DNMT3A is a gene frequently mutated in human acute myeloid leukemia (AML), with DNMT3A R882H as the hot spot. It had been long postulated that DNMT3A mutation should play a key role in AML pathogenesis, so far the main animal models used were Dnmt3a-/- or transplantation of retrovirally transduced bone marrow cells expressing human DNMT3A R882H mutations (BMT). To recapitulate the features of human AML associated with DNMT3A mutation, this study generated a conditional knock-in mouse model to express Dnmt3a R878H mutation (homologous to human DNMT3A R882H) from the endogenous promoter/enhancer. We investigated epigenetic changes, including gene expression profiles, DNA methylation, and chromatin modification as affected by the mutation. We also explored the potential mechanisms that can explain the process by which DNMT3Amutation hierarchically induces abnormal hematopoiesis and the manner by which specific regulators of relevant pathways in murine and human settings can be targeted for potential therapeutic applications.


We performed the single-cell RNA-seq (scRNA-seq) of LSKs and MEPs, RNA-seq and Methylated DNA immunoprecipitation sequencing (MeDIP-seq) of Gr-1 cells and whole exome sequence (WES) of BMs and tails in Dnmt3a R878H conditional knock-in mice.


Approximately 4-6 months after birth with interferon induction, all Dnmt3aR878H/WTMx1-Cre+ knock-in mice developed AML of myelomonocytic subtype, characterized by massive expansion of immature cells and infiltration of bone marrow, spleen and lymph node, along with hyperleukocytosis, thrombocytosis, splenomegaly and lymphadenectasis. The leukemic mice also showed severe diffuse skin ulceration and alopecia. The transcriptome and DNA methylation profiling of bulk Gr-1 leukemic cells and the single-cell RNA-sequencing of LSKs/MEPs revealed significant changes in gene expression and epigenetic regulatory patterns that could cause differentiation arrest and growth advantage. Consistent with leukemic cell accumulation in G2/M phase, CDK1 was found overexpressed as a result of mTOR gene activation due to DNA hypomethylation in the gene body region. We then discovered that overexpressed CDK1 could compete with EZH2 in binding to DNMT3A, induce EZH2 phosphorylation and reduction, and result in abnormal histone methylation. Notably, we showed a very significant response from Dnmt3aR878H/WTto the therapeutic effect of the mTOR inhibitor rapamycin, particularly in terms of prolongation of lifespan in treatment group as compared to the control group (p<0.001). Moreover, rapamycin exerted strong inhibitory effects, including anti-proliferative and apoptosis-induction ones, on human AML cells lines and primary samples from AML patients harboring DNMT3A mutation.


We established a novel mouse model for the expression of mutant Dnmt3a R878H from endogenous locus to investigate the role of Dnmt3a abnormality in leukemogenesis. Indeed, Dnmt3aR878H/WTMx1-Cre+ mice developed AML of myelomonocytic subtype with skin injury. We discovered unique gene expression and DNA methylation patterns in concordance with enhanced proliferation and suppressed differentiation in leukemic cells. The heterogeneity of gene expression in individual leukemic stem/progenitor cells implied the presence of clonal diversity, which could underlie disease evolution. The activation of mTOR and the resultant overexpression of CDK1 might contribute to malignant transformation. Evidence has been obtained in both murine and human settings to suggest DNMT3A mutation-related AML as a potential disease target for rapamycin.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.