Abstract 2406 Inhibition of Aurora B kinase autoactivation by lysine methylation of the activation loop

Most protein kinases are capable of autoactivating by phosphorylating their activation loops, but their activities in cells are tightly constrained. The factors responsible for fine-tuning kinase activity are clinically significant because many human diseases are driven by kinase hypo- or hyperactivation. The phosphatase-mediated inactivation of an already active kinase by dephosphorylation of the activation loop is well understood, but significantly less attention has been paid to a complementary and fundamentally distinct mode of regulation, in which an inactive kinase is outright prevented from undergoing autoactivation. This kind of regulation has been shown to modulate the activities of several kinases in cells, and it often proceeds by modification of the activation loop, rendering it a poorer substrate for the kinase. Here, we document a new example of such regulation for the mitotic kinase Aurora B, demonstrating that its autoactivation can be blocked by methylation of a single lysine residue within its activation loop. We show, using in vitro biochemistry with purified components, that Aurora B kinase is a substrate for the lysine methyltransferases (KMTs) G9a and SETD6, which are recently established regulators of related mitotic kinase Plk1. Incubation of initially inactive Aurora B with either KMT yields a methylated kinase, which, upon addition of ATP, fails to phosphorylate its own activation loop as well as other substrates. We find that the activity of Aurora B in synchronized, mitotic human RPE-1 cells is increased by the addition of inhibitors against the same KMTs. We provide evidence that this regulation is conferred by a single lysine residue within the activation loop of Aurora B, because purified Aurora B harboring an arginine substitution at the position of the critical lysine is not inhibited by KMTs. We conclude that the activation loop of Aurora B kinase contains an unexpected site for negative regulation of autoactivation, allowing fine-tuning of Aurora B activity during mitosis. We speculate that similar mechanisms may be important for the regulation of other kinases. Our findings strengthen our emerging understanding of how KMTs modulate kinase signaling during mitosis. We anticipate that continued investigation of how other kinases might be similarly regulated will broaden our understanding of complex cellular signaling networks and potentially reveal new therapeutic targets. Luke Eldredge was supported by 5T32GM008136-35. Todd Stukenberg was supported by 5R35GM148181-02.