Integrative Analysis of Gut Microbiota, Plasma Metabolome, and 1 Gene Expression Identifies Causal Mediators in Graves' Disease 2 Pathogenesis

Yuying Zhang^1,2,3,4Jine Li^1,2,3,4Yunfei Luo^1,2,3,4Shuang Liu^1,2,3,4Jianping Liu^1,2,3,4*

• ¹Department of Endocrinology and Metabolism, The 2nd affiliated hospital，Jiangxi Medical College, Nanchang University, Nanchang City, China
• ²Institute for the Study of Endocrinology and Metabolism in Jiangxi Province, Nanchang City, China
• ³Branch of National Clinical Research Center for Metabolic Diseases, Nanchang City, China
• ⁴Jiangxi Key Laboratory of Molecular Medicine，The Second Affiliated Hospital of Nanchang University, Nanchang City, China

Background: Graves' disease (GD) is characterized by hyperthyroidism and is influenced by genetic and environmental factors. The "gut-thyroid axis" establishes a connection between the gut microbiota and GD, yet the underlying potential mechanisms remain unclear. This study employed Mendelian randomization to investigate the causal relationships between the gut microbiota and GD, aiming to identify key microbial taxa and their metabolites, as well as to explore the regulatory roles of relevant genes in the pathogenesis of GD. Methods: We utilized the two-sample Mendelian randomization (MR) approach to evaluate the causal effects of gut microbiota and plasma metabolites on GD. Mediation analysis was conducted to explore the associations of metabolites in linking gut microbiota to GD. Additionally, we employed bioinformatics tools to identify GD-regulating genes within the gut microbiome and validated their expression levels in peripheral blood mononuclear cells from GD mouse models. Results: MR analysis identified eight gut microbes associated with GD, six of which were positively correlated with an increased risk, while two were negatively correlated. Additionally, fifty-six plasma metabolites exhibited potential causal relationships with GD; of these, 27 were positively associated with risk and 29 were negatively associated. Mediation analysis revealed that three bacteria influenced GD through five plasma metabolites. Specifically, the mannose to glycerol ratio and 1-(1-enyl-palmitoyl)-GPC (p-16:0) mediated the effect of Dialister on GD. Gamma-glutamylthreonine mediated the effect of Oscillospira on GD, whereas the ratios of acetylcarnitine (C2) to propionylcarnitine (C3) and adenosine 5'-diphosphate (ADP) to ornithine mediated the effect of MollicutesRF9 on GD. These microbiota regulate plasma metabolites, thereby affecting GD. TAGAP and HERC3 were significantly upregulated in the peripheral blood mononuclear cells of GD mice induced by recombinant adenovirus Ad-TSHR289, while NCEH1 and LYN were significantly downregulated, indicating their potential role in the regulation of GD. Conclusion: This study reveals that the composition of gut microbiota and its related metabolites promote the development of GD through the modulation of gene expression in peripheral blood mononuclear cells. Our findings have significant implications for the advancement of gut microbiota-based diagnostic techniques and targeted therapies for GD.