Inhibitor-based deep mutational scanning of MET kinase

Deep mutational scanning (DMS) of the MET kinase domain in wild-type and METdeltaEx14 intracellular background contributed to the identification of conserved regulatory motifs, interactions involving the juxtamembrane and alpha-C-helix, a critical beta-5 motif, clinically documented cancer mutations, and classification of variants of unknown significance (Estevam et al., 2023). Beyond defining phenotypic landscapes, inhibitor-based DMS studies have elucidated the landscape of ATP-competitive resistance across various kinases such as ERK, CDK4/6, Src, EGFR and others (Brenan et al., 2016; Persky et al., 2020; Chakraborty et al., 2023; An et al., 2023).Here, we further explore the landscape of TKI resistance with the MET kinase domain, utilizing the murine Ba/F3 cell line in a constitutively active, TPR fusion background to again investigate mutational responses in two intracellular domain isoforms: MET and METdeltaEx14 (Estevam et al., 2023). Against a panel of 11 MET inhibitors, we have identified novel resistance mutations and uncovered common MET resistance mechanisms across inhibitor types. Additionally, we adopt Rosace as a growth-based, Bayesian fitness scoring framework, which shows reduced false discovery rates and allows for post-processing normalization of inhibitor treatments (Rao et al., 2023). With our dataset, we have analyzed differential sensitivities to inhibitor pairs and provide a framework for assessing inhibitor efficacy based on mutational sensitivity and likelihood. Lastly, we utilize the DMS data generated for MET as a machine learning training dataset to predict resistance mutations for inhibitors that have not been experimentally screened, to act as a basepoint in the assessment of MET-drugs.