SUMMARY Acute myeloid leukemia (AML) is the most common acute leukemia in the US, with ~20,000 new diagnoses and ~11,000 deaths per year. This pernicious hematological malignancy has rapid progression and startling mortality in untreated patients. This is particularly true for elderly patients (i.e., > 65 years of age), where up to 70% of newly diagnosed patients succumb within a year. These patients have worse prognoses, often cannot tolerate the most aggressive (and effective) treatment courses, are less likely to obtain complete remission af- ter standard induction and consolidation treatment, and are more likely to have serious complications and side effects. Using bioinformatic and bench approaches, we found the leukemic cells show striking sensitivity to mi- tochondrial toxins. Our subsequent research identified a correlation between lower coupling efficiency in the mitochondria of AML cells and increased sensitivity to mitochondrial damage. We showed strong synergy be- tween multiple combinations of mitochondria-targeting molecules and known anti-cancer agents in AML, but not healthy cells, selectively killing cancer cells. We hypothesized that hematological cancers accrue substan- tial mitochondrial damage but reduce activation of mitochondrial recycling (mitophagy), likely in an effort to maintain their cell division potential. However, over-activating mitophagy causes a crisis that triggers pro- grammed cell death pathways. To leverage these findings, we performed a high-throughput screen and identi- fied 8 compounds that increased levels of mitophagic activator kinase PINK1, driving mitochondrial turnover. Iterative structure-activity relationship studies yielded two analogs, named PS127B and PS127E, with CC50AML in the nM range; healthy cells survived the drug at up to 20 times these concentrations. These molecules inhib- it mitochondrial function, reduce oxygen consumption and ATP production, and activate several programmed cell death pathways. These compounds also synergized with known AML chemotherapeutics, were effective against primary AML cells (including leukemia stem cells), and reduced tumor burden and extended survival in mice engrafted with human leukemia cells. A subsequent in silico screen of over 4 M compounds using two predicted activities (apoptotic agonist and thioredoxin glutathione reductase inhibition) shared by multiple cyto- toxic PS127-family members (but missing for non-cytotoxic members) identified 213 hits. 93 of these hits were in a cluster of compounds with chemical similarity to the PS127 family. 23 others were in a second group, structurally related to each other, but not to PS127. In this project, we will validate cheminformatics predictions of these comounds’ molecular targets, determine their effects on cellular metabolism, and confirm that they trigger mitophagic activation via the PINK1/Parkin axis in mammalian cells. Completion of the proposed exper- iments will identify and characterize high-value leads for future pre-clinical studies of in murine models engraft- ed with patient-derived leukemia cells, including PK/PD and treatment efficacy. This work will also validate tar- geting mitochondria for autophagic degradation as a therapeutic strategy for AML.

NARRATIVE Acute myeloid leukemia (AML), a hematological malignancy, remains stubbornly difficult to treat in both children and adults, with fewer than 30% of patients surviving five years after diagnosis. This project tests a novel treatment target (mitochondrial recycling—aka mitophagy) and novel small molecules identified using cheminformatics, in AML cell lines and primary cells. This project will facilitate the identification and characterization of promising leads for the development of future therapies.