Background: Oxidative stress contributes to the development of prostate adenocarcinoma (PRAD). However, the prognosis prediction and therapy response predicted based on oxidative stress-associated genes lacked comprehensive study. Herein, an integrated bioinformatics approach was adopted to identify the prognosis-associated oxidative stress genes for PRAD. Methods: From Gene-Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases, the transcriptomic data and clinical data were collected. Genes related to oxidative stress were retrieved from oxidative stress pathway "GOBP_RESPONSE_TO_OXIDATIVE_STRESS" in the MsigDB. A risk model was constructed based on the hub genes selected by both WGCNA analysis and LASSO analysis. GEO cohort was used for verifying the model robustness. By running CIBERSORT and ESTIMATE algorithm, immune cell infiltration was quantified. TIDE algorithm and Spearman correlation analysis were used for evaluating the immunotherapy response and drug sensitivity. RT-qPCR analysis and CCK-8 were carried out to validate the results. Results: We established oxidative stress-related gene signature (NUDT7, NTRK3, MAP3K12, DRD5, C3orf18, and B3GALT2) as an independent factor for the prognostic survival of PRAD. In virto experiments showed that MAP3K12 had a higher expression in PRAD cell lines, and knockdown of MAP3K12 inhibited PC3 cell viability. The risk score was positively linked with T_ cells_ regulatory_ Tregs and Macrophages_ M2 and negatively linked with Plasma_cells. High-risk patients showed higher expressions of PDCD1, CD274, CTLA4, LAG3, PDCDILG2, BTLA, HAVCR2, TIGIT, and higher myeloid-derived suppressor (MDSC) score. Docetaxel, Cisplatin, and Bicalutamide could benefit low-risk patients more. Calibration curves and DCA showed an accurate prediction by the nomogram. Conclusion: We established a novel and reliable prognostic model for PRAD patients.
Read full abstract