Massively-parallel functional characterization of MAPK1 missense mutants

Genome-directed oncology has the potential to revolutionize patient treatment, but is limited by an abundance of rare, uncharacterized and therapeutically uninformative somatic variants. To accelerate characterization of the “long tail” of rare somatic variants, we quantified the activity and drug responsiveness of virtually all possible (99.84%) missense variants in the Ser/Thr kinase MAPK1/ERK2. We identified recurrent and rare hypermorphic and loss-of-function alleles, revealing that variant activity is uncorrelated with mutational frequency. Somatic ERK2 variants displayed variable responses to RAF-, MEK- and ERK-directed therapies, potentially informing clinical treatment strategies for patients whose tumors harbor these alterations. A subset of recurrent and rare somatic variants co-localized on ERK2 protein-protein interfaces, yet engendered contrasting phenotypes based on their specific sub-domain localization. The approach presented here represents an allele-characterization framework that compliments existing computational efforts and supports current and future somatic variant discovery efforts, advancing the promise of genome-guided treatment strategies. Overall design: Examination of the transcriptional output (mRNA) for a set of ERK2 Alleles under both trametinib drug treatment and DMSO controls. All experimental conditions were done in duplicates and sequenced in quadruplicates. Thus, each biological sample (48 total) was sequenced in 4 different lanes, resulting in total of 192 sequenced samples