SRSF2 mutations occur in up to 25% of acute myeloid leukemia (AML) and 17% of myelodysplastic syndrome (MDS) cases and are associated with poor prognosis, yet no mutation-directed therapy exists. Here, we aimed to identify therapeutically targetable vulnerabilities in MDS/AML with SRSF2 mutations. Ex vivo drug-sensitivity testing of bone marrow cells from AML patients and healthy donors showed that SRSF2-mutant cells are sensitive to inhibitors of CHK1, and WEE1 DNA damage response (DDR) kinases. To test causality, we engineered isogenic K562 cell line clones expressing SRSF2P95H/L/R mutations. RNA sequencing confirmed splicing aberrations characteristic of MDS/AML in these clones. We found that SRSF2P95H/L/R sensitize leukemia cells to ATR-CHK1-WEE1 inhibition. Bone marrow progenitors from Srsf2P95H and U2AF1S34F knock-in mice showed heightened sensitivity to CHK1 inhibition, corroborating the human data. In contrast, RUNX1 mutations were linked to resistance against CHK1 and WEE1 inhibition in SRSF2-mutant AML samples. Runx1 loss also caused resistance to CHK1 inhibitors in knock-in mouse progenitors harboring Srsf2P95H or U2AF1S34F, indicating that RUNX1 loss is a mechanism of resistance. In conclusion, SRSF2 and U2AF1 mutations are biomarkers of sensitivity to ATR-CHK1 pathway inhibitors, while RUNX1 mutations cause resistance. These biomarkers can support patient stratification in MDS/AML.
Eldfors et al. (Tue,) studied this question.