Metabolic reprogramming involves in transition of activated/resting CD4+ memory T cells and prognosis of gastric cancer

Abstract

BackgroundLittle is known on how metabolic reprogramming potentially prompts transition of activated and resting CD4+ memory T cells infiltration in tumor microenvironment of gastric cancer (GC). The study aimed to evaluate their interactions and develop a risk model for predicting prognosis in GC.MethodsExpression profiles were obtained from TCGA and GEO databases. An immunotherapeutic IMvigor210 cohort was also enrolled. CIBERSORT algorithm was used to evaluate the infiltration of immune cells. The ssGSEA method was performed to assess levels of 114 metabolism pathways. Prognosis and correlation analysis were conducted to identify metabolism pathways and genes correlated with activated CD4+ memory T cells ratio (AR) and prognosis. An AR-related metabolism gene (ARMG) risk model was constructed and validated in different cohorts. Flow cytometry was applied to validate the effect of all-trans retinoic acid (ATRA) on CD4+ memory T cells.ResultsSince significantly inverse prognostic value and negative correlation of resting and activated CD4+ memory T cells, high AR level was associated with favorable overall survival (OS) in GC. Meanwhile, 15 metabolism pathways including retinoic acid metabolism pathway were significantly correlated with AR and prognosis. The ARMG risk model could classify GC patients with different outcomes, treatment responses, genomic and immune landscape. The prognostic value of the model was also confirmed in the additional validation, immunotherapy and pan-cancer cohorts. Functional analyses revealed that the ARMG model was positively correlated with pro-tumorigenic pathways. In vitro experiments showed that ATRA could inhibit levels of activated CD4+ memory T cells and AR.ConclusionOur study showed that metabolic reprogramming including retinoic acid metabolism could contribute to transition of activated and resting CD4+ memory T cells, and affect prognosis of GC patients. The ARMG risk model could serve as a new tool for GC patients by accurately predicting prognosis and response to treatment.