Abstract

Background Clear cell renal cell carcinoma (ccRCC), the most common subtype of kidney cancer, exhibits high molecular heterogeneity and poor clinical prognosis. Current prognostic models often lack functional relevance and fail to reflect the immunometabolic complexity of ccRCC. This study aimed to identify robust, functionally relevant prognostic biomarkers through integrative multi-omics analysis and to develop a clinically applicable risk stratification model. Methods Transcriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), in-house proteomic profiles, and Weighted Gene Co-expression Network Analysis (WGCNA)-derived modules were integrated to identify consistently dysregulated genes in ccRCC. A four-gene prognostic signature [15-hydroxyprostaglandin dehydrogenase (HPGD), PSAT1, GLOD5, and SMIM24] was constructed using least absolute shrinkage and selection operator (LASSO) Cox regression. Model performance was evaluated by Kaplan-Meier survival analysis with log-rank tests after median risk-based stratification and by time-dependent concordance indices (C-index) at 1, 3, and 5 years in training and validation cohorts. Gene expression was validated at the mRNA level using TCGA-Kidney Renal Clear Cell Carcinoma (KIRC) RNA sequencing data and quantitative real-time polymerase chain reaction (RT-qPCR) in 10 paired clinical samples, and at the protein level using the Clinical Proteomic Tumor Analysis Consortium (CPTAC) and Human Protein Atlas (HPA). Functional relevance was explored using gene set enrichment analysis (GSEA) and immune cell infiltration profiling. Based on multi-omics findings, HPGD was selected for functional validation and stably overexpressed in Caki-1 cells, followed by Cell Counting Kit-8 (CCK-8), colony formation, wound-healing, and Transwell assays. Results All four genes were consistently downregulated in ccRCC at both mRNA and protein levels. The prognostic model effectively stratified patients into high- and