RESUMEN
Achieving selectivity across the human kinome is a major hurdle in kinase inhibitor drug discovery. Assays using active, phosphorylated protein kinases bias hits toward poorly selective inhibitors that bind within the highly conserved adenosine triphosphate (ATP) pocket. Targeting inactive (vs active) kinase conformations offers advantages in achieving selectivity because of their more diversified structures. Kinase cascade assays are typically initiated with target kinases in their unphosphorylated inactive forms, which are activated during the assays. Therefore, these assays are capable of identifying inhibitors that preferentially bind to the unphosphorylated form of the enzyme in addition to those that bind to the active form. We applied this cascade assay to the emerging cancer immunotherapy target hematopoietic progenitor kinase 1 (HPK1), a serine/threonine kinase that negatively regulates T cell receptor signaling. Using this approach, we discovered an allosteric, inactive conformation-selective triazolopyrimidinone HPK1 inhibitor, compound 1. Compound 1 binds to unphosphorylated HPK1 >24-fold more potently than active HPK1, is not competitive with ATP, and is highly selective against kinases critical for T cell signaling. Furthermore, compound 1 does not bind to the isolated HPK1 kinase domain alone but requires other domains. Together, these data indicate that 1 is an allosteric HPK1 inhibitor that attenuates kinase autophosphorylation by binding to a pocket consisting of residues within and outside of the kinase domain. Our study demonstrates that cascade assays can lead to the discovery of highly selective kinase inhibitors. The triazolopyrimidinone described in this study may represent a privileged chemical scaffold for further development of potent and selective HPK1 inhibitors.