Abstract
BACKGROUND: Systemic lupus erythematosus (SLE) is an autoimmune disease with unclear etiology. T cell exhaustion (TEX) suppresses the immune response and can be a potential therapeutic strategy for autoimmune diseases. Therefore, this study primarily investigated the mechanism by which TEX influences SLE, offering a novel target for its treatment. METHODS: GSE72326 and GSE81622 were utilized in this study. TEX related genes (TEX-RGs) were obtained from the published literature. Differentially expressed genes (DEGs) were obtained through differential expression analysis. Subsequently, candidate genes were selected by overlapping DEGs and TEX-RGs. These candidate genes underwent protein-protein interactions (PPIs) analysis for further screening. Machine learning was applied to identify candidate key genes from the PPI-identified genes. The candidate key genes exhibiting an area under the receiver operating characteristic (ROC) curve (AUC) greater than 0.7, along with consistent expression trends and significant differences in GSE72326 and GSE81622 were defined as biomarkers. Additionally, enrichment analysis, immune infiltration analysis, chemical compounds prediction and molecular docking were carried out. Importantly, the biomarkers were validated for expression by reverse transcription-quantitative polymerase chain reaction (RT-qPCR). RESULTS: The biomarkers MX1, LY6E, IFI44 and OASL were screened by overlapping 327 DEGs and 1,408 TEX-RGs. Gene set enrichment analysis (GSEA) showed that there was a significant positive correlation between the expression of these biomarkers and immune-related pathways, such as the NOD-like receptor signaling pathway, Toll-like receptor signaling pathway and RIG-I-like receptor signaling pathway significant positive correlation. The immune infiltration of 8 types of immune cells differed significantly in SLE. Naive B cells, resting memory CD4 T cells and resting NK cells were significantly down-regulated in the SLE group. 4 biomarkers showed the highest correlation with resting memory CD4 T cells. Bisphenol A targeted OASL and LY6E, whereas acetaminophen targeted IFI44 and MX1.The binding activity between the biomarkers and the chemical compounds targeting them was very strong. Finally, RT-qPCR expression of MX1, LY6E, IFI44 and OASL was consistent with the results of the dataset. CONCLUSION: MX1, LY6E, IFI44 and OASL were identified as biomarkers related to TEX in SLE. These biomarkers could be detected in the blood for early diagnosis of the disease or to monitor the efficacy of the disease treatment, thus providing a new target for the management of SLE.