Combined Proteomic and Genetic Interaction Mapping Reveals New RAS Effector Pathways and Susceptibilities

Abstract

Activating mutations in RAS GTPases drive one fifth of cancers, but poor understanding of many RAS effectors and regulators, and of the roles of their different paralogs, continues to impede drug development. We developed a multi-stage discovery and screening process to understand RAS function and identify RAS-related susceptibilities in lung adenocarcinoma. Using affinity purification mass spectrometry (AP/MS), we generated a protein-protein interaction map of the RAS pathway containing thousands of interactions. From this network we constructed a CRISPR dual knockout library targeting 119 RAS-related genes that we screened for genetic interactions (GIs). We found important new effectors of RAS-driven cellular functions, RADIL and the GEF RIN1, and over 250 synthetic lethal GIs, including a potent KRAS-dependent interaction between RAP1GDS1 and RHOA. Many GIs link specific paralogs within and between gene families. These findings illustrate the power of the multiomic approach to identify synthetic lethal combinations for hitherto undruggable cancers. STATEMENT OF SIGNIFICANCEWe present a thorough survey of protein-protein and genetic interactions in the Ras pathway. These interactions suggested new discoveries that we validate here, and demonstrate important new paralog specificities and redundancies. By comparing synthetic lethal interactions across KRAS-dependent and -independent tumors, we identify new combination therapy targets against Ras-driven cancers.