Histone lysinemethyltransferases (HMTs), which catalyze the methylation of histone lysines, are an important class of epigenetic regulators. G9a and GLP are a pair of HMTs, which catalyze mono- and dimethylation of histone H3 lysine 9 (H3K9). Great progress has been made in developing small molecules targeting both G9a and GLP. They are widely applied in the studies of different disease models including cancers, AD and Prader-Willi syndrome (PWS). Compare with traditional non-covalent inhibitors, covalent inhibitors offer several potential advantages including sustained duration of action, enhanced capability of targeting shallow binding sites, improved selectivity and solution to the resistant mutation problem.
After analysis of our previously published co-crystal structure of UNC0638 and MS012 in complex with G9a or GLP, we found that both G9a and GLP have an identical cysteine residue in their SET domains. which are suitable for covalent modification. Therefore, we designed and synthesized the first G9a/GLP covalent inhibitor (MS8511). This inhibitor covalently modified both G9a and GLP effectively using the MS-based analysis. Interestingly, we also observed this inhibitor modified G9a faster than GLP. In addition, MS8511 inhibited G9a more effectively than GLP in the SAHH-coupled enzymatic assays (IC₅₀ = 100 nM vs IC₅₀ = 140 nM). Similar trend was observed in the radioactivity-based scintillation proximity biochemical assays. Furthermore, MS8511 was more potent than the previously reported G9/GLP inhibitor UNC0642 (100 nM vs 220 nM) in inhibiting G9a in the SAHH-coupled enzymatic assays and displayed higher binding affinity to G9a than UNC0642 (ITC, 44 nM vs 230 nM). While MS8511 and UNC0642 exhibited similar potency in inhibiting GLP (IC₅₀ = 140 nM vs IC₅₀ = 180 nM ) and similar binding affinity to GLP (ITC_,46 nM vs 62 nM). In cellular assay, MS8511 was slightly more potent than UNC0642 in reducing H3K9 dimethylation. Importantly, MS8511 was selective for G9a and GLP over 21 other PKMTs and PRMTs. Moreover, we also solved the co-crystal structures of G9a and GLP in complex with another covalent inhibitor whose structure is very close to MS8511. These structures confirmed the covalent modification of G9a cys1098 and GLP cys1186. We strongly believe, MS8511 would be a useful chemical tool for further investigating physiological and pathophysiological functions of G9a and GLP.

Jian Jin

Appearances