO3, and 100 μg proteinase K) overnight at 65 °C. DNA was purified by QIAquick PCR Purification Kit (QIAGEN) and quantified using Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific)19,24.
For RNA-Seq, mRNAs were purified using Dynabeads mRNA Purification Kit (Invitrogen), and then they were used to synthesize double-stranded cDNAs using SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen). Sequencing libraries were constructed using NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB). All ChIP-Seq and RNA-Seq samples were sequenced on the Illumina HiSeq 2500.
Computational analysis
To identify Brd4 binding regions, we used ‘SICER’ method with a window size of 50 bp and a gap size of 50 bp40. To eliminate non-specific binding of Brd4 antibody, we compared Brd4 ChIP-Seq data from Brd4 KO cells with that from two biological replicates of control (Pparγf/f) cells24, and kept only the identified Brd4 binding regions with enrichment level significantly higher in control cells than in the Brd4 KO cells, with an estimated false discovery rate (FDR) of <10−3. Then, we chose only the overlapping Brd4 binding regions from two biological replicates. For Brd4 ChIP-Seq data in Mll3 −/− Mll4 f/f cells, an FDR of <10−10 was used to find the high-confidence ChIP-enriched regions. Other published ChIP-Seq data sets were downloaded (GSE74189, GSE50466, and GSE44824)19,24,41.
For motif analysis of Brd4 binding regions (Fig. 4a), we used SeqPos motif tool in Galaxy Cistrome (http://cistrome.org/ap/root)42. The algorithm used in this motif tool is previously described in detail43. We selected the top 3000 Brd4 binding regions to screen enriched TF motifs at each time point.
Data availability
All data sets described in the paper have been deposited in NCBI Gene Expression Omnibus under accession number GSE99101.
References
Heintzman, N. D. et al. Histone modifications at human enhancers reflect global cell-type-specific gene expression. Nature 459, 108–112 (2009).
Heinz, S., Romanoski, C. E., Benner, C. & Glass, C. K. The selection and function of cell type-specific enhancers. Nat. Rev. Mol. Cell Biol. 16, 144–154 (2015).
Froimchuk, E., Jang, Y. & Ge, K. Histone H3 lysine 4 methyltransferase KMT2D. Gene 627, 337–342 (2017).
Jin, Q. et al. Distinct roles of GCN5/PCAF-mediated H3K9ac and CBP/p300-mediated H3K18/27ac in nuclear receptor transactivation. EMBO J. 30, 249–262 (2011).
Bonn, S. et al. Tissue-specific analysis of chromatin state identifies temporal signatures of enhancer activity during embryonic development. Nat. Genet. 44, 148–156 (2012).
Visel, A. et al. ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature 457, 854–858 (2009).
Zeng, L. & Zhou, M. M. Bromodomain: an acetyl-lysine binding domain. FEBS Lett. 513, 124–128 (2002).
Dey, A., Nishiyama, A., Karpova, T., McNally, J. & Ozato, K. Brd4 marks select genes on mitotic chromatin and directs postmitotic transcription. Mol. Biol. Cell 20, 4899–4909 (2009).
Wu, S. Y. & Chiang, C. M. The double bromodomain-containing chromatin adaptor Brd4 and transcriptional regulation. J. Biol. Chem. 282, 13141–13145 (2007).
Loven, J. et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell 153, 320–334 (2013).
Roe, J. S., Mercan, F., Rivera, K., Pappin, D. J. & Vakoc, C. R. BET bromodomain inhibition suppresses the function of hematopoietic transcription factors in acute myeloid leukemia. Mol. Cell 58, 1028–1039 (2015).
Baud, M. G. et al. Chemical biology. A bump-and-hole approach to engineer controlled selectivity of BET bromodomain chemical probes. Science 346, 638–641 (2014).
Bhagwat, A. S. et al. BET bromodomain inhibition releases the mediator complex from select cis-regulatory elements. Cell Rep. 15, 519–530 (2016).
Kanno, T. et al. BRD4 assists elongation of both coding and enhancer RNAs by interacting with acetylated histones. Nat. Struct. Mol. Biol. 21, 1047–1057 (2014).
Lefterova, M. I., Haakonsson, A. K., Lazar, M. A. & Mandrup, S. PPARgamma and the global map of adipogenesis and beyond. Trends Endocrinol. Metab. 25, 293–302 (2014).
Rosen, E. D. & MacDougald, O. A. Adipocyte differentiation from the inside out. Nat. Rev. Mol. Cell Biol. 7, 885–896 (2006).
Ge, K. Epigenetic regulation of adipogenesis by histone methylation. Biochim. Biophys. Acta 1819, 727–732 (2012).
Berkes, C. A. & Tapscott, S. J. MyoD and the transcriptional control of myogenesis. Semin. Cell Dev. Biol. 16, 585–595 (2005).
Lee, J. E. et al. H3K4 mono- and di-methyltransferase MLL4 is required for enhancer activation during cell differentiation. eLife 2, e01503 (2013).
Seale, P. et al. PRDM16 controls a brown fat/skeletal muscle switch. Nature 454, 961 (2008).
Wu, S. Y., Lee, A. Y., Lai, H. T., Zhang, H. & Chiang, C. M. Phospho switch triggers Brd4 chromatin binding and activator recruitment for gene-specific targeting. Mol. Cell 49, 843–857 (2013).
Cohen, D. M. et al. ATF4 licenses C/EBPbeta activity in human mesenchymal stem cells primed for adipogenesis. eLife 4, e06821 (2015).
