DIEAML · Deciphering immune evasion of Acute Myeloid Leukemia

Acute Myeloid Leukemia (AML) is the most common adult leukemia, with a high frequency of relapse and mortality. After therapy, the persistence of rare leukemia stem cells (LSCs) with disease initiating and therefore stem cell-like properties is a likely cause for relapse. Therefore, novel therapeutic strategies to specifically eradicate LSCs are urgently needed. Natural Killer-cells (NK-cells) are innate immune cells capable of detecting cell surface ligands to trigger a cytotoxic response against diverse tumor cell types. NK-cells are activated either by missing inhibitory or activating ligands, such as known natural killer group 2D receptor ligands (NKG2DL), which become expressed on target cells upon cellular stress or malignant transformation. The groups of Drs. Trumpp, Lengerke and Salih recently described that LSCs evade the recognition from NK-Cells by repressing NKG2DL expression. Importantly, inhibition of Poly [ADP-ribose] polymerase 1 (PARP1) in a subclass of CD34+ AML patient samples reactivated NKG2DL expression and, when combined with allogeneic NK-Cell therapy, eradicated AML in xenotransplantation models. The goal of this proposal is to systematically discover novel NKG2DL regulators in AML to improve immunotherapeutic therapy options for a broad group of AML patients. I have identified monocytic differentiation and DNA hypermethylation as potentially opposing mechanisms of NKG2DL expression. I will combine single-cell transcriptomic and epigenomic interrogation of AML patient samples to discover how NKG2DL are activated during malignant differentiation and how these processes are impeded by DNA hypermethylation. These approaches, together with genome-wide CRISPR-dCas9 gain of function screens, will help identify master regulators of NKG2DL expression. Lastly, validation of these regulators in AML patient samples will expose vulnerabilities of AML for NK-cell based therapies and elaborate on the previously identified link with AML stemness.