Enhanced tools to model preleukemia to leukemia transformation

Abstract

Preleukemia to leukemia transformation is one of the least well-understood processes in the hematopoiesis field, which is in no small part due to the lack of models that faithfully recapitulate this phenomenon. In this thesis, I provide three novel model systems that help to elucidate crucial factors that contribute to leukemia progression. First, I developed a zebrafish model that expresses human hematopoietic cytokines, which critically preserves the clonal diversity present in the original patient-derived leukemia. These human cytokine fish enhance the self-renewal capacity of transplanted human hematopoietic stem cells and also lead to multi-lineage differentiation. Secondly, I developed a tet2 loss-of-function zebrafish to mimic clonal hematopoiesis. TET2 mutations are seen in older individuals and significantly increases the chance of acquiring leukemia. I observed that tet2-deficient zebrafish had restrictive hematopoiesis resulting in fewer cells across multiple blood lineages. Induction of emergency hematopoiesis in these zebrafish resulted in a proliferative phenotype with increased blasts and progenitors. A key translational finding from the tet2 zebrafish is that growth factor treatment may accelerate leukemia in the subset of patients with acute myeloid leukemia and a TET2 mutation. Finally, I examined the KIT D816V mutation commonly seen in a preleukemic neoplasm called systemic mastocytosis, which can evolve into aggressive mast cell leukemia. Upon examination of the mutant D816V KIT receptor, I discovered that dimerization does not occur as it does in the wild type in the presence of ligand, and downstream signalling is activated independently of dimerization. Phosphoproteome analysis highlighted increased activity of the pro-survival mTOR axis, revealing that mTOR inhibiting agents (rapamycin and its analogues) may have a previously unrecognized therapeutic benefit in preleukemic conditions associated with mutant KIT. These models will provide useful preclinical tools for future candidates and unbiased drug screens to identify prospective drugs and effective drug combinations to improve the outcome of human preleukemic diseases.