Lipid Metabolic Rewiring During Continuous Prostate Cancer Progression Defines a Biologically Relevant Four-Gene Prognostic Signature.

Abstract

Background: Lipid metabolic reprogramming is increasingly recognized as a critical feature of prostate cancer progression, but the lipid metabolism-related genes that remain continuously dysregulated from normal tissue to primary tumor and metastatic disease have not been systematically characterized, and their biological and prognostic relevance remains incompletely understood.

Objective: To identify lipid metabolism-related genes associated with continuous prostate cancer progression and develop a prognostic signature for survival stratification.

Methods: Clinical prostate cancer specimens and a high-fat diet (HFD)-driven RM-1 tumor model were first used to evaluate lipid metabolic alterations in vivo. GSE6919 transcriptomic data were used to identify genes shared between the Normal-Primary and Primary-Metastatic transitions. These genes were intersected with a curated lipid metabolism-related gene set, followed by GO and KEGG enrichment analyses. TCGA prostate adenocarcinoma expression and clinical data were used for LASSO regression to construct a prognostic model. The four core genes were further evaluated by clinicopathological correlation analysis, protein- and transcript-level validation in clinical tissues and prostate cancer cell lines, and functional assays under oleic acid-induced lipid stress. Immune infiltration analysis, ssGSEA, and nomogram analysis were performed to assess the biological and clinical relevance of the model.

Results: Clinical tissues showed increased PLIN3 expression, and HFD feeding promoted tumor growth and reinforced lipid metabolic alterations in vivo. A total of 44 lipid metabolism-related genes were identified as continuously dysregulated during prostate cancer progression. These genes were mainly enriched in fatty acid metabolism, lipid catabolism, peroxisome, lipid droplet, glycolysis/gluconeogenesis, arachidonic acid metabolism, and PPAR signaling. Eight genes were significantly associated with overall survival in TCGA, and a four-gene signature comprising ALDH3A2, ENO2, PPP1CB, and PTGIS was established. This model effectively stratified patients into high- and low-risk groups with significantly different survival outcomes. The risk score was positively associated with clinical T stage and Gleason score. The four core genes were also associated with lipid metabolic enzymes, immune infiltration patterns, and multiple metabolism-related pathways. Protein- and transcript-level validation in clinical tissues and prostate cancer cell lines supported the biological relevance of the signature, although PTGIS showed a more context-dependent pattern. Functionally, silencing ENO2 reduced oleic acid-induced lipid peroxidation, whereas silencing PPP1CB enhanced it, while ALDH3A2 showed a more context-dependent effect. A nomogram integrating the risk score with clinical variables improved individualized survival prediction.

Conclusion: We identified lipid metabolism-related genes continuously dysregulated during prostate cancer progression and established a four-gene prognostic signature with potential value for survival prediction and risk assessment. These findings highlight lipid metabolic rewiring as an important component of prostate cancer evolution and provide candidate biomarkers for future mechanistic and translational studies.