UCSF

Unique among metazoan repressive histone methyltransferases, G9a and GLP, which target histone 3 lysine 9 (H3K9), require dimerization for productive H3K9 mono (me1)- and dimethylation (me2) in vivo. Intriguingly, even though each enzyme can independently methylate H3K9, the predominant active form in vivo is a heterodimer of G9a and GLP. How dimerization influences the central H3K9 methyl binding (“reading”) and deposition (“writing”) activity of G9a and GLP, and why heterodimerization is essential in vivo remains opaque. Here, we examine the H3K9me “reading” and “writing” activities of defined, recombinantly produced homo- and heterodimers of G9a and GLP. We find that both reading and writing are significantly enhanced in the heterodimer. Compared to the homodimers, the heterodimer has higher recognition of H3K9me2, and a striking ~ 10-fold increased kcat for nucleosomal substrates under multiple turnover conditions, which is not evident on histone tail peptide substrates. This however is not encoded by altered nucleosome affinity, which is dominated by the G9a protomer and comparable across the homo- and heterodimer. Our results indicate that heterodimerization may be required to relieve autoinhibition of H3K9me reading and chromatin methylation evident in G9a and GLP homodimers. Relieving this inhibition may be particularly important in early differentiation when large tracts of H3K9me2 are deposited by G9a-GLP, which may require a more active form of the enzyme.