Development of a Ferroptosis-based Molecular Markers for Predicting RFS in Prostate Cancer Patients

Abstract

Abstract The goal of this study was to develop a ferroptosis-based molecular signature that can predict recurrence-free survival (RFS) in patients with prostate cancer (PCa). In this study, we obtained ferroptosis-related genes (FRGs) in FerrDb database and clinical transcriptome data in TCGA database and GEO database. Consensus cluster analysis was used to identify three molecular markers of ferroptosis in PCa with differential expression of 40 FRGs, including PD-L1 expression levels.We conducted a new ferroptosis- related signature for PCa RFS using four FRGs identified through univariate and multivariate Cox regression analyses. The signature was validated in the training, testing, and validation cohorts, and it demonstrated remarkable results in the area under the time-dependent receiver operating characteristic (ROC) curve of 0.757, 0.715, and 0.732, respectively. Additionally, we observed that younger patients, those with stage T III and stage T IV, stage N0, cluster 1, and cluster 2 PCa were most accurately predicted by the signature as independent predictors of RFS.DU-145 and RWPE-1 cells were successfully analyzed by qRT-PCR and Western blot for ASNS, GPT2, RRM2, and NFE2L2. We conducted a new ferroptosis-based signature for PCa RFS using four FRGs identified through univariate and multivariate Cox regression analyses. The signature was validated in training, testing, and validating cohorts, with an excellent performance based on the ROC curves respectively of 0.757, 0.715, and 0.732. Furthermore, we found that younger patients or those with stage T III and stage T IV, stage N0, cluster 1and cluster 2 were best predicted by the signature as independent predictors of RFS. DU-145 and RWPE-1 cell lines were successfully analyzed by qRT-PCR and Western blot for ASNS, GPT2, RRM2, and NFE2L2.