Abstract. Allogeneic hematopoietic stem cell transplantation (alloSCT) is a curative therapy most commonly applied in treatment of patients with acute myeloblastic leukemia (AML). The progeny of alloreactive ab T cells in the allograft can kill recipient leukemia cells, thereby mediating the graft-vs-leukemia effect (GVL). Nevertheless, disease relapse, indicative of insufficient GVL, is the single most common cause of death post-transplant. GVL-resistance and sensitivity are not equal across classes of leukemias, despite all expressing alloantigens. For example, chronic phase chronic myelogenous leukemia (CP-CML) is exquisitely GVL-sensitive whereas blast crisis CML (BC-CML) is GVL-resistant, despite sharing common biology. Likewise, AML and acute lymphoblastic leukemia are relatively GVL-resistant. Because GVL-resistance and -sensitivity track with the class of neoplasm, we reasoned that they are leukemia cell-intrinsic properties. We therefore established GVL against mouse models of mCP-CML and mBC-CML created with authentic human oncogenes. Through the creation of gene-deficient leukemias we discovered that for effective GVL, mBC-CML requires IFN-γR stimulation whereas GVL is fully intact against STAT1/STAT2-/- mCP-CML that cannot respond to any type of IFN. IFN-γR-/- MLL-AF9 was also GVL-resistant. These data indicate that effective GVL against myeloblastic leukemias requires a high magnitude alloreactive T cell response that generates IFN-g whereas a more smoldering T cell response that does not create much IFN-g can be effective against CP-CML. These results also strongly suggest that the delivery of IFN-g, in conjunction with alloreactive or leukemia-reactive T cells, would have a major clinical impact. However, key questions need to be answered to optimally apply IFN-g or agents that induce some of its effects in the clinic. First, while all AML subtypes share some common biology, there is substantial heterogeneity in driver genes. It will therefore be critical to understand which AML molecular subtypes require IFN-g for optimal GVL. We aim to do so by using CRISPR-Cas9 to delete the IFN- γR from transplantable leukemias that develop in compound gene-edited mice that express key AML-driver genes, sometimes in conjunction with a gene deletion. We have already successfully edited the IFN-γR from JAK2V617F/p53-/- and FLT3-ITD/DNMT3-/-/NPM1c AMLs. Second, IFN-g is a blunt tool with pleiotropic effects that could promote graft-vs-host disease. Therefore, we aim to understand under what alloimmune conditions IFN-g is required and what specific downstream effects are essential to sensitize AML cells to GVL. We have already developed approaches to restore expression of key IFN-g-induced genes (e.g. CIITA, NLRC5, ICAM-1) to IFN-γR-/- leukemias in order to specifically interrogate their relevance. Further targets will be discovered through single cell RNAseq, focusing on IFN-g-induced gene expression changes in subpopulations with stem cell qualities. These targets could then be deleted or expressed in IFN-γR leukemias to further define their importance.

Project Narrative Many patients with blood cancers (leukemias) are treated with a blood stem cell transplant from another person. Much of the benefit of such transplants are from donor immune cells that attack the leukemia cells. However, some types of leukemias benefit more so than other types from this immune effect. In prior work we discovered that leukemias that benefit less need a substance called interferon-gamma in order for them to be killed by the immune cells. In this proposal we test whether this is true for a broader range of leukemias and investigate exactly how interferon-gamma works to sensitize leukemias to immune cell killing. This knowledge could make blood stem cell transplants more effective.