Mechanistic consequences of splicing factor mutations and their use for targeted cancer therapy

Abstract

In 2011 Yoshida et al. identified mutations in RNA splicing factors as some of the most common mutations in patients with myelodysplastic syndrome (MDS). Mutations in splicing machinery result in the dysregulation of hundreds of isoforms, which lead to disease.The dysregulation of these isoforms is consistent and splicing factor specific. The most commonly mutated splicing factors are SF3B1 and SRSF2. While we know that mutations in each have distinct consequences on splicing, the molecular drivers of disease and sequence characteristics which influence mutant promoted alternative splicing remain unknown. To address this, I first show through RNA-seq analysis that co-occurring mutations in SRSF2 and SF3B1 are synthetically lethal, and both converge on the NF-kB signaling pathway dysregulation. I then show that a less common mutation in SF3B1 allows cells to escape synthetic lethality through allele- specific splicing. Finally, I use the splicing patterns specific to SF3B1 we learned in the previous studies to identify introns that respond to the most common SF3B1 mutations across more than 12 different cancer types. I inserted these introns into a synthetic killer gene regulated by splicing and then systematically redesigned the introns for a high throughput screen to identify sequence variants that allow the introns to only be spliced in the presence of an SF3B1 mutation and be used in a therapeutic context.