BACKGROUND: While chimeric antigen receptor (CAR)-T cell therapy has revolutionized the treatment of relapsed/refractory acute lymphoblastic leukemia (ALL), treatment failures continue to occur. In studying therapeutic T cell function, it has become clear that achieving a memory-like phenotype is ideal for CAR-T production. This is likely related to the enhanced oxidative metabolic potential of this subset, which allows for improved persistence and enhanced anti-leukemia activity in vivo. However, current expansion protocols drive T cells towards terminal differentiation, decreasing the number of T cells fit for the in vivo environment. Finding methods to improve the yield of memory-like cells without sacrificing T cell expansion has been challenging. AMP-activated protein kinase (AMPK) is a key metabolic regulator responsible for promoting mitochondrial biogenesis and oxidative metabolism, and is more active in memory T cells at baseline. It is similarly induced by TCR ligation, making it unlikely that it would significantly detract from proliferation. These properties make activation of AMPK a potential candidate pathway for improving the yield of more functional T cells for CAR-T cell therapy.
METHODS: AMPK is a heterotrimeric protein complex consisting of alpha, beta, and gamma domains. Functionally, the alpha subunit contains the kinase domain, which is activated by phosphorylation. The gamma subunit controls the phosphorylation, and therefore the activity, of the alpha domain. To increase AMPK signaling in T cells, we cloned the gamma subunit into a lentiviral plasmid containing the elongation factor 1a (EF1a) promoter and a green fluorescent protein (GFP) tag. An empty vector, containing GFP only, served as a negative control. Human T cells were isolated from three separate donors, transduced with our lentiviral construct, and expanded in vitro in the presence of IL-2. AMPK activity was assessed by phosphorylation of Thr172 on the AMPKα subunit as well as phosphorylation of S555 on downstream target Unc-51-like autophagy activating kinase (ULK1) using western blot densitometry, normalized to the total protein amounts. Memory marker expression and mitochondrial density (using Mitotracker Red) were analyzed by flow cytometry. Oxidative metabolism and spare respiratory capacity (SRC) were determined using the Seahorse Metabolic Analyzer. Fold changes for in vitro expansion were calculated by adjusting manual cell counts to reflect GFP positivity and CD4+/CD8+ surface staining.
RESULTS: The AMPK gamma subunit was efficiently transduced and expressed by human T cells as measured by GFP expression, qRT-PCR, and western blot analysis. Further, AMPK activity increased in GFP+ cells as indicated by the phosphorylation of AMPKα Thr172 (1.93 +/- 0.05 vs 0.6 +/- 0.09, p<0.001) and ULK1 S555 (1.28 +/- 0.11 vs 0.67 +/- 0.08, p<0.01). Cells transduced with AMPK augmented expression of memory markers CD62L, CD27, and CCR7, with an increased yield of stem cell memory-like T cells marked by co-expression of CD45RA and CD62L (Figure 1). In addition, AMPK-transduced T cells showed a statistically significant increase in mitochondrial density along with notable enhancement of SRC and maximal oxygen consumption rates (Figure 2A,B). Furthermore, the rate of expansion of AMPK-transduced T cells did not differ significantly from Empty-transduced controls, and in fact trended towards increased in both CD4+ and CD8+ cells (Figure 3A). Indeed, the improved rate of expansion in AMPK-transduced CD4+ T cells led to a measurable increase in CD4+ T cell percentages by flow cytometry (Figure 3B).
DISCUSSION: Here we present an efficient and direct method to increase AMPK activity in human T cells and demonstrate that increased AMPK activity endows T cells with a variety of characteristics ideal for CAR-T cell therapy. These features include increased memory-marker expression, enhanced SRC and oxidative metabolism, equivalent to augmented in vitro expansion, and improved CD4+ T cell yields. Further studies are ongoing to assess the activity and function of AMPK-transduced CAR-T cells both in vitro and in vivo.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.