Engineering oncogenic hotspot mutations on SF3B1 via CRISPR-directed PRECIS mutagenesis.

SF3B1 is the most recurrently mutated RNA splicing gene in cancer; However, the study of its pathogenic role has been hindered by a lack of disease-relevant cell line models. Here, we compared four genome engineering platforms to establish SF3B1 mutant cell lines: CRISPR-Cas9 editing, AAV HDR editing, base editing (ABEmax, ABE8e), and prime editing (PE2, PE3, PE5Max). We showed that prime editing via PE5max achieved the most efficient SF3B1 K700E editing across a wide range of cell lines. We further refined our approach by coupling prime editing with a fluorescent reporter that leverages a SF3B1 mutation-responsive synthetic intron to mark successfully edited cells. By applying this approach, called prime editing coupled intron-assisted selection (PRECIS), we introduced the K700E hotspot mutation into two chronic lymphocytic leukemia (CLL) cell lines, HG-3 and MEC-1. We demonstrated that our PRECIS-engineered cells faithfully recapitulate known mutant SF3B1 phenotypes including altered splicing, copy number variations, and cell growth defect. Moreover, we uncovered that SF3B1 mutation can cause the loss of Y chromosome in CLL. Our results showcase PRECIS as an efficient and generalizable method for engineering genetically faithful SF3B1 mutant models. Our approach provides new insights on the role of SF3B1 mutation in cancer and enables the generation of SF3B1 mutant cell lines in relevant cellular context.