BATF3/MYC axis drives hyper-proliferation of TET2 deficient CAR T cells

Abstract

Abstract A recent report described a patient presenting a dominant chimeric antigen receptor (CAR) T cell clone at the height of anti-tumor response wherein a TET2 allele was disrupted by the integration of the lentiviral 4-1BB CAR vector (Fraietta et. al. Nature 2018). The enhanced proliferative ability conferred by this chance TET2 disruption opened up the possibility of generating TET2 edited CAR T cells for an enhanced anti-tumor efficacy. We modeled and evaluated loss of TET2 in human T cells engineered to express different CAR receptors in a murine model of human acute lymphoblastic leukemia (ALL). Our results indicate that CAR design plays a critical role in determining anti-tumor efficacy and early stage T cell phenotype upon TET2 disruption. We further observe that TET2 edited CAR T cells, over time, enter a hyper-proliferative state with a near total loss of effector function. The frequency with which TET2 edited T cells achieve hyper-proliferation is dependent on the signaling properties of the CAR receptor. This state is in contrast to the canonical exhaustion state wherein loss of proliferative ability precedes loss of effector function. This hyper-proliferative state shares some transcriptional features with T cell leukemia/lymphoma with sustained upregulation of cell cycle associated factors. Exome analysis identified point mutations and chromosomal aberrations in hyper-proliferative cells, but they were not conserved across different populations. Chromatin accessibility analysis revealed that loss of TET2 sets an epigenetic state that allows for sustained elevated levels of BATF3 which in turn drives a MYC-dependent proliferative program in TET2 deficient CAR T cells.