Investigating and validating the molecular mechanisms of lipid metabolism regulation in myasthenia gravis development based on single-cell, bulk transcriptomics, and RT-qPCR.

چکیده اصلی

Myasthenia gravis (MG) is a complex autoimmune neuromuscular disorder, and the role of lipid metabolism dysregulation in MG pathogenesis remains unclear. This study aimed to investigate the molecular mechanisms of lipid metabolism regulation in myasthenia gravis development by integrating single-cell and bulk transcriptomic data. This study analyzed bulk RNA sequencing data from the GSE85452 dataset (13 MG patients and 12 healthy controls) and single-cell RNA sequencing data from the GSE227835 dataset (10 MG patients and 10 healthy controls). Differential expression analysis was performed, and weighted gene co-expression network analysis (WGCNA) was conducted to identify disease-related modules. Key genes were screened through the intersection of differentially expressed genes (DEGs), WGCNA hub genes, and lipid metabolism-related genes. Functional enrichment analysis, protein‒protein interaction (PPI) network construction, immune infiltration analysis, and regulatory network analysis were performed. Single-cell analysis was used to characterize cellular heterogeneity and intercellular communication features. A nomogram prediction model was constructed and internally validated using leave-one-out cross-validation (LOOCV). Potential therapeutic compounds were identified through drug prediction and molecular docking analysis. Furthermore, key genes were validated by RT-qPCR in an independent cohort of 5 MG patients and 5 healthy controls. A total of 823 DEGs and 13 co-expression modules were identified, of which 4 modules were significantly associated with MG. Twenty-one candidate genes were screened, and 2 key genes (IRS2 and ALDH2) were ultimately determined. IRS2 was significantly downregulated while ALDH2 was significantly upregulated in MG patients. The nomogram model based on key genes demonstrated excellent predictive performance (AUC = 0.897). Immune infiltration analysis showed increased regulatory T cells (Tregs) and decreased CD4+ memory activated T cells in MG patients. Single-cell analysis identified 10 major cell types. Cell‒cell communication analysis revealed dense interactions among CD4+ T cells, B cells, and CD14+ monocytes. Drug prediction identified metformin and cyclophosphamide as potential therapeutic candidates, and molecular docking confirmed favorable binding affinities. This integrative study generated the hypothesis that lipid metabolism dysregulation, potentially mediated by IRS2 and ALDH2, may contribute to immune dysfunction in MG pathogenesis. These findings provide preliminary insights into the molecular mechanisms underlying MG development and suggest potential diagnostic biomarkers and therapeutic targets. However, these results are hypothesis-generating, and clinical translation is contingent upon rigorous mechanistic validation through functional experiments, animal models, and larger multicenter clinical cohorts.

متن کامل اصلی