We use adipogenesis as a model system to study the roles of histone methyltransferases and demethylases, and the dynamics of site-specific histone methylation, in regulation of gene expression and cell differentiation. We are interested in methylations on K4, K9, K27 and K36 of histone H3 (H3K4, H3K9, H3K27 and H3K36, respectively). We reported previously that Ezh2 and its H3K27 methyltransferase activity are required for adipogenesis and that Ezh2 constitutively represses Wnt genes to facilitate adipogenesis (Wang L, PNAS 2010). Histone methyltransferase G9a is responsible for H3K9 di-methylation (H3K9me2), an epigenetic mark for gene repression. We reported that G9a represses PPARgamma expression and adipogenesis (Wang L, EMBO J 2013). Together with our reports that H3K4me1/2 methyltransferases MLL3/MLL4 are required for PPARgamma and C/EBPa expression and adipogenesis (Lee JE, eLife 2013; Lai B, Nucleic Acids Res 2017), these findings provide an initial view of epigenetic regulation of adipogenesis, and suggest that histone methylations control expression of positive and negative master regulators of adipogenesis (reviewed in BBA 2012 and Cell & Biosci 2014). Much of our knowledge on adipogenesis comes from cell culture models such as the widely used 3T3L1 cells. Adipogenesis is induced by treating 3T3L1 cells with an adipogenic cocktail, which activates transcription factors (TFs) including glucocorticoid receptor (GR) within minutes and increases the expression of TFs including KLF4 and Krox20 within hours. GR, KLF4 and Krox20 are believed to be important for adipogenesis because knocking down each of them in 3T3L1 cells appears to impair adipogenesis in culture. Using conditional knockout mice and derived preadipocytes, we show surprisingly that endogenous GR, KLF4 and Krox20 are largely dispensable for adipogenesis in culture and brown adipose tissue development in mice. In contrast, the master adipogenic TF PPARgamma is essential (Park Y, MCB 2017a and MCB 2017b). These results challenge the existing model on transcriptional regulation of adipogenesis and highlight the need of studying adipogenesis in vivo, a process that is very poorly understood. By screening histone H3.3 mutants carrying a lysine-to-methionine (K-to-M) mutation at K4, K9, K27 or K36, we have found that site-specific depletion of histone H3K4 and H3K36 methylation by H3.3K4M and H3.3K36M, respectively, strongly inhibits adipogenesis. H3.3K36M inhibits adipogenesis by increasing H3K27me3 to prevent the induction of C/EBPalpha and other targets of the master adipogenic transcription factor PPARgamma. Depleting H3K36 methyltransferase Nsd2, but not Nsd1 or Setd2, phenocopies the effects of H3.3K36M on adipogenesis and PPARgamma target expression. In mice, adipose-selective expression of H3.3K36M leads to profound whitening of brown adipose tissue, severe defects in thermogenesis and lipolysis, and insulin resistance in white adipose tissue (WAT). These mice are resistant to high fat diet-induced WAT expansion and show severe lipodystrophy. Together, these results suggest a critical role of Nsd2-mediated H3K36 methylation in adipose tissue development and function.
Showing the most recent 10 out of 26 publications