Temporal and Spatially Regulated Oncogenic KIT Expression and Loss of Dnmt3a Cooperate to Drive MPN Development: Role of PI3Kinase in Dnmt3a Loss Induced Hyperproliferation in Myeloid Cells

Abstract

Recent studies, including studies from our group have shown that somatic mutations in DNA methyltransferase type 3A (DNMT3A) are frequently associated with systemic mastocytosis (SM), myelodysplastic syndrome and AML. Importantly, the presence of this mutation is associated with poor prognosis and overall patient survival. We have shown that approximately 12% of patients with SM possess DNMT3A mutations, which are commonly co-expressed with an activating mutation of KIT (KITD816V in humans and KITD814V in mice). However, it is unclear how DNMT3A mutations contribute to myeloid lineage derived mast cell growth and differentiation either alone or in conjunction with KITD816V. Further, the mechanisms by which loss of DNMT3A impairs various mast cell functions either alone or in cooperation with KITD816V are poorly understood. Utilizing mice conditionally deficient in Dnmt3a in the hematopoietic compartment, we show that loss of Dnmt3a in vivo results in increased numbers of mast cells relative to controls. In vitro, loss of Dnmt3a results in accelerated and enhanced differentiation of mast cells from its bone marrow (BM) precursors. Previous studies have shown that the level of expression of the transcription factor, Microthalamia (MITF) (high MITF expression drives mast cell commitment) and CEBPα (high C/EBPα drives basophil commitment) in Basophil-Mast cell common progenitor (BMPs) drives the fate of these cells towards either basophilic lineage or mast cell lineage. To assess why loss of Dnmt3a results in increased mast cell lineage commitment, we performed RNAseq analysis and found elevated expression of MITF and a profound repression in the expression of C/EBPα in BM precursors. Importantly, restoring the expression of Dnmt3a in Dnmt3a deficient BM cells, completely restored enhanced differentiation, along with the expression of MITF and C/EBPα. To our surprise, no appreciable differences in the expression of Gata-1, Gata-2 or PU.1 were noted in the absence of Dnmt3a relative to controls. These results demonstrate that early loss of C/EBPα in Dnmt3a null cells contributes to enhanced maturation of mast cells from their precursors. We next assessed the growth and survival potential of Dnmt3a mast cells and found a significant increase in SCF-mediated growth of these cells compared to controls. Biochemical analysis revealed greater PI3Kinase activation in Dnmt3a null cells, and these findings were supported by genome wide transcriptome analysis involving Ingenuity pathway analysis (IPA) and Gene set enrichment analysis (GSEA). Importantly, mast cells derived from Dnmt3a null BM lacking the expression of p85α regulatory subunit of PI3Kinase (Dnmt3a-/-:p85α-/-) or pharmacologic inhibition of PI3Kinase completely corrected SCF induced hyper proliferation in these cells to near normal levels. The increase in PI3Kinase activity in Dnmt3a null cells was associated with the loss of PTEN expression. To understand how loss of Dnmt3a and expression of an oncogenic form of KIT cooperate to drive MPN, we generated mice in which the expression of oncogenic KIT was temporally and spatially regulated in a cell type and tissue specific manner using the Mx-Cre system along with the expression of Dnmt3a. Consistent with earlier studies, we observed that KitD814V/+ mutation alone only resulted in the development of lymphoid neoplasms. In contrast, Dnmt3a-/- :KitD814V/+ mice manifested robust signs of MPN development including splenomegaly with increased numbers of neutrophils and monocytes in the peripheral blood associated with a significant reduction in lymphoid cells including B cells. Importantly, we noted a 5-fold increase in the fraction of LSKs and GMPs in the BM of Dnmt3a-/- :KitD814V/+ mice compared to controls and consequently an increase in the presence of mature Gr-1 and Mac-1 positive myeloid cells in the BM, spleen and in the PB of these mice relative to controls. The observed MPN was more aggressive in Dnmt3a-/- :KitD814V/+ mice compared to Dnmt3a+/- :KitD814V/+ mice and of stem cell origin as transplantation of BM cells from these mice recapitulated the development of MPN in recipient mice, similar to that observed in the original animals. The impact of PI3Kinase inhibition on mitigating oncogenic KIT and loss of Dnmt3a induced MPN and overall survival will be discussed. DisclosuresNo relevant conflicts of interest to declare.