AUTHOR=Zhao Jianan , Wei Kai , Shi Yiming , Jiang Ping , Xu Lingxia , Chang Cen , Xu Linshuai , Zheng Yixin , Shan Yu , Liu Jia , Li Li , Guo Shicheng , Schrodi Steven J. , Wang Rongsheng , He Dongyi 

TITLE=Identification of immunological characterization and Anoikis-related molecular clusters in rheumatoid arthritis

JOURNAL=Frontiers in Molecular Biosciences

VOLUME=10

YEAR=2023

URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2023.1202371

DOI=10.3389/fmolb.2023.1202371

ISSN=2296-889X

ABSTRACT=<p><bold>Objective:</bold> To investigate the potential association between Anoikis-related genes, which are responsible for preventing abnormal cellular proliferation, and rheumatoid arthritis (RA).</p><p><bold>Methods:</bold> Datasets GSE89408, GSE198520, and GSE97165 were obtained from the GEO with 282 RA patients and 28 healthy controls. We performed differential analysis of all genes and <italic>HLA</italic> genes. We performed a protein-protein interaction network analysis and identified hub genes based on <italic>STRING</italic> and cytoscape. Consistent clustering was performed with subgrouping of the disease. SsGSEA were used to calculate immune cell infiltration. Spearman’s correlation analysis was employed to identify correlations. Enrichment scores of the GO and KEGG were calculated with the ssGSEA algorithm. The WGCNA and the <italic>DGIdb</italic> database were used to mine hub genes’ interactions with drugs.</p><p><bold>Results:</bold> There were 26 differentially expressed Anoikis-related genes (<italic>FDR</italic> = 0.05, log2FC = 1) and HLA genes exhibited differential expression (<italic>P</italic> &lt; 0.05) between the disease and control groups. Protein-protein interaction was observed among differentially expressed genes, and the correlation between <italic>PIM2</italic> and <italic>RAC2</italic> was found to be the highest; There were significant differences in the degree of immune cell infiltration between most of the immune cell types in the disease group and normal controls (<italic>P</italic> &lt; 0.05). Anoikis-related genes were highly correlated with HLA genes. Based on the expression of Anoikis-related genes, RA patients were divided into two disease subtypes (cluster1 and cluster2). There were 59 differentially expressed Anoikis-related genes found, which exhibited significant differences in functional enrichment, immune cell infiltration degree, and <italic>HLA</italic> gene expression (<italic>P</italic> &lt; 0.05). Cluster2 had significantly higher levels in all aspects than cluster1 did. The co-expression network analysis showed that cluster1 had 51 hub differentially expressed genes and cluster2 had 72 hub differentially expressed genes. Among them, three hub genes of cluster1 were interconnected with 187 drugs, and five hub genes of cluster2 were interconnected with 57 drugs.</p><p><bold>Conclusion:</bold> Our study identified a link between Anoikis-related genes and RA, and two distinct subtypes of RA were determined based on Anoikis-related gene expression. Notably, cluster2 may represent a more severe state of RA.</p>