Enhancers are gene regulatory elements critical for cell-type-specific gene expression in eukaryotes (Bulger and Groudine, 2011). Lineage-determining or signal-dependent transcription factors (TFs) bind to enhancers and recruit chromatin modifying and remodeling enzymes and the Mediator coactivator complex to initiate RNA polymerase II (Pol II)-mediated transcription at promoters (Roeder, 2005). Recent genome-wide analyses of histone modifications, binding of transcription coactivators p300 and MED1 and lineage-determining TFs, coupled with functional assays of gene regulatory elements, have enabled the identification of enhancer chromatin signatures (Heintzman et al., 2007; Lupien et al., 2008; Wang et al., 2008; Heintzman et al., 2009; Visel et al., 2009; Creyghton et al., 2010; Ghisletti et al., 2010; Heinz et al., 2010; Ernst et al., 2011; Rada-Iglesias et al., 2011). In contrast to active promoters that are marked by H3K4me3, enhancers are marked by H3K4me1 and H3K4me2 (H3K4me1/2) but little H3K4me3, and are often bound by cell-type-specific TFs. Active enhancers are further marked by histone acetyltransferases CBP/p300-mediated H3K27ac (Creyghton et al., 2010; Rada-Iglesias et al., 2011). H3K4me1 on enhancers often precedes H3K27ac and activation of enhancers. A large number of recent publications have validated the predictive power of such enhancer chromatin signatures in identification of novel enhancers critical for cell-type-specific gene expression and cell differentiation (reviewed in Calo and Wysocka, 2013).
In yeast, the Set1 complex is the sole H3K4 methyltransferase. Through the enzymatic subunit SET1, it is responsible for all mono-, di- and tri-methylations on H3K4 (Ruthenburg et al., 2007). In Drosophila, there are three Set1-like H3K4 methyltransferase complexes that use dSet1, Trithorax (Trx) and Trithorax-related (Trr) as the respective enzymatic subunits (Mohan et al., 2011). Mammals possess six Set1-like H3K4 methyltransferase complexes (Figure 1—figure supplement 1A). Based on the sequence homology of enzymatic subunits and the subunit compositions (Cho et al., 2007; Ruthenburg et al., 2007; Vermeulen and Timmers, 2010; Cho et al., 2012), they fall into three sub-groups: MLL1/MLL2 (also known as KMT2A and KMT2B, respectively), MLL3/MLL4 (MLL3 is also known as KMT2C; MLL4 is also known as KMT2D, ALR and sometimes MLL2), and SET1A/SET1B (also known as KMT2E and KMT2F, respectively), which correspond to Drosophila Trx, Trr and dSet1 complexes, respectively. dSet1 is responsible for the bulk of H3K4me3 in Drosophila (Ardehali et al., 2011; Mohan et al., 2011). Consistently, depletion of the unique CFP1 subunit of SET1A/SET1B complexes in mammalian cells markedly decreases global H3K4me3 level, suggesting that SET1A/SET1B are the major H3K4 tri-methyltransferases in mammals (Clouaire et al., 2012). In contrast, knockdown of Trr, the Drosophila homolog of MLL3/MLL4, decreases global H3K4me1 levels, indicating that Trr regulates H3K4me1 in Drosophila (Ardehali et al., 2011; Mohan et al., 2011). However, the histone methyltransferases (HMTs) responsible for H3K4me1/2 on mammalian enhancers remain elusive. Further, the functions of these H3K4 mono-/di-methyltransferases on enhancers and in regulating cell-type-specific gene induction and cell differentiation are unclear. Finally, how these HMTs are recruited to enhancers needs to be clarified (Calo and Wysocka, 2013).
Adipogenesis and myogenesis are robust and synchronized models of cell differentiation. Differentiation of preadipocytes towards adipocytes, that is adipogenesis, is regulated by a network of sequentially expressed adipogenic TFs (Rosen and MacDougald, 2006). Peroxisome Proliferator-Activated Receptor-γ (PPARγ) is considered the master regulator of adipogenesis and controls adipocyte gene expression cooperatively with CCAAT/enhancer-binding protein-α (C/EBPα) (Rosen et al., 2002; Lefterova et al., 2008). The early adipogenic TF C/EBPβ marks a large number of TF ‘hotspots’ before induction of adipogenesis. C/EBPβ not only controls the induction of PPARγ and C/EBPα expression but also acts as a pioneer factor to facilitate the genomic binding of PPARγ, C/EBPα and other adipogenic TFs during adipogenesis (Siersbaek et al., 2011). Adipogenesis in cell culture is synchronized, with the vast majority of cells in the confluent population differentiating into adipocytes within 6-8 days, thus providing a robust model system for studying transcriptional and epigenetic regulation of gene expression during cell differentiation (Ge, 2012). Myogenesis is another robust model system for cell differentiation. Ectopic expression of the myogenic TF MyoD in fibroblasts and preadipocytes is sufficient to induce muscle differentiation program characterized by expression of myogenesis markers such as Myogenin (Myog) and Myosin (Tapscott et al., 1988; Lassar et al., 1991).
Using adipogenesis and myogenesis as model systems, here we show MLL4 is partially redundant with MLL3 and is required for cell differentiation and cell-type-specific gene expression. By ChIP-Seq analyses, we observe cell-type- and differentiation-stage-specific genomic binding of MLL4. MLL4 is mainly localized on enhancers and co-localizes with lineage-determining TFs on active enhancers during differentiation. We demonstrate that MLL4 is partially redundant with MLL3 and is a major H3K4 mono- and di-methyltransferase in mouse and human cells. Furthermore, MLL4 is required for H3K4me1/2, H3K27ac, Mediator and Pol II levels on active enhancers, indicating that MLL4 is required for enhancer activation. Finally, we provide evidence to suggest that lineage-determining TFs recruit and require MLL4 to establish cell-type-specific enhancers.
