Abstract
Purpose: Currently, the early diagnosis and treatment of osteoarthritis (OA) remain a challenge. In the present study, we attempted to explore potential biomarkers for the diagnosis and treatment of OA. Methods: The differentially expressed genes (DEGs) were identified based on three mRNA datasets of synovial tissues for OA patients and normal controls downloaded from the Gene Expression Omnibus (GEO) database. Furthermore, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were used for evaluating gene function related categories. Then, miRNA sequencing was performed for differentially expressed miRNAs' identification. Finally, weighted gene co-expression network analysis (WGCNA) was performed for genes detected by the three mRNA datasets and a competing endogenous RNA (ceRNA) network with DEGs and differentially expressed microRNAs (miRNAs) was constructed for central genes identification. In addition, the relationship between central gene expression and immune infiltration was analyzed, and the candidate agents for OA were predicted based on the Connectivity Map database. Quantitative RT-PCR (qRT-PCR), Western blotting analysis, and immunofluorescent staining were performed to validate the expression levels of differentially expressed miRNAs and differentially expressed target genes in normal and OA tissues and chondrocytes. MiRNA-mRNA network was also validated in chondrocytes in vitro. Results: A total of 259 DEGs and 26 differentially expressed miRNAs were identified, among which 94 miRNA-mRNA interactions were predicted. The brown module in WGCNA was most closely correlated with the clinical traits of OA. After overlapping the brown module genes with miRNA-mRNA pairs, 27 miRNA-mRNA pairs were obtained. A ceRNA network was constructed with 5505 lncRNA-miRNA-mRNA interactions. B-cell translocation gene 2(BTG2), Abelson-related gene (ABL2), and vascular endothelial growth factor A (VEGFA) were identified to be the central genes with good predictive performance, which were significantly correlated with immune cell infiltration in OA, reflected by declined activated dendritic cells (aDCs), and elevated contents of B cells, macrophages, neutrophils, and T helper cells. Anisomycin, MG-132, thapsigargin, and lycorine were predicted to be the potential candidate agents for OA intervention. In vitro, the expression levels of differentially expressed miRNAs and biomarkers identified in the present study were consistent with the results obtained in normal or OA knee cartilage tissues and chondrocytes. Furthermore, BTG2 was identified to be negatively regulated by miR-125a-5p. Conclusion: BTG2, ABL2, and VEGFA can be regarded as potential predictive and treatment biomarkers for OA, which might guide the clinical therapy of OA.