Identification of potential biomarkers for differential diagnosis between rheumatoid arthritis and osteoarthritis via integrative genome‑wide gene expression profiling analysis

Abstract

The present study aimed to identify potential novel biomarkers in synovial tissue obtained from patients with Rheumatoid Arthritis (RA) and Osteoarthritis (OA) for differential diagnosis. The genome‑wide expression profiling datasets of synovial tissues from RA and OA cohorts, including GSE55235, GSE55457 and GSE55584 datasets, were retrieved and used to identify differentially expressed genes (DEGs; P<0.05; false discovery rate <0.05 and Fold Change >2) between RA and OA using R software. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses of DEGs were performed to determine molecular and biochemical pathways associated with the identified DEGs, and a protein‑protein interaction (PPI) network of the DEGs was constructed using Cytoscape software. Significant modules in the PPI network and candidate driver genes were screened using the Molecular Complex Detection Algorithm. Potential biomarkers were evaluated by receiver operating characteristic and logistic regression analyses. Large numbers of DEGs were detected, including 273, 205 and 179 DEGs in the GSE55235, GSE55457 and GSE55584 datasets, respectively. Among them, 80 DEGs exhibited identical expression trends in all the three datasets, including 49 upregulated and 31 downregulated genes in patients with RA. DEGs in patients suffering from RA compared with patients suffering from OA were predominantly associated with the primary immunodeficiency pathway, including interleukin 7 receptor (IL7R) and signal transducer activator of transcription 1 (STAT1). The sensitivity of IL7R + STAT1 to differentiate RA from OA was 93.94% with a specificity of 80.77%. The results generated from analyses of the GSE36700 dataset were closely associated with results generated from analyses of GSE55235, GSE55457 and GSE55584 datasets, which further verified the reliability of the aforementioned results. The results of the present study suggested that increased expression of IL7R and STAT1 in synovial tissue as well as in the primary immunodeficiency may be associated with RA occurrence. These identified novel biomarkers may be used to predict disease occurrence and clinically differentiate RA from OA.

Keywords: rheumatoid arthritis; osteoarthritis; pathway; protein-protein interaction; sensitivity; specificity.

Figure 1.

Identification of DEGs between patients with RA and OA. (A) In total, 140 upregulated and 133 downregulated genes were identified in patients with RA from GSE55235 datasets. (B) In total, 103 upregulated and 102 downregulated genes were identified in patients with RA from GSE55457 datasets. (C) In total, 95 upregulated and 84 downregulated genes were identified in patients with RA from GSE55584 datasets. (D) In total, 50 upregulated DEGs in patients with RA were identified as being overlapped between the three datasets. (E) In total, 31 downregulated DEGs in patients with RA were identified as being overlapped between the three datasets. One DEG without a symbol was excluded from the upregulated DEGs, therefore 80 DEGs in total were included in the final analysis. RA, rheumatoid arthritis; OA, osteoarthritis; DEG, differentially expressed gene.

Figure 2.

PPI network analysis, Core network and Gene-pathway network. (A) PPI network of differentially expressed genes (light red, upregulated; green, downregulated). (B) Core of the specific network affecting RA development. (C) Gene-pathway network associated with the development of RA. Larger circles represent genes in the core network. In Cytoscape 3.5.1 software, there are two visual styles, groups and significance. When ‘significance’ and ‘show only pathways with P-values <0.05’ were selected, the colors and the names of the enriched pathways in the figures are consistent with the P-values, therefore the colors of the pathway circles and their accompanying names represent P-values. RA, rheumatoid arthritis; PPI, protein-protein interaction.

Figure 3.

Receiver operating characteristic curves of the six genes in the core network to distinguish rheumatoid arthritis from osteoarthritis using data from the GSE55235, GSE55457 and GSE55584 datasets. Receiver operating characteristic curves of (A) CD3D, (B) CXCR4, (C) IL2RG, (D) IL7R, (E) LCK, (F) STAT1 and (G) IL7R+STAT1 are presented. CD3D, T-cell surface glycoprotein CD3 δ chain; CXCR4, C-X-C motif chemokine receptor 4; IL2RG, interleukin 2 receptor γ; IL7R, interleukin 7 receptor; LCK, LCK proto-oncogene, Src family tyrosine kinase; STAT1, signal transducer and activator of transcription 1.

Figure 4.

Receiver operating characteristic curves of the six genes in the core network to investigate the differentiation between rheumatoid arthritis and osteoarthritis using data from the GSE36700 dataset. Receiver operating characteristic curves of (A) CD3D, (B) CXCR4, (C) IL2RG, (D) IL7R, (E) LCK and (F) STAT1 are presented. CD3D, T-cell surface glycoprotein CD3 δ chain; CXCR4, C-X-C motif chemokine receptor 4; IL2RG, interleukin 2 receptor γ; IL7R, interleukin 7 receptor; LCK, LCK proto-oncogene, Src family tyrosine kinase; STAT1, signal transducer and activator of transcription 1.