Abstract

Background Gliomas are the most common primary brain tumors and show marked molecular heterogeneity and variable clinical outcomes, highlighting the need for reliable prognostic biomarkers. Methods We analyzed TCGA and CGGA datasets to identify prognosis-related genes in glioma and performed single-cell analysis and functional validation in glioma cell lines. Results Ten prognosis-related genes were identified, among which GLA showed the strongest association with survival. Elevated GLA expression was significantly associated with higher WHO grade, therapy resistance, and worse prognosis. Time-dependent ROC analysis showed favorable predictive performance for 1-, 3-, and 5-year survival. Single-cell RNA sequencing revealed preferential GLA expression in tumor-associated astrocytes and grade-dependent upregulation. GLA expression was associated with sphingolipid metabolism, hypoxia-related pathways, and ERAD-related pathways, including EDEM2 expression. In glioma cell lines, GLA knockdown suppressed cell viability and downregulated EDEM2. Cross-database analyses further supported the diagnostic and prognostic relevance of GLA across cohorts. Conclusion GLA may serve as a prognostic biomarker in glioma and is associated with metabolic and ERAD-related alterations, providing a potential direction for further mechanistic and therapeutic investigation.