Mutation in DNA Methyltransferase DNMT3A Confers Enhanced Self-Renewal Capacity Onto Multipotent Progenitor Cells and Predisposes to Acute Myeloid Leukemia (AML)

The clinical significance of clonal hematopoiesis (CHIP or ARCH) remains a barrier to predicting the risk of hematologic malignancy. DNMT3A is a de novo DNA methyltransferase frequently mutated in clonal hematopoiesis, myelodysplastic syndrome and acute myeloid leukemia. Loss of or mutation in DNMT3A has been demonstrated to enhance self-renewal of hematopoietic stem cells (HSCs), suggesting that this is the predominant cell population driving clonal hematopoiesis. How DNMT3A-mutant cells become at risk for transformation is unclear, in part due to our limited understanding of how DNMT3A mutation confers a selective advantage and the cooperating mechanisms required for progression to MDS or AML.

To address this gap in knowledge, we generated a cre-inducible Dnmt3a-LSL-R878H mouse model (representing the DNMT3A-R882H mutation commonly found in human AML), in which wild-type Dnmt3a expression is preserved prior to recombination. Heterozygous Dnmt3aR878H mice exhibit an expansion in both HSCs and multipotent progenitor (MPP) cell subsets with distinct kinetics. Transcriptional profiling of sorted HSC and MPP populations by RNA-seq revealed distinct transcriptional signatures indicating that different mechanisms underlie expansion of Dnmt3aR878H/+ HSCs and MPPs. Dnmt3a-mutant HSCs exhibit downregulation of genes important for differentiation, while Dnmt3a-mutant MPPs exhibit upregulation of genes associated with stem cell self-renewal, including Jam2 and Ryk. Functionally, we observe that Dnmt3a-mutant MPPs have enhanced serial replating capacity in in vitro colony assays. These data suggest that mutation in DNMT3A may cause clonal hematopoietic expansion through distinct mechanisms dependent on the cell-of-origin which incurs this mutation.

To determine whether clonal hematopoiesis driven by Dnmt3aR878H/+ was sufficient to predispose to a hematologic malignancy, we generated an independent, Flp-inducible Npm1-FSF-cA mouse model (representing the NPM1cA mutation commonly found in human AML), in which wild-type Npm1 expression is preserved prior to recombination. Inducing Npm1cA mutation in hematopoietic stem and progenitor cells carrying Dnmt3aR878H caused development of a fully penetrant myeloproliferative disorder upon transplant into recipient mice. Transplantation of these cells into secondary recipient mice led to a fully penetrant AML with accelerated disease kinetics compared to primary transplant recipients. These data suggest that the combination of DNMT3A mutation followed by NPM1 mutation is sufficient to cause AML.

In summary, this study reveals a novel cell context-specificity of how DNMT3A mutation confers a selective advantage and demonstrates that NPM1 mutation can cooperate with DNMT3A mutation to cause AML. This work has implications for predicting individuals at risk of progression from clonal hematopoiesis to AML.