Metabolic Reprogramming By IDH1 Neomorph Epigenetically Guides CAR T Cell Differentiation

Adoptive cell therapy using Chimeric Antigen Receptor (CAR) T cells has revolutionized cancer immunotherapy for a number of hematologic diseases. However, challenges remain in Chronic Lymphocytic Leukemia (CLL) where durable remission rates are low. Previously, we found that CAR vector-mediated disruption of the epigenetic regulator TET2 led to the clonal expansion of CD8+ CAR T cells in a CLL patient (Fraietta et al., Nature 2018). TET2 is an ɑ-ketoglutarate (ɑKG)-dependent dioxygenase that catalyzes the conversion of methyl-Cytosine in DNA to hydroxy-methyl-cytosine, which culminates in the removal of methylation marks in silenced genes. The TET2-disrupted CAR T cells exhibited immunophenotypic and epigenetic characteristics of early memory T cells. Given that TET2 is a known tumor suppressor gene, complete ablation via CRISPR/Cas9 could risk malignant transformation of engineered T cells (Jain et al., Nature 2023). Isocitrate dehydrogenase 1 (IDH1) mutant R132H had previously demonstrated exhibition of neomorphic activity that producing excess amounts of D-2-hydroxyglutarate (D-2HG), which is a metabolic inhibitor of TET2 via competitor ɑKG (Figueroa et al., Cancer Cell 2010). We hypothesized that IDH1 neomorph overexpression in CAR T cells safely inhibits TET2 activity and thereby phenocopies TET2 genetic disruption as observed in our patient.

To investigate the biological impact of IDH1 R132H neomorph overexpression (IDH1NM OE) in CAR T cells, we developed a bicistronic expression system for CD19-specific 2^nd generation CAR with 4-1BB/CD3ζ to co-express IDH1 R132H neomorph (IDH1NM) and used wildtype IDH1 (IDH1WT) as a control. Western blotting confirmed IDH1WT and IDH1NM overexpression in manufactured CAR19 T cells. IDH1NM OE led to significant D-2HG accumulation in CAR19 T cells, unlike IDH1WT. The expression of IDH1NM did not affect the cytolytic activity of CAR T cells, confirming the preservation of CAR their effector function. IDH1NM OE, however, significantly augmented CAR T cell response in a repeat tumor challenge model, similar to our previously published TET2 knock-down studies in normal donor T cells (Fraietta et al., Nature 2018). Further, IDH1NM CAR T cells showed significantly increased outgrowth in the CCR7+CD45RA+ central memory phenotype during chronic CAR stimulation. Additionally, we observed that effector cytokine levels, including IFNɣ and TNFɑ, were significantly reduced along with a decrease in CD107a, a major degranulation marker, while IL-2 levels were increased which is consistent with our TET2 knock-down studies. This suggests that IDH1NM OE improves the maintenance of early memory as well as effector function in the face of repeated tumor cell encounter.

Next, we aimed to identify the relevance between metabolic reprogramming and memory function mediated by IDH1NM OE in CAR T cells, based on our hypothesis that IDH1NM OE impact ɑKG-dependent dioxygenase activity including various histone demethylase as well as TET2 in CAR T cells. We observed an increase in methylated histone and a decrease in oxidated cytosine in IDH1NM CAR T cells which experienced repeated CAR stimulation in vitro. This suggests that metabolic reprogramming by IDH1NM OE in CAR T cells leads to epigenetic alterations toward a memory-like phenotype in CAR T cells. Our studies therefore confirm that the metabolic reprogramming of CAR T cells via IDH1 NM OE phenocopies TET2 disruption. Pre-clinical mouse studies are ongoing and will be presented at the ASH.