Abstract
BACKGROUND: Recent studies have increasingly reported abnormal glutathione (GSH) metabolism within the tumor microenvironment across various solid tumors. However, the specific mechanisms underlying aberrant GSH metabolism in pancreatic cancer (PC) remain unclear. This study aims to investigate the prognostic significance of GSH metabolism-related genes in PC and to identify key molecular targets, thereby providing novel perspectives for targeted PC therapy. METHODS: The GSH metabolism gene set was retrieved from the KEGG database. Utilizing single-cell transcriptomic data from the GSE205049 dataset, this study analyzed the variation in GSH metabolic signaling intensity across distinct cell types within the tumor microenvironment of PC. Additionally, transcriptomic data from multiple repositories, including TCGA, ICGC, and GEO, comprising a total of 930 patients with PC, were integrated to construct a prognostic molecular classifier related to GSH metabolism. Furthermore, the role of the key gene GSTA4 in PC was experimentally validated through a series of in vitro assays. RESULTS: Significant differences in GSH metabolic signaling intensity were observed across various cell types in both normal pancreatic and PC tissues. A prognostic signature comprising six GSH metabolism-related genes (GSTA5, PGD, IDH2, GSTA4, GPX2, and GPX3) was established, wherein a high-risk score was associated with a poorer patient prognosis. Notably, GSTA4 expression was significantly reduced in PC tissues, and higher GSTA4 levels were linked to a favorable prognosis. In vitro functional analyses demonstrated that GSTA4 overexpression markedly inhibited PC cell proliferation and migration. CONCLUSION: The GSH metabolism-associated prognostic signature developed in this study effectively identifies high-risk patients with PC. As a prognostic protective factor, GSTA4 exhibits downregulated expression in PC tissues and suppresses tumor proliferation and migration, highlighting its potential as a therapeutic target.