Polyamine metabolism related prognostic genes and risk model in prostate cancer insights into tumor microenvironment and drug sensitivity

Abstract Background Polyamine metabolism (PM) is linked to the progression and prognosis of several cancers, but the specific mechanism of Polyamine metabolism-related genes (PMRGs) in Prostate carcinoma (PCa) is not fully understood. This study aims to construct and validate a PMRG-based prognostic risk model for PCa via bioinformatics. Methods In this study, PCa-related datasets (TCGA-PRAD and GSE70769) were used. Candidate genes were identified by intersecting DEGs from differential expression evaluation in TCGA-PRAD dataset with PMRGs. Subsequently, selected genes underwent univariate Cox and LASSO analyses to identify prognostic markers, which was utilized to develop a risk model. This model was validated using GSE70769 dataset. Furthermore, GSEA, tumor microenvironment, drug sensitivity and single-cell RNA sequencing (scRNA-seq) analysis were performed. Results Nine candidate genes were obtained by intersecting 6,668 DEGs and 59 PMRGs. Then, SRM, SAT1, ODC1, SMOX, PAOX, and OAZ3 were identified as prognostic genes, which were markedly up-regulated in the PCa samples compared to control samples. The risk model had a moderate predictive accuracy for the risk of developing PCa. Meanwhile, we also found that prognostic genes were associated with multiple immune factors. The differential immune cells showing significant positive correlations, and all prognostic genes except SAT1 being negatively correlated with most differential immune cells (such as immature dendritic cells). The response to various drugs was significantly different between the two risk cohorts, such as Lapatinib, Bleomycin, Pyrimethamine. Helper T cells and epithelial cells were identified as key cell types, the occurrence of PCa was found to induce alterations in their communication function, and trajectory analysis of these cells revealed that prognostic gene expression changed with differentiation. Conclusions The present study identified six prognostic genes (SRM, SAT1, ODC1, SMOX, PAOX, and OAZ3) related to polyamine metabolism in PCa, which may offer novel insights for the clinical management of patients with PCa.