Shared molecular biomarkers and therapeutic targets in rheumatoid arthritis and osteoarthritis: Focus on EIF3B, KHSRP, NCL, PDCD1LG2, and SLC25A37

Abstract

Rheumatoid arthritis (RA) and osteoarthritis (OA) are the most common joint diseases globally, posing major public health challenges. Although they are often distinguished from each other in clinical diagnoses, we hypothesize that RA and OA could have overlapping molecular pathways. Hence, this study was designed to explore the shared molecular changes and potential therapeutic targets for RA and OA. Transcriptome data were obtained from the Gene Expression Omnibus (GEO) database (GSE51588, GSE12021 and GSE55235), which included 117 synovial membrane samples (33 healthy, 50 OA, and 34 RA). Differentially expressed genes (DEGs) were identified using the “limma” package in R, and functional enrichment analyses were conducted using Gene ontology and Kyoto Encyclopedia of Genes and Genomes frameworks. Protein-protein interaction networks were constructed through STRING, and further analyzed using GeneMANIA. Immune cell infiltration in RA and OA samples was evaluated using the CIBERSORT algorithm; microRNA-messenger RNA interactions were predicted through miRanda, miRDB, miRWalk, and TargetScan databases; and lncRNAs targets were identified via the SpongeScan database. Gene-drug interactions were analyzed using DGIdb, and the results were validated in RA and OA mouse models via immunohistochemistry and western blot. In RA and OA, 38 DEGs were identified, including 23 downregulated and 15 upregulated genes (FDR < 0.05, |log2FC| > 0), associated with key pathways such as ubiquitin-mediated proteolysis, mTOR, JAK-STAT, and Wnt signaling. Hub genes, including EIF3B, KHSRP, nucleolin (NCL), PDCD1LG2, SLC25A37, demonstrated significant differential expression (p < 0.05). In addition, the receiver operating characteristic (ROC) curve analysis indicated good diagnostic potential, with areas under the curve (AUC) values ranging from 0.795 to 0.958. Furthermore, immune cell infiltration analysis revealed significant involvement of plasma cells, T cells, monocytes, and dendritic cells (p < 0.05). Several hub genes were targeted by existing drugs, such as NCL by AS1411, and PDCD1LG2 by Pembrolizumab. In vivo validation revealed that EIF3B, KHSRP, NCL, and PDCD1LG2 were downregulated in both RA and OA mouse models compared to controls (p < 0.01), with EIF3B exhibiting higher expression in RA than in OA (p < 0.01). Mitoferrin 1 expression showed no significant differences among groups. These findings suggest that RA and OA share common molecular pathways that may serve as promising diagnostic biomarkers and therapeutic targets.