Background: Remarkable durable responses are seen with chimeric antigen receptor (CAR) T cell therapy in B cell lymphoma, however the majority of patients relapse (Locke et al. Lancet Oncol. 2019). Improvements enabling CAR T cells (CAR-T) to circumvent mechanisms of resistance may increase efficacy. Hypoxia, nutrient deprivation and acidosis, all common in the tumor microenvironment (TME), impair metabolic function necessary for CAR-T to kill tumor (Chang et al. Cell 2015). The metabolic response gene peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1α) co-activates genes that upregulate mitochondrial and glycolytic machinery for ATP synthesis from myriad carbon sources. Post translational modifications (PTM) fine tune PGC-1α activity to meet energy demands (Luo et al. IJC 2019). We hypothesized that CAR-T co-expressing full-length PGC-1α or the truncated (ie. short) NT-PGC-1α isoform, with mutations that prevent suppressive PTMs, would confer metabolic flexibility to improve function under TME conditions.
Methods: We constructed four PGC-1α encoded retroviral vectors with an IRES and DsRed fluorescent protein: full-length wild type (WT); full-length mutant (GA); wild type short isoform (NT); and mutant short isoform (mNT). GA contained T295A and S571A mutations to abrogate GSK3β and Akt mediated PTMs. mNT sequence contained K to A mutations at K78/K145/K184/K254 to prevent acetylation by GCN5, and L to A mutations of the nuclear export sequence corresponding to L29/L33/L36/L38. Human CD8 T cells were activated with αCD3/αCD28 beads + 100 IU IL-2/mL, and transduced at 48 hr. to express FMC63-CD28/CD3z CAR and non-functional truncated CD34. Cells were co-transduced with WT, or in the case of metabolically flexible CAR T cells (mfCAR-T) with a mutant and/or short isoform PGC-1α vector. After 7 days of expansion CD34+DsRed+ cells were isolated by FACS. In vitro experiments were performed within 2 weeks to characterize mitochondrial dynamics/oxidative stress (flow cytometry), cytokine secretion (ELISA), and real-time cytotoxicity (xCelligence). The effect of glucose restriction was evaluated in normal (10 mM) and low glucose (0.01 mM) medium. A Mitochondrial stress test (Seahorse) was performed 30 days after FACS. CAR-T (WT and control w/o co-transduction) and mfCAR-T were stimulated with CD19+ K562 or 3T3 cells.
Results: Representative PGC-1α metabolic fitness target genes (ERRα, TFAM, and NRF2) were increased in mfCAR T cells (p≤0.001). mfCAR-T exhibited decreased mitochondrial biomass (p≤0.01) and mitochondrial membrane potential (MMP) (p≤0.01) in both glucose conditions. However, MMP:mitochondrial biomass and autophagy were greater (p≤0.01, p≤0.001), suggesting accelerated mitochondrial quality control (MQC). Oxidative stress was generally decreased (p≤0.01) in mfCAR-T, accompanied by reduced apoptosis. All mfCAR and control CAR T cells cytolysed 100% of targets at a 1:1 ratio but differed in cytolytic rate. Relative to CAR only, WT CAR-T and GA mfCAR-T killed 1.6 and 1.9 times faster, while shorter isoforms required 1.9 times longer to lyse all targets. IFNγ and IL-2 secretion by GA-mfCAR-T was increased above control CAR-T and other mfCAR T cells (p≤0.01), while others were similar. At 30 days both WT-CAR-T and all 3 mfCAR-T had increased spare respiratory capacity (SRC) compared to control CAR-T (p≤0.05); however ATP production and OCR/ECAR was increased (p≤0.001, p≤0.0.05) in mfCAR-T above control CAR-T and WT-CAR-T.
Conclusion: Enforced expression of mutant or truncated PGC-1α in CAR-T enhanced mitochondrial quality control with commensurate function. mfCAR-T cells exhibited equivalent cytotoxicity in vitro, improved survival, and a metabolism less reliant on glucose. Stark differences in SRC, OCR/ECAR, and mitochondrial ATP production between WT and mfCAR-T suggest signaling pathways in CAR T cells may target PTM mediated suppression of PGC-1α and lead to metabolic exhaustion in the TME. mfCAR-T are a promising new strategy to improve the function of CAR-T cells in the TME. Further in vitro and in vivo experiments are needed to validate the approach.
Locke:Kite, a Gilead Company: Consultancy, Research Funding; Celgene/Bristol-Myers Squibb: Consultancy; Cellular Biomedicine Group: Other: Consultancy with grant options; Wugen: Consultancy; GammaDelta Therapeutics: Consultancy; Calibr: Consultancy; Allogene: Consultancy; Novartis: Consultancy.
Author notes
Asterisk with author names denotes non-ASH members.