Mitochondrial dysfunction driven by splicing factor mutations in MDS

Project Summary Myelodysplastic syndrome (MDS) is characterized by clonal expansion of dysplastic hematopoietic cells and reduction in circulating white blood cells, red blood cells, and platelets. Splicing factor mutations are the most common driver mutations in MDS. We have identified a new mechanism of mitochondrial quality control that is essential for cell survival in MDS caused by mutations in the splicing factor SRSF2. Targeting this pathway is lethal to SRSF2 mutant cells, suggesting a therapeutic approach for MDS driven by splicing factor mutations. Dysfunctional mitochondria are targeted for lysosomal degradation through a pathway termed mitophagy, which is activated by the protein kinase PINK1. We have found that mitochondrial dysfunction enhances the splicing, stability, and abundance of PINK1 mRNA to support an increased demand for mitophagy. This mechanism is essential for the survival of cells from MDS patients with the SRSF2P95H/+ mutation, which is associated with high grade MDS and therapy resistance. We show that the SRSF2P95H mutation alters splicing of nuclear-encoded mitochondrial genes and impairs oxidative phosphorylation. This mitochondrial dysfunction then signals in a retrograde manner to the nucleus to enhance PINK1 splicing. Inhibition of the splicing regulator glycogen synthase kinase 3 (GSK-3) impairs PINK1 splicing, reduces mitophagy, and is lethal to SRSF2 mutant cells. Furthermore, MDS and AML patients with SRSF2 mutations in the Cancer Genome Atlas express robustly higher levels of mitophagy genes than patients with wild-type SRSF2, indicating that increased mitophagy is a common feature of SRSF2 mutant MDS. We therefore propose that mitochondrial dysfunction enhances PINK1 splicing to allow increased mitophagy and that SRSF2P95H/+ activates this pathway indirectly by disrupting oxidative phosphorylation. Inhibition of PINK1 splicing with GSK-3 inhibitors is lethal to SRSF2 mutant cells, suggesting a therapeutic approach for MDS driven by recurrent mutations in SRSF2. Our specific aims are to: 1) address genetically how impaired oxidative phosphorylation alters PINK1 splicing using a series of patient-derived mitochondrial DNA mutations that disrupt oxidative phosphorylation, 2) apply long-read RNA-sequencing to identify full length transcripts whose splicing is altered by the SRSF2P95H mutation and, more generally, altered by mitochondrial dysfunction, and 3) explore a new mechanism for mitochondrial surveillance through alternative splicing of PINK1.