Abstract
BACKGROUND: Prostate cancer (PCa) is a highly prevalent malignant tumor in males. Lysine β-hydroxybutyrylation (Kbhb), an emerging post-translational modification, plays a critical role in tumorigenic processes. However, its functional mechanism in PCa remains elusive. This study aims to identify Kbhb-related prognostic genes in PCa and provide novel insights for its diagnosis, prognosis and treatment. METHODS: Candidate genes were obtained through differential expression analysis and intersection analysis. Univariate Cox proportional hazards model and least absolute shrinkage and selection operator (LASSO) regression were employed to screen Kbhb-related prognostic genes, and a risk model was constructed and validated. The expression levels of Kbhb modification and related prognostic genes were validated using Western blotting (WB), real-time quantitative PCR (RT-qPCR) and immunohistochemistry (IHC). Additionally, clinical correlation analysis, nomogram development and validation, immune infiltration analysis, tumor mutation burden (TMB) analysis, gene set enrichment analysis (GSEA), drug sensitivity prediction, molecular regulatory network analysis, drug targeting analysis, and molecular docking were performed. RESULTS: Three genes--kinesin family member 4A (KIF4A), TPX2 microtubule nucleation factor (TPX2), and aurora kinase B (AURKB)-were identified as Kbhb-related prognostic genes. Results of WB indicated that the butyrylation level of the H3K27 (H3K27-Bu) protein was significantly upregulated in PCa tissues. RT-qPCR and IHC results demonstrated that the expression levels of KIF4A, TPX2, and AURKB were significantly higher in the PCa than normal tissues. A risk model based on these genes demonstrated discriminatory ability (AUC >0.6) and served as an independent prognostic factor alongside prostate-specific antigen (PSA). The prognostic nomogram showed high accuracy (AUC 0.82-0.88). High-risk patients exhibited distinct immune infiltration profiles and higher mutation frequencies in tumor protein 53 (TP53), Titin (TTN), and speckle-type POZ protein (SPOP). Drug sensitivity analysis linked the risk score to 24 compounds, while molecular docking suggested that estradiol and bisphenol A could target the identified hub genes. CONCLUSION: This study identified KIF4A, TPX2, and AURKB as reliable prognostic biomarkers for PCa. These findings provide a theoretical basis for understanding Kbhb-mediated mechanisms in PCa and offer novel targets for early diagnosis and precision therapy.