Lassar, A. B. et al. Functional activity of myogenic HLH proteins requires hetero-oligomerization with E12/E47-like proteins in vivo. Cell 66, 305–315 (1991).
Lai, B. et al. MLL3/MLL4 are required for CBP/p300 binding on enhancers and super-enhancer formation in brown adipogenesis. Nucleic Acids Res. 45, 6388–6403 (2017).
Wang, C. et al. Enhancer priming by H3K4 methyltransferase MLL4 controls cell fate transition. Proc. Natl Acad. Sci. USA 113, 11871–11876 (2016).
Thomas, M. C. & Chiang, C. M. The general transcription machinery and general cofactors. Crit. Rev. Biochem. Mol. Biol. 41, 105–178 (2006).
Jang, M. K. et al. The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19, 523–534 (2005).
Wang, Z. V., Deng, Y., Wang, Q. A., Sun, K. & Scherer, P. E. Identification and characterization of a promoter cassette conferring adipocyte-specific gene expression. Endocrinology 151, 2933–2939 (2010).
Dawson, M. A. et al. Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529–533 (2011).
Filippakopoulos, P. et al. Selective inhibition of BET bromodomains. Nature 468, 1067–1073 (2010).
Najafova, Z. et al. BRD4 localization to lineage-specific enhancers is associated with a distinct transcription factor repertoire. Nucleic Acids Res. 45, 127–141 (2017).
Stonestrom, A. J. et al. Functions of BET proteins in erythroid gene expression. Blood 125, 2825–2834 (2015).
Bolden, J. E. et al. Inducible in vivo silencing of Brd4 identifies potential toxicities of sustained BET protein inhibition. Cell Rep. 8, 1919–1929 (2014).
Bao, Y. et al. Brd4 modulates the innate immune response through Mnk2-eIF4E pathway-dependent translational control of IkappaBalpha. Proc. Natl Acad. Sci. USA 114, E3993–e4001 (2017).
Hu, Y. J. et al. Transcriptional and post-transcriptional control of adipocyte differentiation by Jumonji domain-containing protein 6. Nucleic Acids Res. 43, 7790–7804 (2015).
Ge, K. et al. Alternative mechanisms by which mediator subunit MED1/TRAP220 regulates peroxisome proliferator-activated receptor {gamma}-stimulated adipogenesis and target gene expression. Mol. Cell. Biol. 28, 1081–1091 (2008).
Yang, Z. et al. Recruitment of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol. Cell 19, 535–545 (2005).
Park, Y. K. et al. Distinct roles of transcription factors KLF4, Krox20, and peroxisome proliferator-activated receptor gamma in adipogenesis. Mol. Cell. Biol. 37, e00554–16 (2017).
Park, Y. K. & Ge, K. Glucocorticoid receptor accelerates, but is dispensable for, adipogenesis. Mol. Cell. Biol. 37, e00260–16 (2017).
Zang, C. et al. A clustering approach for identification of enriched domains from histone modification ChIP-Seq data. Bioinformatics 25, 1952–1958 (2009).
Marinov, G. K., Kundaje, A., Park, P. J. & Wold, B. J. Large-scale quality analysis of published ChIP-seq data. G3 4, 209–223 (2014).
Liu, T. et al. Cistrome: an integrative platform for transcriptional regulation studies. Genome Biol. 12, R83 (2011).
He, H. H. et al. Nucleosome dynamics define transcriptional enhancers. Nat. Genet. 42, 343–347 (2010).
Acknowledgements
We thank John Seavitt and the KOMP2 program at Baylor College of Medicine for Brd4 tm1a(EUCOMM)Wtsi mice, Philipp Scherer for Adipoq-Cre mice, Erin Koh and Jaena Taitague for technical support, and NIDDK Genomics Core for sequencing. This work was supported by the Intramural Research Program of NIDDK, NIH to KG.
Author information
Authors and Affiliations
Contributions
J-E.L. and K.G. conceived and designed the experiments. J-E.L., Y-K.P., S.P., and Y.J. performed the experiments. J-E.L., Y-K.P., S.P., Y.J., N.W., B.L., W.P., and K.G. analyzed the data. A.D. and K.O. generated Brd4 f/f mice. J-E.L., N.W., B.L., and W.P. performed computational analyses. J-E.L., S.P., and K.G. wrote the manuscript. K.G. supervised all the experiments.
Corresponding author
Ethics declarations
Competing interests
The authors declare that they have no competing financial interests.
Additional information
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Lee, JE., Park, YK., Park, S. et al. Brd4 binds to active enhancers to control cell identity gene induction in adipogenesis and myogenesis. Nat Commun 8, 2217 (2017). https://doi.org/10.1038/s41467-017-02403-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-017-02403-5
This article is cited by
-
ARID1A safeguards the canalization of the cell fate decision during osteoclastogenesis
Nature Communications (2024)
-
Znf687 recruits Brd4-Smrt complex to regulate gfi1aa during neutrophil development
Leukemia (2024)
-
A genome-wide CRISPR screen identifies BRD4 as a regulator of cardiomyocyte differentiation
Nature Cardiovascular Research (2024)
-
BRD4 facilitates osteogenic differentiation of human bone marrow mesenchymal stem cells through WNT4/NF-κB pathway
Journal of Orthopaedic Surgery and Research (2023)
-
Connectome and regulatory hubs of CAGE highly active enhancers
Scientific Reports (2023)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.