My labs initial effort was to isolate novel transcriptional cofactors for PPARgamma. Using GST-fused PPARgamma ligand binding domain (GST-PPARgLBD) as bait, we pulled down PTIP, a nuclear protein that has been implicated in DNA damage response, from cell nuclear extracts. PTIP functions as a transcription coactivator for PPARgamma in reporter assay. However, no direct interaction was observed between recombinant PTIP and PPARgamma proteins, suggesting that PPARgamma may interact indirectly with PTIP through PTIP-associated proteins. By using proteomic approaches to isolate PTIP-associated proteins, we found that in cells, endogenous PTIP and a novel protein PA1 are both subunits of a Set1-like histone H3K4 methyltransferase complex (i.e. MLL3/MLL4 complex) that contains H3K4 methyltransferases MLL3 and MLL4, and the JmjC domain-containing protein UTX (Cho, Y.-W., et al., J. Biol. Chem., 2007. 282: p. 20395-20406.) Further, we demonstrate that the JmjC domain-containing proteins UTX and JMJD3 are histone H3K27-specific demethylases (Hong, S., et al., PNAS, 2007. 104: p. 18439-18444). Methylation on H3K4 is an activating epigenetic mark while methylation on H3K27 is a repressive one. Based on our finding that H3K4 methyltransferases MLL3/MLL4 physically associate with H3K27 demethylase UTX, we propose that by adding an activating epigenetic mark and removing a repressive one, the MLL3/MLL4 complex may use two distinct histone modifying activities to synergistically activate target gene expression. We will use MEF cell lines derived from MLL3-/- and MLL4-flox/flox mice to investigate how MLL3/MLL4 complex regulates ligand-induced PPARgamma target gene expression.
Showing the most recent 10 out of 23 publications