Pan-cancer analysis identifies mutations in SUGP1 that recapitulate mutant SF3B1 splicing dysregulation
Menée à l'aide de données génétiques issues du projet "The Cancer Genome Atlas Program" et menée in vitro, cette étude met en évidence, au niveau du gène du facteur d'épissage SUGP1, des mutations pouvant altérer l'épissage de l'ARN messager d'une manière analogue à celles observées dans les cancers présentant une mutation du gène du facteur d'épissage SF3B1
SF3B1 is the most commonly mutated splicing factor in cancers, and SF3B1 mutations result in aberrant 3′ splice site usage during splicing of a subset of introns. Here, we utilized an unbiased computational biology approach to determine whether cancer-associated mutations in any other genes produce the same pattern of aberrant splicing as do SF3B1 mutations. We found that, while rare, the most frequently mutated gene that causes such splicing defects is SUGP1. This is striking because previous biochemical studies indicated that loss of SF3B1 interaction with SUGP1 underlies the effects of SF3B1 mutations on splicing. Our findings thus establish the cancer relevance of the biochemical link between SF3B1 and SUGP1 and also identify SUGP1 as a cancer-associated gene.The gene encoding the core spliceosomal protein SF3B1 is the most frequently mutated gene encoding a splicing factor in a variety of hematologic malignancies and solid tumors. SF3B1 mutations induce use of cryptic 3′ splice sites (3′ss), and these splicing errors contribute to tumorigenesis. However, it is unclear how widespread this type of cryptic 3′ss usage is in cancers and what is the full spectrum of genetic mutations that cause such missplicing. To address this issue, we performed an unbiased pan-cancer analysis to identify genetic alterations that lead to the same aberrant splicing as observed with SF3B1 mutations. This analysis identified multiple mutations in another spliceosomal gene, SUGP1, that correlated with significant usage of cryptic 3′ss known to be utilized in mutant SF3B1 expressing cells. Remarkably, this is consistent with recent biochemical studies that identified a defective interaction between mutant SF3B1 and SUGP1 as the molecular defect responsible for cryptic 3′ss usage. Experimental validation revealed that five different SUGP1 mutations completely or partially recapitulated the 3′ss defects. Our analysis suggests that SUGP1 mutations in cancers can induce missplicing identical or similar to that observed in mutant SF3B1 cancers.