Bulk and single-cell transcriptome-based integrative analysis of an rRNA processing-related prognostic signature with functional validation of DDX47 in clear cell renal cell carcinoma

Abstract

Clear cell renal cell carcinoma (ccRCC) exhibits pronounced molecular heterogeneity and variable clinical outcomes, complicating prognosis and treatment. While many prognostic models exist, signatures centered on ribosomal RNA (rRNA) processing remain underexplored. This study aimed to establish a prognostic signature associated with rRNA processing and to investigate the functional role of DDX47 in ccRCC. rRNA processing-related genes were curated from Gene Ontology. Prognostic candidates were identified via differential expression and Cox regression in the TCGA-KIRC cohort. A total of 101 machine-learning model combinations were evaluated, and the optimal approach was used to develop the RPPS, with its predictive performance further confirmed in the E-MTAB-1980 and GSE167573 datasets. Prognostic accuracy was evaluated by Kaplan-Meier, ROC, Cox regression, calibration, and decision curve analyses. Immune features, drug sensitivity, and immune escape potential were assessed. Single-cell datasets were integrated to compute an rRNA processing activity score (RPAS), followed by cell–cell communication and pseudotime analyses. DDX47 was knocked out in B cells via scTenifoldKnk, and the resulting perturbed pathways were investigated. DDX47 function was examined in 786-O and A498 cells via siRNA knockdown, Western blotting, EdU, and Transwell assays. The eight-gene RPPS effectively stratified patients into high- and low-risk groups, yielding significantly distinct survival outcomes across all cohorts. The Riskscore independently predicted overall survival and enhanced a nomogram incorporating Age and Stage. High-risk patients showed increased immune infiltration, higher TIDE scores, distinct mutation patterns, and lower predicted IC50 values for selected TKIs. RPAS-low tumor cells displayed enhanced SPP1-mediated communication. Given its high B-cell expression, virtual knockout of DDX47 via scTenifoldKnk revealed enrichment of perturbed genes in immune-related processes, suggesting that DDX47 may be involved in immune-related regulatory programs. High DDX47 expression predicted poor prognosis, and its knockdown suppressed ccRCC cell proliferation, migration, and invasion. The RPPS provides a potential tool for individualized prognostic assessment and molecular stratification in ccRCC. RPAS-defined tumor-cell states may contribute to ccRCC progression through distinct mechanisms, while DDX47 may serve as a candidate prognostic biomarker and warrants further mechanistic investigation in ccRCC.