A Gene Signature Developed Based on Reactive Oxygen Species to Predict the Metabolism, Immunity, Mutational Status, and Prognostic Survival for Glioblastoma.

Abstract

Aim: To explore the mechanism of regulatory genes related to reactive oxygen species (ROS) in glioblastoma (GBM).

Background: GBM is a brain malignancy with a poor prognosis. ROS plays a critical role in cellular metabolism, signaling, and senescence, and abnormalities in ROS are closely associated with cancer initiation and development. However, the role of ROSregulated genes in GBM remained unknown.

Objective: To explore the role of ROS-regulated genes in GBM and to build a ROS-related prognostic model.

Methods: RNA sequencing and clinical data from GBM patients were collected from public databases. The enrichment scores of ROS-correlated pathway gene sets obtained from The Molecular Signatures Database (MSiDB) were calculated using single sample gene set enrichment analysis (ssGSEA). Subsequently, key ROS-correlated gene modules were sectioned by weighted gene co-expression network analysis (WGCNA). Least absolute shrinkage and selection operator (LASSO) and Cox regression analyses were performed to screen ROS-related genes, which were used to develop a risk model. In addition, the correlation between patients in high-risk and low-risk groups and clinicopathological features, metabolism-related pathways, and pathways related to tumor progression was analyzed. Finally, the difference in immune cell infiltration between patients in the two risk groups was calculated using CIBERSORT.

Results: We found that ROS-related genes could predict the prognosis of patients suffering from GBM and that abnormal activation of the ROS pathway increased the metabolism of sugars, fats, and amino acids. WGCNA identified gene modules closely associated with ROS. A prognostic risk model was created using three key genes (OSMR, SLC6A6, and UPP1). Immune infiltration analysis showed that high-risk Patients had higher levels of macrophage infiltration, and a high-RiskScore was positively correlated with multiple metabolism processes, programmed death, and epithelial-mesenchymal transition (EMT) pathway activity.

Conclusion: The ROS-associated risk model could accurately predict tumor immunity and progression for GBM patients, acting as an effective predictor of GBM prognosis. The present discovery provided a novel understanding of the diagnosis and treatment of GBM patients.