A chemical method for mapping protein kinase substrate interactions in a fundamentally new way from current approaches is proposed. The method will be applied to identification of the kinases responsible for phosphorylation and inhibition of glycogen synthase kinase 3�, (GSK-3�) an enzyme which is hyperactivated in a number of important human disorders including schizophrenia and bi- polar disorder. These diseases affect approximately 5.7 million American adults or about 2.6 percent of the population age 18 and older. A method for understanding the signaling defects in these diseases will potentially lead to new drugs which are safer and have fewer side effects than current therapies.
Protein phosphorylation is a ubiquitous post-translational modification used to control central functions of life including cell division, cell growth and cell death. There are 518 human protein kinases and thousands of protein phosphorylation sites identified to date in the human proteome. Multiple complementary methods exist for forward mapping of kinase signaling pathways starting with a specific kinase and searching for downstream substrates. No methods exist for determining which kinase is responsible for a specific phosphorylation event. We term this approach mapping kinase pathways in reverse. The absence of a method to address this challenge has created a fundamental gap in our understanding of kinase networks. How many different upstream kinases are capable of phosphorylating a given serine? What degree of crosstalk exists at the level of a single phosphorylation site? In this proposal we focus on the development of a chemical method which will allow for the identification of the kinase or kinases responsible for specific serine or threonine phosphorylation events in mammalian cells. We propose to identify the kinases which phosphorylate Glycogen Synthase Kinase- 32 (GSK-32) at it's inhibitory Ser-9 residue, a subject of much study and debate in multiple biological contexts from metabolism (insulin signaling) to neuropsychiatric disorders. Recent studies have revealed the presence of hypo-phosphorylated GSK-32 in schizophrenic patients and in animal models. Furthermore, several drugs which are used to treat schizophrenia cause an increase in phosphorylation of Ser9 of GSK-32, by unknown kinases. We propose to identify the kinases responsible for phosphorylation of GSK-32 under different cell stimulation conditions through development of a kinase-substrate selective crosslinking reaction capable of covalently crosslinking GSK-32 to its upstream kinases allowing their identification using traditional affinity isolation and mass spectrometry. The crosslinking reaction we propose involves a three-component chemical reaction between a catalytically essential lysine residue conserved in all kinases, a cysteine engineered at Ser-9 of GSK-32, and an ATP analog bearing a functional group, o-pthalaldehyde, capable of chemoselectively reacting with lysine and cysteine moieties. The proposal first focuses on chemical design and optimization of the crosslinking reaction and then its application to identification of the kinases activated by diverse cellular stimuli which cause phosphorylation of Ser- 9 of GSK-32.