%0 Journal Article %A Dennis Liang Fei %A Hayley Motowski %A Rakesh Chatrikhi %A Sameer Prasad %A Jovian Yu %A Shaojian Gao %A Clara Kielkopf %A Robert K. Bradley %A Harold Varmus %T Wild-type Splicing Factor U2AF1 inhibits splicing associated with a recurrent U2AF1 mutant in human lung cancers and is required for cell survival %D 2016 %R 10.1101/048553 %J bioRxiv %P 048553 %X The splicing factor gene, U2AF1, is recurrently mutated in a variety of human cancers, including lung adenocarcinomas. The most frequent U2AF1 mutant, U2AF1 p.Ser34Phe (S34F), induces specific changes in splicing that we collectively refer to as “S34F-associated splicing”, but it is unclear how these splicing changes are regulated. Moreover, while a wild-type U2AF1 allele is retained in all cancers expressing a U2AF1 mutation, the functional significance of the retained wild-type allele is unknown.Our analysis of published data on human lung adenocarcinomas indicates that lung adenocarcinomas carrying a U2AF1 S34F allele exhibit a wide range of mutant to wild-type U2AF1 (S34F:WT) mRNA ratios, which can be partially attributed to copy number variation at the U2AF1 locus. S34F:WT mRNA ratios, rather than absolute levels of U2AF1 S34F or total U2AF1 mRNA, correlate positively with the magnitude of S34F-associated splicing in lung adenocarcinoma transcriptomes. To examine the effect of S34F:WT ratios on S34F-associated splicing directly, we modified a human bronchial epithelial cell line so that U2AF1 S34F is expressed at one of the two endogenous U2AF1 loci and the S34F:WT mRNA ratio approximates one. By altering the levels of mutant or wild-type U2AF1 in this engineered cell line, we show that the degree of S34F-associated splicing is proportional to the ratio of S34F:WT gene products and not to absolute levels of either the mutant or wild-type factor. Further, we show that in nearly all cases, S34F-associated splicing alterations are largely explained by the different RNA binding affinities of recombinant protein complexes containing wild-type or mutant U2AF1. Together, these observations suggest that wild-type U2AF1 is a negative regulator of S34F-associated splicing, at least in part through differential binding to 3′ splice sites. Finally, we show that the U2AF1 S34F allele does not behave like some oncogenes: the mutated gene does not induce cell transformation, and lung adenocarcinoma cell lines are not dependent on it for growth in vitro or in vivo. Wild-type U2AF1, however, is absolutely required for cell survival, regardless of whether the cells carry the U2AF1 S34F allele.We conclude that wild-type U2AF1 has two important functions in U2AF1 S34F-expressing lung cancers: it controls the magnitude of S34F-associated splicing, and it is essential for cell survival. %U https://www.biorxiv.org/content/biorxiv/early/2016/04/13/048553.full.pdf