Stat3 Inhibition Is A Synthetic Lethal Vulnerability For Splicing Factor-Mutated Hematopoietic Stem Cells

Hematopoietic stem and progenitor cells (HSPCs) sustain adult hematopoiesis, and their malfunction causes blood disorders including myelodysplastic syndrome (MDS). Mutations in pre-mRNA splicing machinery predominate in MDS, however the mechanism of action remains unclear. We discovered a selective dependency of HSPCs defective in Splicing Factor 3B, subunit 1 (SF3B1), the top mutated splicing factor in MDS. Using loss-of-function sf3b1 mutant zebrafish, we showed that impaired splicing hindered HSPC specification. RNA-sequencing of kdrl:gfp+ endothelial cells from sf3b1 mutants and siblings revealed ∼900 mis-spliced genes including Stat3 pathway components. Stat3 target genes were significantly downregulated in sf3b1 mutants, demonstrating decreased Stat3 function. Constitutively active Stat3 overexpression partially suppressed the sf3b1 mutant HSPC defect. Morpholino-induced stat3 mis-splicing or pharmacological Stat3 inhibition with STATTIC significantly decreased HSPCs in sf3b1 heterozygotes but had no effect in wild-type embryos, demonstrating synthetic lethality. Similarly, we demonstrated that murine Sf3b1+/K700E mutant HSPCs were more susceptible to eradication by STATTIC treatment than wild-type cells both in vivo and ex vivo. Moreover, primary MDS patient cells with SF3B1 mutations produced fewer hematopoietic colonies with STATTIC treatment versus controls, further demonstrating conservation of the SF3B1-STAT3 interaction. Using zebrafish, murine and human models, we also showed this STAT3 synthetic lethal interaction was evident in other MDS-associated splicing factor mutants, but not mutants for the epigenetic regulator TET2. Our findings demonstrate a conserved and selective synthetic lethal interaction between STAT3 function and splicing factor defects that is a novel vulnerability for splicing factor mutant HSPCs with important implications for MDS treatment.