Background
Multiple sclerosis (MS) is a chronic debilitating disease characterized by inflammatory demyelination of the central nervous system. Grey matter (GM) lesions have been shown to be closely related to MS motor deficits and cognitive impairment. In this study, GM lesion-related genes for diagnosis and immune status in MS were investigated.
Conclusions
TLR9, CCL5, CXCL8 and PDGFRB were identified as potential biomarkers for GM injury in MS. The effectively predicted diagnosis model will provide guidance for therapeutic intervention of MS.
Methods
Gene Expression Omnibus (GEO) databases were utilized to analyze RNA-seq data for GM lesions in MS. Differentially expressed genes (DEGs) were identified. Weighted gene co-expression network analysis (WGCNA), least absolute shrinkage and selection operator (LASSO) algorithm and protein-protein interaction (PPI) network were used to screen related gene modules and candidate genes. The abundance of immune cell infiltration was analyzed by the CIBERSORT algorithm. Candidate genes with strong correlation with immune cell types were determined to be hub genes. A diagnosis model of nomogram was constructed based on the hub genes. Gene set enrichment analysis (GSEA) was performed to identify the biological functions of hub genes. Finally, an MS mouse model was induced to verify the expression levels of immune hub genes.
Results
Nine genes were identified by WGCNA, LASSO regression and PPI network. The infiltration of immune cells was significantly different between the MS and control groups. Four genes were identified as GM lesion-related hub genes. A reliable prediction model was established by nomogram and verified by calibration, decision curve analysis and receiver operating characteristic curves. GSEA indicated that the hub genes were mainly enriched in cell adhesion molecules, cytokine-cytokine receptor interaction and the JAK-STAT signaling pathway, etc. Conclusions: TLR9, CCL5, CXCL8 and PDGFRB were identified as potential biomarkers for GM injury in MS. The effectively predicted diagnosis model will provide guidance for therapeutic intervention of MS.