SF3B1 Mutation Drives Loss of Y Chromosome in Chronic Lymphocytic Leukemia

RNA splicing factor SF3B1 is the most recurrently mutated gene in chronic lymphocytic leukemia (CLL), a mature B cell malignancy. Mutations on SF3B1 are associated with poorer clinical outcomes in patients and accelerated onset of CLL in murine models, highlighting an oncogenic role for mutant SF3B1 in leukemogenesis. While typically indolent, CLL can progress to an aggressive state characterized by increased copy number variations (CNVs) and genetic lesions. We previously reported that SF3B1 mutations not only impacted RNA splicing but also contributed to chromosomal instability (Yin et al, Cancer Cell 2019; Cusan et al, JCI 2023). These findings support a model in which SF3B1 mutations can potentiate downstream cytogenetic changes to drive the clonal evolution of CLL. However, to date, little evidence exists to support this model.

To study this, we first developed a novel methodology called prime editing coupled intron-assisted selection (PRECIS) that pair prime editing with a SF3B1 mutation-responsive intron reporter to engineer SF3B1 mutant cell lines. Our PRECIS workflow leveraged the PE5max system to introduce the K700E hotspot mutation into two CLL cell lines: HG-3 and MEC-1.Through a synthetic intron reporter that fluorescently labeled SF3B1 mutant cells with GFP, we were able to derive pure mutant clones following FACS enrichment and single-cell cloning, achieving cancer-relevant allelic frequencies of 50% and 33% in HG-3 and MEC-1, respectively. Through RNA-seq, we showed that the RNA splicing profiles in our CLL SF3B1 mutant cell lines were consistent with those of SF3B1-mutated primary CLL. We further confirmed additional SF3B1 mutation phenotypes including G₁-S phase cell cycle arrest, impacted cellular growth, and loss of CD5 expression.

To study the effect of SF3B1 mutation on CNVs, we conducted whole genome sequencing (WGS) using genomic DNA derived from these cell lines. When comparing SF3B1 mutant to WT cells, WGS analyses revealed widespread CNV events including both focal and arm level deletions and amplifications. The most significant CNV event that we observed was the loss of the Y chromosome (LOY) in SF3B1 mutant cells. We confirmed our observation using RNA-seq calls and differential gene expression (DEG) analysis that showed global downregulation of Y chromosome genes such as DDX3Y, KDM5D, UTY, etc. Furthermore, leveraging genomic DNA samples retained at each step of the PRECIS engineering process, we profiled and mapped out the clonal evolutionary trajectory of a few SF3B1 mutant clones, identifying the first LOY event – a focal deletion of KDM5D – at 35 days after editing and a complete deletion of the Y chromosome at day 90. Our results support LOY as a general SF3B1 mutation phenotype.

We next focused on validating the linkage between SF3B1 mutation and LOY in primary CLL patient samples. Using RNA-seq data (n=395) derived from CLLmap (https://cllmap.org/), we created a Y chromosome gene signature score based on the 43 most expressed Y genes to measure LOY in male CLL patients. Using this, we observed increased levels of LOY in SF3B1 mutant samples (n=84) compared to WT samples (n=311). Known tumor suppressors such as DDX3Y and KDM5D were among the most significantly affected genes, suggesting a potential impact of LOY on patient outcomes. To determine this, we utilized the Y signature score to stratify patients into Y+ and Y- groups based on high (top 30%) and low (bottom 30%) Y chromosome expression. Compared to Y+ patients (n=116), Y- patients (n=124) had a shorter overall survival (median survival: 5701 vs 3254 days, p=0.017). Compared to WT samples (n=84), SF3B1 mutation (n=40) further accelerated disease progression in Y- patients (median survival: 3902 vs 2348 days, p=0.042). Moreover, LOY was associated with an overall poorer clinical outcome (HR=1.75, 95% CI: 1.10-2.79, p=0.019, Cox univariate test). These results suggest that SF3B1 mutation not only drives LOY but also synergizes with LOY to accelerate CLL aggressiveness.

In summary, through our innovative PRECIS gene editing strategy, we were able to study the effect of SF3B1 mutation on CLL biology, confirming previously known phenotypes such as aberrant splicing while uncovering novel genetic events such as LOY. We were able to model for the clonal evolutionary trajectory of SF3B1 mutation, discovering LOY as a novel SF3B1 mutation phenotype and providing evidence for LOY as a contributor to aggressive CLL.