Identification and diagnostic model construction of cuproptosis-related biomarkers in bone metastatic prostate cancer

Prostate cancer is the predominant malignancy in males, with bone metastasis representing a defining characteristic of its progression. Bone tissue involvement significantly deteriorates patient quality of life and incurs substantial healthcare expenditure. Emerging evidence suggests a potential link between cuproptosis and tumor biology, thereby identifying cuproptosis-related genes (CRGs) as promising biomarkers for bone metastatic prostate cancer (BMPC). Consequently, this bioinformatics study aimed to identify specific biomarkers and establish a robust diagnostic model.

To conduct this study, we downloaded publicly available datasets on prostate cancer bone metastasis from the Gene Expression Omnibus (GEO) database and integrated them after a rigorous standardization process to address batch effects. Following differential expression analysis to identify the upregulated and downregulated genes, comprehensive functional annotation was performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Gene set enrichment analysis (GSEA), gene set variation analysis (GSVA), and immune infiltration profile analyses were performed. Multidimensional analysis identified 24 cuproptosis-related differentially expressed genes (CRDEGs). Based on 23 of these CRDEGs, a support vector machine (SVM) diagnostic model was created and further optimized using the least absolute shrinkage and selection operator (LASSO) regression, ultimately focusing on 13 key genes. The predictive accuracy of the model was validated using a receiver operating characteristic curve analysis. To elucidate the molecular interactions, protein-protein interaction (PPI) networks and GeneMANIA prediction results were analyzed, along with regulatory network analysis to identify potential therapeutic targets.

In total, 3,401 differentially expressed genes (DEGs) (1,744 upregulated and 1,657 downregulated genes) were identified. Among them, 24 CRDEGs were closely linked to copper metabolism and likely influenced tumor cell survival and migration, including Metallothionein 1M (MT1M), Heat Shock Protein Family D (Hsp60) Member 1 (HSPD1), and Glutathione Peroxidase 4 (GPX4). GO and KEGG enrichment highlighted processes, such as "response to copper ion" and "copper ion detoxification", implicating these genes in prostate cancer bone metastasis. The diagnostic model, constructed via SVM using 13 key genes selected by LASSO regression, exhibited excellent discriminatory ability, with an area under the curve (AUC) of 0.97. PPI analysis revealed dense interactions among the key genes, which was corroborated by GeneMANIA predictions. Exploration of the regulatory network revealed upstream transcription factors (TFs) and downstream pathways that may serve as therapeutic targets for BMPC.

This bioinformatics study identified cuproptosis-related biomarkers and established a high-performance diagnostic model specifically targeting BMPC, despite limitations in sample size and experimental validation. These findings suggested potential molecular mechanisms underlying the pathogenesis of bone metastases in prostate cancer and provided a computational foundation for the development of early detection strategies and targeted therapeutic interventions. Future research should focus on increasing cohort sizes and integrating in vitro and in vivo experiments to further confirm the potential clinical implications of the identified biomarkers.