Novel therapeutics for SF3B1 mutant cancers which exploit the missplicing of DCAF16

Abstract Oncogenic mutations of the core RNA splicing factor SF3B1 are common in myeloid cancers, chronic lymphocytic leukemia (CLL) and select solid tumors, and act as clear disease drivers. Despite multiple efforts, therapeutic modulation of RNA splicing has not yet demonstrated benefit for SF3B1 mutant cancers in the clinic, and there are no FDA approved disease-modifying therapies for cancers with this mutation. This highlights an urgent unmet medical need for targeted therapies. Here we identified the first known gain-of-function target of oncogenic SF3B1, i.e. DCAF16, and exploit it to selectively target SF3B1 mutant cancers – thus demonstrating a novel therapeutic approach to this oncogene. Mutant SF3B1 drives oncogenesis through a variety of complex mechanisms. Among the best-studied are its impact on the stereotyped missplicing of mRNA coding regions (CDS) in target genes like MAP3K7 and BRD9. Ordinarily, these misspliced gene targets yield loss-of-function, contributing to disease pathogenesis. Here we systematically analyzed the impact of mutant SF3B1 on mRNA untranslated regions (UTRs) o in wildtype (WT) vs. SF3B1 mutant cancers across primary myeloid, lymphoid and solid tumors, as well as isogenic cell lines. This identified numerous novel and highly reproducible missplicing events. Among the most prominent targets with UTR missplicing is DCAF16, a substrate recognition adapter for the DDB1/CUL4 E3 ubiquitin ligase complex and emerging pharmacologic target. RNA sequencing (RNA-seq), 3' end sequencing (3'-seq), and Nanopore long read sequencing (LRS) experiments revealed complex mutation-induced alterations in DCAF16 5' and 3' UTRs. Specifically, mutant SF3B1 mediates exclusion of exons 2 and 3 in the DCAF16 5' UTR to generate CDS isoforms with an alternative 5' UTR. Further, in other isoforms mutant SF3B1 promotes use of an alternative 5' splice site in exon 4, which ordinarily contains the CDS and 3' UTR. As a result, one novel DCAF16 transcript skips the CDS to encode a long noncoding RNA (lncRNA) by inclusion of exon 5 to generate an alternative mRNA 3' end. Remarkably, mutant SF3B1 is reproducibly associated with elevated DCAF16 protein levels in multiple isogenic hematologic and solid tumor SF3B1 mutant cell lines over WT, as well as primary myeloid cancers and CLL specimens vs. WT controls. Further, DCAF16 isoforms skipping exon 3 in the 5‘ UTR, or the DCAF16 associated lncRNA, both of which are promoted by mutant SF3B1, directly cause increased DCAF16 protein levels. These transcript species and the presence of mutant SF3B1 are not associated with changes in DCAF16 mRNA half-life, subcellular localization, expression levels of its CDS, or DCAF16 protein half-life. Conversely, the presence or absence of exon 3 in dual luciferase reporter systems, as well as polysome profiling of WT vs. SF3B1 mutant cells indicate that elements within exon 3 of the DCAF16 5‘ UTR, which are misspliced by mutant SF3B1, may regulate DCAF16 translation efficiency. Together, these studies represent the first time that oncogenic SF3B1 has been shown to increase levels of a target protein in a gain-of-function manner, by altering the composition of a target gene's untranslated regions. Finally, DCAF16 is a novel target for protein degrader therapeutics such as proteolysis targeting chimeras (PROTACs) and molecular glues which co-opt it to mediate the proteosomal degradation of neosubstrates of therapeutic importance. Small molecules that utilize DCAF16 to degrade key cancer protein targets including BRD4, DYRK1A and RPS6KA4 among others demonstrated preferential selectivity for SF3B1 mutant cancer cell lines and primary CLL patient specimens due to increased DCAF16 protein. The preferential sensitivity of SF3B1 mutant cells to DCAF16-based degraders was associated with accentuated loss of protein degradation targets such as BRD4, as well as consistent alterations in downstream targets such as MYC and anti-apoptotic BH3 family members in the case of DCAF16-BRD4 degraders. Comprehensive gene expression profiling confirmed reproducible effects of DCAF16-based degrader compounds. Further, solid tumors bearing SF3B1 mutations also exhibit enhanced sensitivity to DCAF16-based protein degraders over WT. This reveals the therapeutic relevance of mutant SF3B1 dysregulation of mRNA UTRs and uncovers targeting DCAF16 as a novel strategy for their selective treatment across a range of solid and hematologic tumors.