Integrating multi-omics, machine learning, and molecular dynamics simulations to identify glutamate metabolism-related biomarkers and drug candidates in rheumatoid arthritis.

BACKGROUND

Rheumatoid arthritis (RA) is a chronic autoimmune disorder marked by progressive joint destruction and functional impairment. Increasing data indicate that glutamate metabolism is critically involved in RA pathogenesis. This analysis aimed to identify glutamate metabolism-related biomarkers and potential RA therapeutics.

METHODS

Integrated analysis of data sourced from the GeneCards and Gene Expression Omnibus databases detected differentially expressed glutamate metabolism genes (DEGMGs). Functional enrichment analysis was implemented. Weighted gene co-expression network analysis and three machine learning algorithms were combined to detect potential RA biomarkers. Immune infiltration characteristics were evaluated via the CIBERSORT algorithm. Single-cell RNA sequencing delineated the cellular localization of biomarkers. Molecular docking and dynamics simulations screened for small-molecule drugs. Finally, quantitative real-time polymerase chain reaction and Western blot experiments in a fibroblast-like synoviocyte model verified the expression levels of detected biomarkers.

RESULTS

This analysis identified 322 DEGMGs. Enrichment analysis revealed their involvement in biological processes, including the tumor necrosis factor, phosphatidylinositol 3-kinase-Akt, and Janus kinase-signal transducer and activator of transcription signaling pathways. Machine learning algorithms ultimately pinpointed four core biomarkers. Combined molecular docking and dynamics simulations revealed favorable binding between azacitidine and the target proteins, characterized by high affinity and complex stability. In vitro experimental results were consistent with the bioinformatic predictions.

CONCLUSION

This study identified four glutamate metabolism-related genes-CXCL10, ENTPD1, GPX3, and PSMB9-as potential biomarkers for RA. Azacitidine was also predicted as having therapeutic potential for RA. Together, these findings advance the understanding of RA pathogenesis and provide a novel theoretical foundation and candidate targets for its clinical diagnosis and targeted drug development.