Project Summary The long-term goal of this project is to develop effective, precision therapies directed against the initiating mutations of Acute Myeloid Leukemia (AML). During the current funding period, we performed a series of genomic and epigenomic studies that have clarified the mechanisms that AML initiating mutations use to "reprogram" hematopoietic stem progenitor cells (HSPCs), increasing their fitness for transformation. In the next funding period, we will use primary human AML samples, induced pluripotent stem cells (iPSCs), and genetically engineered mouse models to further evaluate the molecular mechanisms involved in preleukemic reprogramming, and progression to AML. Four well characterized events (that initiate more than half of AML cases) will be studied in detail. We will continue our work with DNMT3A mutations and PML- RARA, and add the study of Core Binding Factor AML fusions (RUNX1-RUNX1T1 and CBFB-MYH11). The "toolkit" for these studies will involve the analysis of preleukemic and fully transformed hematopoietic cells from these models, using bulk DNA and RNA sequencing, whole genome bisulfite sequencing, ATAC-seq, ChIP- seq and/or CUT&RUN to detect the genomic locations of activating and repressive histone marks (and the fusions themselves), and single cell technologies for RNA, DNA, and ATAC-seq. We will also be performing comprehensive proteomic studies to complete "proteogenomic" datasets for these initiating events, including 1) the identification of the hematopoietic proteins that interact with the initiating proteins listed above, and 2) the development of quantitative deep-scale proteomic and phosphoproteomic datasets broadly representative of all AML subtypes. The integrative analysis of these datasets (and their availability to the AML community) should provide important new insights about AML pathogenesis, and suggest mechanistically targeted therapies. In this proposal, we provide one representative example of this process: using novel methods to identify proteins that interact with DNMT3A, we discovered several mutations that disrupt the normal interaction of DNMT3A with an inactive isoform of DNMT3B (DNMT3B3); these mutations destabilize DNMT3A and decrease its activity. Remarkably, we found that we can restore the activity of many mutant DNMT3A proteins by retrovirally overexpressing DNMT3L, a protein that normally interacts with DNMT3A and 3B in embryonic cells to increase their activity. "Addback" of DNMT3L into hematopoietic cells with the Dnmt3aR878H mutation remethylates DNA, and decreases the growth of AML cells initiated by this mutation. Since DNMT3L is epigenetically silenced in nearly all AMLs, a program to identify drugs and genetic strategies to reactivate DNMT3L in AML cells will be developed. We have already found that Romidepsin, an HDAC1 inhibitor, potently induces DNMT3L expression, and a clinical trial designed to evaluate the activity of this drug in DNMT3A mutant AMLs is planned. Additional...