Spliceosome induction is a druggable dependency of RAS-driven senescence and cancer

Abstract

Abstract RAS family proteins, including HRAS, NRAS, and KRAS, are frequently mutated in cancer. Although there has been recent success in designing inhibitors that target oncogenic RAS, they elicit resistance and treating RAS-driven cancer remains difficult. Here, employing a proteomic analysis, we find that multiple spliceosome components are upregulated in the nuclei of cells undergoing RAS-induced senescence. This upregulation depends on RAS signalling and occurs in both senescent preneoplastic and fully transformed cancer cells. Spliceosome components are also highly expressed in preneoplastic and cancerous lesions in human and murine lung, liver, colorectal, and pancreatic cancers. Using siRNA screens, we identify six spliceosome components, including SF3B1 and RBM39, that are essential in cells expressing oncogenic RAS. We find that SF3B1 is required in these cells for maintaining splicing fidelity. By combining transcriptome and splicing analyses with functional screens, we identify the RNA Pol II-associated factor SPT5 as a key mediator of the SF3B1 effects. Importantly, using mouse models of liver cancer, we show that RBM39 and SF3B1 inhibitors are effective in targeting both preneoplastic lesions and aggressive tumours expressing oncogenic RAS. In summary, our study highlights the spliceosome as a promising target for RAS-driven cancers capable of inhibiting both cancer initiation and progression.