Metformin fails to replicate the metabolic phenotype of Compound 991 treatment in primary human T cells 2309421

Introduction Chimeric Antigen Receptor T cell (CART) therapy, while highly effective at inducing remission in leukemia patients, only has long-term cure rates of 50%, due in part to short in vivo persistence of the CAR T cells. Previous data demonstrate that pre-infusion treatment of CAR T cells with Compound 991 (991), a direct agonist of AMP-activated protein kinase (AMPK), increases their in vivo potency. However whether metformin, an FDA approved compound, is equally efficacious is unknown. Methods To determine if metabolic priming with metformin produces a metabolic phenotype similar to 991, human T cells (n = 4) were stimulated with CD3/CD28 Dynabeads, cultured with DMSO (vehicle control), metformin, or 991 on days 5 and 7, then rested on day 9. Glucose assays were conducted on culture media from day 9 and oxygen consumption was measured via Seahorse following restimulation on day 11. Immunoblots for phosphorylated Raptor, a downstream target of AMPK, determined pathway activation. Results Doses of 300μM Metformin and 25μM 991 were deemed equivalent based on phosphorylation of ULK1, a downstream target of AMPK. Mitostress assays demonstrated significantly higher basal (p = 0.007) and maximal (p = 0.014) oxygen consumption rates, as well as a higher spare respiratory capacity (p = 0.028), in human T cells treated with 991 as compared to metformin. 991-treated cells also showed significantly reduced glucose utilization (p < 0.0001), a proxy for increased autophagy, with a trend towards elevated Raptor phosphorylation. Conclusion Our results suggest that in human T cells, AMPK activation by metformin does not replicate the metabolic changes conferred by 991 treatment . Based on these results, we predict that distinct pathways of AMPK activation will likely result in different functional outcomes, a premise we are currently testing with 41BB CART cells in vivo.