U2AF1 is a haplo-essential gene required for hematopoietic cancer cell survival in mice

Somatic mutations in the spliceosome gene U2AF1 are common in patients with myelodysplastic syndromes. U2AF1 mutations that code for the most common amino acid substitutions are always heterozygous, and the retained wild-type allele is expressed, suggesting that mutant hematopoietic cells may require the residual wild-type allele to be viable and cause disease. We show that hematopoiesis and RNA splicing in U2af1 heterozygous knock-out mice was similar to control mice, but that deletion of the wild-type allele in U2AF1 (S34F) heterozygous mutant expressing hematopoietic cells (i.e., hemizygous mutant) was lethal. These results confirm that U2AF1 mutant hematopoietic cells are dependent on the expression of wild-type U2AF1 for survival in vivo and that U2AF1 is a haplo-essential cancer gene. Mutant U2AF1 (S34F) expressing cells were also more sensitive to reduced, but not absent, expression of wild-type U2AF1 than non-mutant cells. Furthermore, mice transplanted with leukemia cells expressing mutant U2AF1 had significantly reduced tumor burden and improved survival after the wild-type U2af1 allele was deleted compared to when it was not deleted. These results suggest that selectively targeting the wild-type U2AF1 allele in heterozygous mutant cells could induce cancer cell death and be a therapeutic strategy for patients harboring U2AF1 mutations. Overall design: Mouse E14.5 fetal liver cells or adult bone marrow were harvested and RNA was then isolated from sorted hematopoietic progenitor cells (Lin- cKit+ Sca1-) using the NucleoSpin RNA Plus XS kit (Macherey-Nagel), per manufacturer's instructions. The RNA-sequencing libraries were prepared using KAPA RNA Hyper Prep Kit with RiboErase, per manufacturer's protocol. After amplification, ~300 bp fragments were sequenced on Illumina NovaSeq 6000, generating 2 x 150bp paired-end reads. The reads were then mapped using HISAT2 (version 2.1.0) against the GRCm38 version of the mouse genome from Ensembl consortium. Differentially expressed genes were identified using DESeq2, and splicing event classification was performed using rMATS (version 4.1.0). For Fetal liver samples due to experimental limitations few samples have potential confounding batch effects that may impact differential gene expression. Therefore, we focused using samples within each batch for splicing analysis only.