A Murine Model Harboring Cooperating DNMT3A and SF3B1 Mutations Phenocopies SF3B1 Driven Myelodysplastic Syndrome

Myelodysplastic syndrome (MDS) is a heterogeneous group of clonal hematopoietic stem cell (HSC) disorders associated with peripheral blood cytopenias, cellular dysplasia, and a risk of leukemic transformation. Genome sequencing studies have demonstrated that MDS develops as a result of the sequential acquisition of somatic mutations in HSCs. Splicing factor 3b subunit 1 (SF3B1) mutations are found in ~25% of MDS cases and are associated with lower risk disease characterized by ineffective erythropoiesis. We previously used a conditional knockin mouse model of the most common SF3B1 mutation, Sf3b1 K700E, to show that point mutations in Sf3b1 are sufficient to cause MDS and aberrant splicing of genes involved in erythroid maturation. DNA methyltransferase three alpha (DNMT3A) is mutated in ~10-15% of patients with MDS, making it one of the most frequently mutated epigenetic regulators in MDS. Prior studies have shown that HSCs with DNMT3A loss of function mutations have increased stem cell self-renewal at the expense of HSC differentiation. SF3B1 and DNMT3A mutations co-occur frequently and arise early in MDS pathogenesis, making them attractive therapeutic targets. However, the mechanisms by which these mutations cooperate to cause HSC dysfunction are not well understood.

Clinical studies found patients with SF3B1/DNMT3A co-mutations have shorter median survival and worse progression free survival than those with just SF3B1 mutations. To further investigate how DNMT3A and SF3B1 mutations disrupt HSC function in vivo, we crossed our Sf3b1 K700E mice with a Dnmt3a R878H conditional KI mouse (the murine equivalent to the human DNMT3A R882H mutation). We found that Sf3b1/Dnmt3a double mutant mice developed a progressive anemia, bone marrow myelodysplasia, and an increase in lineage-, cKit+, Sca1+, CD150+ Long Term Hematopoietic Stem Cells (LT-HSCs); similar to Sf3b1 K700E mice. We have previously shown that Sf3b1 K700E hematopoietic stem and progenitor cells (HSPCs) have a competitive disadvantage in bone marrow transplantation (BMT) assays with peripheral blood chimerism decreasing from 50% at the time of transplant to < 10% as early as 4 weeks post-BMT. Bone marrow chimerism remains low upon serial transplantation. In contrast, Dnmt3a R878H HSPCs have similar peripheral blood chimerism compared to wild type HSPCs (~ 50%) on primary and secondary transplantation, but their bone marrow chimerism is close to 100% after 20 weeks in primary and secondary transplant recipients. Sf3b1/Dnmt3a HSPCs have an intermediate phenotype with low (~ 10%) peripheral blood chimerism but a higher bone marrow chimerism (~ 30%) after secondary transplantation. Similarly, Sf3b1/Dnmt3a HSPCs displayed increased serial replating (up to 4 passages) in methylcellulose colony forming assays compared to 2 passages for Sf3b1 K700E HSPCs. However, Dnmt3a mutant HSPCs were able to be replated for > 6 passages.

To identify factors that contribute to the intermediate behavior of double mutant HSPCs, differential gene expression and alternative splicing were assessed. Transcriptomic analysis showed that Sf3b1/Dnmt3a Lineage-, cKit+, Sca1+ (LSK) cells clustered closer to Sf3b1 LSK cells with shared expression of 74% of upregulated genes and 64% of downregulated genes. Shared canonical pathways, identified using Ingenuity Pathway Analysis (IPA), included oxidative phosphorylation, DNA damage response and cell cycle regulation. In contrast, Dnmt3a and Sf3b1/Dnmt3a LSKs shared expression of approximately 30% of differentially expressed genes and did not have any shared canonical pathways. We identified over 2200 alternative splicing events (FDR < 0.05, inclusion difference > ± 0.15) in Sf3b1 and Sf3b1/Dnmt3a mutant LSK cells compared to wild type LSKs, with the highest number of events in the Sf3b1 single mutant LSKs. We performed IPA on the genes differentially spliced in Sf3b1 singe and double mutant LSKs and identified an enrichment for pathways involved in RNA binding, cell cycle and DNA damage.

Taken together, our in vivo and transcriptomic studies show that Sf3b1 point mutations are a stronger driver of stem cell self-renewal and gene expression than Dnmt3a mutations in Sf3b1/Dnmt3a mutant HSPCs. These findings suggest that agents that interfere with the DNA damage response, cell cycle regulation, or spliceosome function may more effectively target SF3B1 and SF3B1/DNMT3A mutant HSCs in patients with MDS.

No relevant conflicts of interest to declare.