The origin of Juvenile Myelomonocytic Leukemia : Insights from developmental hematopoiesis

Abstract

Hematopoiesis proceeds through three developmental phases, each with a unique and indispensable function. The individual roles of these phases in the pathogenesis of blood disorders is unknown. We have adapted murine lineage trace models to identify the relative contributions of embryonic, fetal, and adult hematopoietic phases to the origin of Juvenile Myelomonocytic Leukemia. We hypothesized that the fetal phase would have the most pronounced contribution to the development of JMML, a pediatric myeloproliferative disorder whose disease-initiating somatic mutations occur in utero. Progenitors expressing PTPN11E76K from all three waves were growth hypersensitive to GM-CSF due to hyperactive RAS-ERK signaling. However, fulminant myeloproliferation was only seen in fetal and adult cohorts. We observed equal disease severity in FLT3Cre; PTPN11E76K; ROSA26mTmG and CSF1R-MCM; PTPN11E76K; ROSA26YFP cohorts, which had high and low mutant allele frequencies, respectively. This led to the revelation that all progenitors in the BM niche of mutant animals have equal growth hypersensitivity and RAS-ERK hyperactivation due to non-cell autonomous effects of PTPN11E76K. We further identified that FLT3Cre has hematopoietic-restricted expression, and thereby circumvented morbidity from PTPN11E76K expression in endothelial and stromal cells. This led us to hypothesize that FLT3Cre; KrasG12D; ROSA26mTmG would be the first faithful model of JMML to express this disease-initiating mutation. Indeed, FLT3Cre; KrasG12D mice were born at expected Mendelian ratio and showed normal weight gain to 2 weeks of age. Thereafter, they acquired defining features of JMML including monocytosis, anaemia, thrombocytopenia, and hepatosplenomegaly. All FLT3Cre; KrasG12D mice succumb to a JMML-like disease, which was propagated following transplantation. This is in contrast with CSF1R-MCM; KrasG12D; ROSA26YFP mice, in which low mutant allele frequencies in either fetal or adult HSCs uniformly resulted in T-ALL. Our models reveal previously underappreciated features of JMML including an expansion of dendritic cells and a pronounced defect in T-lymphocyte development. We are the first to demonstrate non-cell autonomous effects of hematopoietic-restricted PTPN11E76K expression. Most importantly, we have shown that both the spatial and the temporal origin of JMML-initiating mutations will affect disease manifestations. Each of our findings suggest novel strategies to treat this intractable disease.