FLT3 ^ITD and NUP98 translocations Cooperate to Induce Type I Interferon Signaling That Impedes Hematopoietic Progenitor Differentiation

FLT3-Internal Tandem Duplication (FLT3 ^ITD) mutations occur at similar frequencies in both childhood and adult acute myeloid leukemia (AML) patients, but they cooperate with different initiating mutations at different ages. In childhood AML, FLT3 ^ITD often cooperates with translocations that generate fusion proteins, such as NUP98 translocations (NUP98-t). In adult AML, FLT3 ^ITD often cooperates with mutations that inactivate RUNX1, DNMT3A, TET2 and other genes that regulate self-renewal and myeloid differentiation. This raises the question of whether FLT3 ^ITD induces similar, or distinct, changes in gene expression and cell fate when it arises in the setting of pediatric and adult driver mutations. To answer this question, we evaluated pre-leukemic changes in hematopoiesis and gene expression in neonatal and juvenile mice that carried pediatric-like FLT3 ^ITD/NUP98-HOXD13 and adult-like FLT3 ^ITD/Runx1 ^DEL/+ mutation profiles. Both NUP98-HOXD13 (NHD13) and Runx1 ^DEL/+ severely depleted phenotypic hematopoietic stem cells (HSCs) when coupled with FLT3 ^ITD. However, FLT3 ^ITD/NHD13 cooperated to enhancer ectopic self-renewal capacity in lineage committed hematopoietic progenitor cells (HPCs) whereas FLT3 ^ITD/Runx1 ^DEL/+ did not. FLT3 ^ITD/NHD13 and FLT3 ^ITD/Runx1 ^DEL/+ both induced changes in gene expression and enhancer activation that were not seen with either mutation alone. These cooperative changes were almost entirely non-overlapping. The most notable cooperative change in FLT3 ^ITD/NHD13 HPCs was activation of a type I Interferon target gene signature. By contrast, this signature was actively repressed in FLT3 ^ITD/Runx1 ^DEL/+ HPCs due to mir155 activation. Thus, FLT3 ^ITD/NHD13 and FLT3 ^ITD/Runx1 ^DEL/+ mutations have largely non-overlapping, cooperative effects on hematopoietic cell fate and gene expression.

Interferon signaling sustained the FLT3 ^ITD/NHD13 HPC population by impeding differentiation into more committed myeloid progenitors. When we deleted the type I interferon receptor (Ifnar) on a FLT3 ^ITD/NHD13 background, the HPC population was virtually eliminated, yet the pre-granulocyte monocyte (pGM) and granulocyte-monocyte progenitor (GMP) populations expanded. HPC proliferation and death did not change in the absence of Ifnar. These findings suggest that type I interferon impedes HPC differentiation into pGMs during the neonatal and juvenile stages of life. Of note, ectopic interferon was insufficient to enhancer HPC numbers in setting of a sole NHD13 mutation. Thus, interferon is necessary to sustain FLT3 ^ITD/NHD13 HPCs, but not sufficient, and other signaling pathways downstream of FLT3 ^ITD must cooperate with NHD13 in parallel to drive HPC expansion. To our surprise, Ifnar deletion did not reduce repopulating activity in FLT3 ^ITD/NHD13 bone marrow - as measured by limiting dilution transplants - it simply shifted repopulating activity to more committed pGMs. Our raise two important conceptual points. First, we show that different initiating mutations can elicit completely different transcriptional responses downstream of FLT3 ^ITD. Precision medicine strategies must account for these differences as effective therapies for adult FLT3 ^ITD AML may not yield similar success in the pediatric population. Second, type I interferon shapes pre-leukemic hematopoiesis in the setting of FLT3 ^ITD/NUP98-t mutations. In doing so, the mutant cells may hijack a program that we recently identified as promoting normal perinatal HPC expansion.