Abstract
BACKGROUND: Clear cell renal cell carcinoma (ccRCC) represents the most common histological subtype of kidney cancer and constitutes a major global health burden. Despite increasing recognition of the roles of migrasome-related long non-coding RNAs (MRLs) in tumor biology, their prognostic relevance in ccRCC remains largely undefined. Therefore, this study aimed to systematically identify MRLs associated with ccRCC prognosis and construct a robust prognostic model to improve risk stratification and guide potential therapeutic strategies. METHODS: To identify MRLs, we initially performed a correlation analysis integrating transcriptomic profiles with clinical parameters of ccRCC patients from The Cancer Genome Atlas (TCGA). Leveraging the expression matrix of MRLs, we subsequently employed the least absolute shrinkage and selection operator (LASSO) regression to construct a prognostic signature. A comprehensive assessment was carried out to determine the model's predictive robustness. At single-cell resolution, we delineated the cellular composition of ccRCC, capturing transcriptional heterogeneity across distinct cell populations. Ultimately, long non-coding RNAs (lncRNAs) derived from the prognostic model were experimentally validated in clinical specimens through reverse transcription-quantitative polymerase chain reaction (RT-qPCR), and the landscape of migrasome-related genes (MRGs) was further defined using single-cell RNA sequencing (scRNA-seq) data. RESULTS: A prognostic signature comprising five MRLs-EMX2OS, AC106897.1, AC087645.2, AC121338.2, and C5orf66-was established to stratify patient risk. The derived risk score was subsequently validated as an independent predictor of overall survival (OS) in ccRCC patients. A nomogram integrating this score exhibited strong predictive capability. Immune landscape analysis uncovered marked differences in functional immune features between high- and low-risk cohorts defined by the lncRNA-based model. Notably, the high-risk group displayed enrichment of immune-related processes, whereas the low-risk group demonstrated enhanced predicted responsiveness to a spectrum of therapeutic compounds. scRNA-seq identified 17 distinct cellular subpopulations and highlighted the involvement of tumor cell-intrinsic vascular endothelial growth factor (VEGF) signaling in modulating migrasome-associated molecular programs. CONCLUSIONS: This study emphasizes the prognostic utility of a signature comprising five MRLs for ccRCC, offering valuable insights for clinical risk stratification and therapeutic decision-making. Additionally, modulation of tumor cell migration and migrasome function via the VEGF signaling pathway offers a mechanistic basis for targeted intervention.