Abstract DNA METHYLTRANSFERASE 3A (DNMT3A) has emerged over the past ~8 years as one of the most important tumor suppressors in the hematopoietic system, being mutated across most types of human hematologic malignancies, and found in greater than 20% of acute myeloid leukemias (AMLs) as well as acute lymphoid leukemias (ALLs) and lymphomas. Through mechanisms that are not understood, DNMT3A mutations are thought to provide a fertile ground for secondary mutations which drive the frank malignancy. In the previous funding period, we sought to establish and study a reliable tumor model of DNMT3A-associated malignancies using the recognized collaboration between DNMT3A-mutation and the internal tandem duplication (ITD) of FLT3 which results in highly penetrant malignancies of both myeloid and lymphoid types. Here, we will study the very earliest events that represent the transition from clonal hematopoiesis to malignacy. We hypothesize that DNMT3A mutations and NPM1 mutations collaborate effectively by enforcing complementary epigenetic changes that serve to maintain mutated cells in an HSC-like state. We expect that a key effect of this dysregulation is aberrant expression of HOX genes that drives self- renewal. We will dissect the mechanisms through which this occurs here using mouse models, human cell lines, and human primary samples. Our long-term goal is to use insights developed here to enforce differentiation and develop new therapeutic strategies. We will (1) Identify the epigenetic and molecular changes associated with the development of malignancies from Dnmt3a-deficient hematopoietic progenitors. Using mice that have mutant alleles of Dnmt3a-KO and inducible NPM1c, we will examine the concerted changes that occur at the epigenetic and transcriptional levels in pre-malignant stem and progenitor cells. (2) Examine the dependencies of AML with mutated DNMT3A, NPM1, and FLT3-ITD. We hypothesize this common sub-type of AML is dependent on the sustained expression of particular genes such as Hox and Meis1. We will examine this and other potential dependencies using CRISPR KO or targeted DNA methylation. (3) Examine in human DNMT3A-mutated AML cells epigenome remodeling and dependencies. We will validate targets identified in Aims 1 and 2, and explore the value of specific modulators such as nuclear re- localization of NPM1, correction of the DNMT3A-mutant allele, and re-methylation of specific target sites. These studies will reveal the stepwise epigenomic changes that occur due to loss of DNMT3A that lead to AML as well as some of their dependencies. This will lead to an improved understanding of how loss of DNMT3A promotes malignancies, and potentially to new therapeutic strategies due to identification of new targets.