Background Rheumatoid arthritis (RA) is a chronic autoimmune disorder characterized primarily by synovial inflammation, often resulting in progressive joint damage and potential multi-system complications. Manganese, an essential trace element, plays a crucial role in various physiological functions; however, its specific mechanisms and implications in autoimmune conditions like RA are not yet fully understood, and investigations in this area remain relatively limited. Methods To identify DEMMRGs, gene expression data from RA patients were obtained from the GEO database, followed by differential expression analysis and WGCNA. Diagnostic genes were filtered through machine learning algorithms (LASSO, SVM, RF) and incorporated into a classification model. V Model validation was first assessed employing ROC curves, a nomogram, and DCA. Subsequently, the biological and translational implications were further explored through functional enrichment, immune infiltration profiling, reconstruction of TF/miRNA networks, and in silico drug sensitivity prediction. Single-cell data processing, clustering, and cell‒cell communication analysis were performed using Seurat and CellChat. Results 6 DEMMRGs were identified via differential analysis and WGCNA. Machine learning selected 5 diagnostic genes (S100A8, ANXA3, C9orf72, FAS, TXN), which showed high diagnostic accuracy (AUC > 0.85). Immune infiltration revealed distinct patterns between RA and controls. Regulatory networks identified key TFs and miRNAs targeting these genes. Two RA subtypes with divergent immune and molecular profiles were identified. Single-cell analysis confirmed elevated expression of diagnostic genes in macrophages and DCs in RA, and cell communication highlighted fibroblasts and macrophages as interaction hubs. Conclusion This study systematically explores the link between manganese metabolism dysregulation and the pathogenesis of RA. It elucidates the disease's underlying mechanisms through integrated omics analysis, revealing changes in cell composition and communication.<Table Float="No" ID="Taba"> <tgroup cols="2"> <colspec align="justify" colname="c1" colnum="1" /> <colspec align="justify" colname="c2" colnum="2" /> <tbody> <row> <entry nameend="c2" namest="c1"> Key Points • Identified 5 manganese metabolism-linked genes (S100A8, ANXA3, C9orf72, FAS, TXN) as highly accurate diagnostic biomarkers for RA (AUC > 0.85), validated across multiple cohorts. • RA was stratified into two subtypes with divergent immune and molecular profiles, revealing the heterogeneity of RA. • Macrophages and dendritic cells in RA synovium overexpress diagnostic genes (S100A8/TXN). Fibroblasts and macrophages drive collagen-mediated cell–cell communication, promoting joint destruction. </entry> </row> </tbody> </tgroup> </Table>

Identification of rheumatoid arthritis manganese metabolism-related diagnostic biomarkers through bulk and single-cell RNA sequencing analysis

期刊论文
发表日期：

2026年3月30日

查看全文

Dong Cheng,
Kaijie Fan,
Xujun Lang,
Yangbiao He,
Xiaoyan Xu

摘要

Background

Rheumatoid arthritis (RA) is a chronic autoimmune disorder characterized primarily by synovial inflammation, often resulting in progressive joint damage and potential multi-system complications. Manganese, an essential trace element, plays a crucial role in various physiological functions; however, its specific mechanisms and implications in autoimmune conditions like RA are not yet fully understood, and investigations in this area remain relatively limited.

Methods

To identify DEMMRGs, gene expression data from RA patients were obtained from the GEO database, followed by differential expression analysis and WGCNA. Diagnostic genes were filtered through machine learning algorithms (LASSO, SVM, RF) and incorporated into a classification model. V Model validation was first assessed employing ROC curves, a nomogram, and DCA. Subsequently, the biological and translational implications were further explored through functional enrichment, immune infiltration profiling, reconstruction of TF/miRNA networks, and in silico drug sensitivity prediction. Single-cell data processing, clustering, and cell‒cell communication analysis were performed using Seurat and CellChat.

Results

6 DEMMRGs were identified via differential analysis and WGCNA. Machine learning selected 5 diagnostic genes (S100A8, ANXA3, C9orf72, FAS, TXN), which showed high diagnostic accuracy (AUC > 0.85). Immune infiltration revealed distinct patterns between RA and controls. Regulatory networks identified key TFs and miRNAs targeting these genes. Two RA subtypes with divergent immune and molecular profiles were identified. Single-cell analysis confirmed elevated expression of diagnostic genes in macrophages and DCs in RA, and cell communication highlighted fibroblasts and macrophages as interaction hubs.

Conclusion

This study systematically explores the link between manganese metabolism dysregulation and the pathogenesis of RA. It elucidates the disease's underlying mechanisms through integrated omics analysis, revealing changes in cell composition and communication.

Key Points

• Identified 5 manganese metabolism-linked genes (S100A8, ANXA3, C9orf72, FAS, TXN) as highly accurate diagnostic biomarkers for RA (AUC > 0.85), validated across multiple cohorts.

• RA was stratified into two subtypes with divergent immune and molecular profiles, revealing the heterogeneity of RA.

• Macrophages and dendritic cells in RA synovium overexpress diagnostic genes (S100A8/TXN). Fibroblasts and macrophages drive collagen-mediated cell–cell communication, promoting joint destruction.