Bioinformatics analysis of hub genes as osteoarthritis prognostic biomarkers

Abstract

Osteoarthritis (OA) is a progressive cartilage degradation disease, concomitant with synovitis, osteophyte formation, and subchondral bone sclerosis. Over 37% of the elderly population is affected by OA, and the number of cases is increasing as the global population ages. Therefore, the objective of this study was to identify and analyze the hub genes of OA combining with comprehensive bioinformatics analysis tools to provide theoretical basis in further OA effective therapies. Two sample sets of GSE46750 contained 12 pairs OA synovial membrane and normal samples harvested from patients as well as GSE98918 including 12 OA and non-OA patients were downloaded from the Gene Expression Omnibus database (GEO) database. Differentially expressed genes (DEGs) were identified using Gene Expression Omnibus 2R (GEO2R), followed by functional enrichment analysis, protein-protein interaction networks construction. The hub genes were identified and evaluated. An OA rat model was constructed, hematoxylin and eosin staining, safranin O/fast green staining, cytokines concentrations of serum were used to verify the model. The hub genes expression level in the knee OA samples were verified using RT-qPCR. The top 20 significantly up-regulated and down-regulated DEGs were screened out from the two datasets, respectively. The top 18 GO terms and 10 KEGG pathways were enriched. Eight hub genes were identified, namely MS4A6A, C1QB, C1QC, CD74, CSF1R, HLA-DPA1, HLA-DRA and ITGB2. Among them, the hub genes were all up-regulated in in vivo OA rat model, compared with healthy controls. The eight hub genes identified (MS4A6A, C1QB, C1QC, CD74, CSF1R, HLA-DPA1, HLA-DRA and ITGB2) were shown to be associated with OA. These genes can serve as disease markers to discriminate OA patients from healthy controls.

Figure 1

The volcano plots and heat maps of GSE46750 and GSE98918 datasets. (A) The volcano plot of GSE46750 dataset. The x-axis represents log2 (Fold Change) and the y-axis represents − log10 (p value). Red dots indicate up-regulated genes and blue dots indicate down-regulated genes. (B) The volcano plot of GSE98918 dataset. (C) The heat map of GSE46750 dataset. Every line represents one gene and every column represents one sample. Red colour indicates high-expression level and blue colour indicates low-expression level. (D) The heat map of GSE98918 dataset. (E) The cross-comparability evaluation of GSE46750 dataset. (F) The cross-comparability evaluation of GSE98918 dataset.

Figure 2

Venn diagram of common DEGs in both GSE46750 and GSE98918 datasets.

Figure 3

Bubble plots of GO and KEGG pathway enrichment analysis results. (A) Bubble plots visualized the Gene Ontology (GO) enrichment analysis results of common DEGs (co-DEGs). The different depths of nodes’ color represent the different adjusted p value. The different sizes of the nodes represent the different number of genes. (B) Bubble plots visualized the Kyoto Encyclopedia of Genes and Genome (KEGG) pathway enrichment analysis results of common DEGs (co-DEGs).

Figure 4

PPI network analysis and hub genes identification. (A) A key cluster with 12 genes was identified by MCODE. (B) Top 8 hub genes explored by CytoHubba.

Figure 5

Hub genes ROC curves. (A) ROC curve analysis of hub genes including MS4A6A, C1QB, C1QC, CD74, CSF1R, HLA-DPA1, HLA-DRA and ITGB2 in the GSE46750 dataset. (B) ROC curve analysis of hub genes in the GSE98918 dataset.

Figure 6

PCA plot of the six hub genes. PC1 and PC2 were the first and second principal components, which were the potential variants’ explanation effect on different expression, respectively. Plots represent different samples, and different colors represent different groups.

Figure 7

Verification hub genes’ expression in OA rat model. (A) HE staining of the growth plates of control and OA rat model joints. Scale bar: 100 μm. (200×) (B) The growth plates of control and OA rat model joints were stained using Safranin-O & fast green. Scale bar: 100 μm. (200×) (C) ELISA results of inflammatory cytokines levels in control and OA rat model serum. (D) RT-qPCR results of MS4A6A, C1QB, C1QC, CD74, CSF1R, HLA-DPA1, HLA-DRA and ITGB2 in control and OA rat model joints tissues.