Dysbiosis of gut microbiota and metabolomic alterations in myasthenia gravis: insights from 16S rRNA sequencing and untargeted metabolomics.
Myasthenia gravis (MG) is an autoimmune disorder of neuromuscular transmission. Gut dysbiosis has been implicated in autoimmune pathogenesis, yet integrated microbial and metabolomic profiling in MG remains scarce. To characterize gut microbiota and the fecal metabolome in MG, identify diagnostic biomarkers, and explore associations between microbial taxa, metabolites, and clinical severity.
Fecal samples from 29 MG patients and 10 healthy controls underwent 16S rRNA sequencing and UHPLC-Q-TOF MS metabolomics. LEfSe identified differential taxa; metabolites were screened by VIP > 1.0, P < 0.05, FDR q < 0.05. Random Forest and Spearman correlation assessed biomarker performance and microbiota - metabolite - clinical associations.
MG patients showed significantly reduced alpha- and beta-diversity. LEfSe identified 232 discriminative taxa, with depletion of butanoic acid-producing commensals (Faecalibacterium prausnitzii, Ruminococcus bromii, Bifidobacterium bifidum) and enrichment of Klebsiella. Metabolomics revealed 567 altered metabolites (424 downregulated), including reduced short-chain fatty acids and secondary bile acids (lithocholic, isolithocholic, and allolithocholic acid). The Random Forest metabolite model achieved AUC = 1.0. Spearman analysis revealed that lithocholic acid (P < 0.05) and allocholic acid (P < 0.001) showed positive correlations with the QMGS, while Ruminococcus abundance was positively correlated with butanoic acid (P < 0.01). KEGG analysis implicated cholinergic synapse, bile secretion, sphingolipid signaling, and mTOR pathways.
MG patients exhibit a distinct profile of gut dysbiosis and metabolic disturbances. The specific microbial and metabolic biomarkers identified in this study may offer novel insights for auxiliary diagnosis of MG and guide future microbiota-targeted intervention strategies.
Fecal samples from 29 MG patients and 10 healthy controls underwent 16S rRNA sequencing and UHPLC-Q-TOF MS metabolomics. LEfSe identified differential taxa; metabolites were screened by VIP > 1.0, P < 0.05, FDR q < 0.05. Random Forest and Spearman correlation assessed biomarker performance and microbiota - metabolite - clinical associations.
MG patients showed significantly reduced alpha- and beta-diversity. LEfSe identified 232 discriminative taxa, with depletion of butanoic acid-producing commensals (Faecalibacterium prausnitzii, Ruminococcus bromii, Bifidobacterium bifidum) and enrichment of Klebsiella. Metabolomics revealed 567 altered metabolites (424 downregulated), including reduced short-chain fatty acids and secondary bile acids (lithocholic, isolithocholic, and allolithocholic acid). The Random Forest metabolite model achieved AUC = 1.0. Spearman analysis revealed that lithocholic acid (P < 0.05) and allocholic acid (P < 0.001) showed positive correlations with the QMGS, while Ruminococcus abundance was positively correlated with butanoic acid (P < 0.01). KEGG analysis implicated cholinergic synapse, bile secretion, sphingolipid signaling, and mTOR pathways.
MG patients exhibit a distinct profile of gut dysbiosis and metabolic disturbances. The specific microbial and metabolic biomarkers identified in this study may offer novel insights for auxiliary diagnosis of MG and guide future microbiota-targeted intervention strategies.