Text

Objectives and questions: Osteoarthritis (OA) is a leading chronic joint disease with rising incidence due to aging. Rho-GTPases regulate chondrocyte differentiation and homeostasis, yet the role of RHO signaling-related genes in OA remains unclear. This study investigates the diagnostic and predictive value of RHO-DEGs in OA pathogenesis.

Material and methods: Primary chondrocytes were isolated from OA patients undergoing total knee replacement. Inflammation was induced with IL1β treatment, while CDC42, RAC1, and RHOA activation were modulated using inhibitors and activators. RNA sequencing, qRT-PCR, and GLISA assays were performed. Hub RHO signaling-related differentially expressed genes (RHO-DEGs) were identified via WGCNA and PPI network analysis. Public datasets (GSE114007, GSE129147) were analyzed using machine learning to develop diagnostic models, including a nomogram and ROC curves. Functional enrichment, gene set variation analysis (GSVA) and gene set enrichment analysis (GSEA) were conducted.

Results: CDC42-GTP and RAC1-GTP levels increased after IL1β stimulation, indicating RHO pathway activation in OA chondrocytes. Eleven hub RHO-DEGs were enriched in cytoskeletal regulation, motor proteins, tight junctions, focal adhesion, and oxytocin signaling. Two OA molecular subtypes were identified, with FLNA, ACTG1, COL2A1, MMP13, NOS2, and ADAMTS4 overexpressed in subtype 2, while ADAMTS5 was downregulated. GSEA showed upregulation of apoptosis, insulin, JAK-STAT, Wnt signaling, and autophagy in subtype 2. Machine learning identified five diagnostic RHO-DEGs with strong predictive potential, validated by ROC analysis. Correlation and PPI analysis revealed strong interactions between core hub RHO-DEGs and diagnostic RHO-DEGs, both enriched in motor protein-related KEGG pathways. GSVA enrichment analysis showed that RAC1 inhibition and Rho activation increased apoptosis and reduced ACAN degradation. CDC42 inhibition promoted cellular senescence but decreased calcified cartilage gene expression, while Rho activation suppressed hypertrophic chondrocyte markers. Autophagy-related genes were significantly upregulated following CDC42 and RAC1 inhibition. Among diagnostic RHO-DEGs, ALDH3A2 and WWP2 were downregulated in OA, with ALDH3A2 further suppressed by CDC42 inhibition and WWP2 upregulated by Rho activation. TUBB3 showed inconsistent expression patterns between in vitro and patient datasets. Among hub RHO-DEGs, RHOQ and DIAPH3 exhibited subtype-specific expression changes and were dynamically regulated by CDC42 and RAC1 inhibition.

Discussion and conclusions: This study highlights the role of RHO-DEGs in OA and their potential as diagnostic and predictive biomarkers. CDC42 and RAC1 regulate hub and diagnostic RHO-DEGs, emphasizing their role in OA pathogenesis and therapeutic potential.