Abstract
PURPOSE: This study aimed to identify centrosome amplification-related genes (CA-RGs) as potential biomarkers in osteoarthritis (OA). METHODS: Publicly available databases provided the transcriptome datasets for OA, while a well-established list of CA-related genes was obtained from published resources. A comprehensive computational biology strategy was implemented, integrating differential gene expression profiling, weighted gene co-expression network analysis (WGCNA), Mendelian randomization (MR) analysis, and ROC curve evaluation to screen for potential diagnostic markers. Following this, a predictive nomogram was developed, accompanied by gene set enrichment analysis (GSEA) and immune cell infiltration characterization utilizing the identified biomarkers. Additionally, single-cell RNA sequencing (scRNA-seq) datasets were examined to determine critical cell subpopulations and track biomarker expression patterns across different cell types. Ultimately, clinical specimen analysis through RT-qPCR was conducted to experimentally confirm the expression profiles of these biomarkers. RESULTS: PLK2 and SUN2 were identified as biomarkers. A nomogram model based on them demonstrated high predictive accuracy for OA. GSEA indicated both biomarkers were strongly correlated with lysosomal pathways. Immune infiltration analysis revealed significant differences in 16 immune cell types between OA and control groups. scRNA-seq analysis pinpointed two key cell clusters: preinflammatory chondrocytes (preInfC) and hypertrophic chondrocytes (HTC). Pseudotime trajectory analysis further uncovered dynamic expression patterns: PLK2 showed a sustained increase in HTC, while SUN2 displayed a dip then rise. In preInfC, PLK2 expression fluctuated (increase-decrease-increase), whereas SUN2 rose initially then declined. CONCLUSION: This study identified 2 biomarkers associated with CA and explored their underlying mechanisms of action, providing new insights into potential therapeutic strategies for OA.