Deletion of AMPK in Human T Cells Decreases Glycolytic Compensation and Severity of Xenogeneic Gvhd without Impairing GVL Potential

Background: Allogeneic hematopoietic stem cell transplantation is a curative therapy for hematological disorders but is often limited by graft-versus-host-disease (GVHD). Following transplantation, alloreactive T cells rapidly increase metabolism which presents a promising area for therapeutic intervention. We have previously shown that GVHD-causing murine T cells upregulate signaling through the cellular energy sensor AMP-activated protein kinase (AMPK), and that transfer of AMPK KO T cells increased recipient survival while preserving leukemia clearance. Here, we follow-up these original results with further mechanistic data and translational studies in AMPK-deficient human T cells.

Methods: AMPK-deficient human T cells, generated by targeting the AMPKa1 locus via CRISPR cas9 editing, were transplanted with autologous APCs into immunodeficient (NSG) mice. To test anti-leukemia efficacy, AMPK was deleted in both unmanipulated and CD19-targeting CAR T cells, followed by coculture with either Molm13 or Nalm6 GFP+ leukemia cells in the Incucyte analyzer. Murine T cells lacking AMPK were transplanted into B6D2F1 recipients and proteins immunoprecipitated from day 7 cells using an antibody recognizing the phosphorylated AMPK specific motif (LxRxx(pS/pT)). For inhibitor studies, murine T cells were treated with SBI-0206965 for 24 hours prior to allogeneic transplant.

Results: Transplantation of AMPK-deficient human T cells improved recipient survival including median day of survival (not reached vs. 78 days) and percentage alive at twelve weeks post-transplant (89 vs. 45%, p=0.041). Furthermore, deletion of AMPK lessened GVHD clinical scores as early as day 50, a phenotype which persisted through study end (p<0.0001). A second set of human T cells recovered at 3-4 weeks post-transplant exhibited decreased CD4/CD8 ratios (p<0.05) and impaired glycolytic compensation (p<0.001). This finding is consistent with data from day 7 murine cells showing decreased immunoprecipitation of AMPK-phosphorylated glycolytic enzymes including aldolase and GAPDH (p<0.001 and <0.01, respectively). Importantly, both polyclonal and CD19-directed CAR cytotoxicity was preserved in AMPK-deficient human T cells when cocultured with two separate types of leukemia cells. Finally, pretreatment of murine T cells with the AMPK inhibitor, SBI-0206965, reduced the frequency of proliferating allogeneic T cells recovered on day 7 post-transplant.

Discussion: Here we show for the first time that deletion of AMPK in human T cells minimizes GVHD severity, with a concomitant decrease in donor CD4/CD8 ratios and an impairment in T cell glycolytic compensation. Maintenance of cytotoxicity across multiple in vitro leukemia models further demonstrates the preservation of intrinsic GVL responses in AMPK deficient cells. Importantly, pretreatment of murine T cells with an AMPK inhibitor decreased allogeneic T cell proliferation post-transplant, portending the potential for future clinical translation. Further studies will explore AMPK inhibitor pretreatment of T cells in additional models of murine and human GVHD.