多组学解析肺纤维化小鼠肠道菌群、血清代谢物与肺部基因的互作网络

Multi-omics analysis of the interaction network among gut microbiota, serum metabolites, and pulmonary genes in the mouse model of pulmonary fibrosis

摘要:

目的 探究肺纤维化小鼠肠道菌群、血清代谢物以及肺组织差异基因的变化特征，并通过多组学联合分析解析其潜在关联。方法 运用动式染尘法构建小鼠肺纤维化模型并进行模型评估。采用宏基因组测序技术分析盲肠内容物微生态变化，利用非靶向代谢组学检测血清代谢物变化，同时采用转录组测序分析肺组织差异基因表达情况。综合运用生物信息学方法，进一步挖掘差异菌群、代谢物与基因之间的关联性及潜在功能模块。结果 模型小鼠成功诱导出肺纤维化病理改变，同时伴有肺组织转化生长因子-β (transforming growth factor-beta, TGF-β)、肿瘤坏死因子-α (tumor necrosis factor-alpha, TNF-α) 以及纤维化相关基因表达上调。组学结果显示，小鼠存在肠道菌群紊乱、血清氨基酸代谢失调以及肺转录组重塑等特征。相关性分析表明，4种差异菌群与多个血清代谢物呈强关联，其中嗜黏蛋白阿克曼氏菌(Akkermansia muciniphila)与鼠乳杆菌(Ligilactobacillus murinus)共同关联22个差异代谢物。基于这22个差异代谢物与差异基因构建跨组学关联网络，通过拓扑分析识别出5个关键子网络：(1) 次黄苷三磷酸作为磷酸供体，经多条通路转化为次黄苷二磷酸；(2) 尿苷三磷酸(uridine triphosphate, UTP)经氨基化反应生成胞苷三磷酸(cytidine triphosphate, CTP)；(3) 丝氨酸/苏氨酸激酶11、Fas活化丝氨酸/苏氨酸激酶以及环鸟苷酸依赖性蛋白激酶作为核心激酶节点；(4) 丝氨酸与同型半胱氨酸的反应桥接甲硫氨酸与半胱氨酸代谢通路；(5) 前列腺素H₂被催化转化为血栓素A₂。结论 肺纤维化小鼠模型中肠道菌群、血清代谢物与肺组织差异基因之间存在显著的统计学关联特征，且识别出核心关联网络及潜在功能模块，为肺纤维化的后续机制探索提供了重要参考。

Abstract:

Objective To investigate the changes in gut microbiota, serum metabolites, and differentially expressed genes (DEGs) in the lung tissue of the mouse model of pulmonary fibrosis and explore the potential associations via multi-omics analysis.Methods A mouse model of pulmonary fibrosis was established by the dynamic inhalation exposure method and evaluated. Metagenomic sequencing was performed to analyze the microecological changes in cecal contents. Untargeted metabolomics was employed to detect serum metabolite alterations, and transcriptomic sequencing was conducted to profile DEGs in the lung tissue. Bioinformatics methods were comprehensively used to explore correlations and potential functional modules among differential microbial taxa, metabolites, and genes.Results Pathological changes of pulmonary fibrosis were successfully induced in the model mice, accompanied by the upregulated expression of transforming growth factor-beta (TGF-β), tumor necrosis factor-alpha (TNF-α), and fibrosis-related genes in the lung tissue. Omics results indicated the presence of gut microbiota dysbiosis, serum amino acid metabolic disorder, and lung transcriptome remodeling in the model mice. Correlation analysis demonstrated that the four differential bacterial species were strongly correlated with multiple serum metabolites, among which Akkermansia muciniphila and Ligilactobacillus murinus were jointly associated with 22 differential metabolites. A cross-omics network was constructed with these 22 differential metabolites and DEGs. Topological analysis identified five key subnetworks: (1) Inosine triphosphate serves as a phosphate donor and is converted to inosine diphosphate via multiple pathways; (2) Uridine triphosphate (UTP) undergoes an amination reaction to form cytidine triphosphate (CTP); (3) Serine/threonine-protein kinase 11, Fas-activated serine/threonine kinase, and cyclic GMP-dependent protein kinase act as core kinase nodes; (4) The reaction between serine and homocysteine bridges the metabolic pathways of methionine and cysteine; (5) Prostaglandin H₂ is catalytically converted into thromboxane A₂.Conclusion There are significant statistical correlations among gut microbiota, serum metabolites, and DEGs in the lung tissue in the mouse model of pulmonary fibrosis. We identify the core association network and potential functional modules, which provide references for the subsequent mechanism exploration of pulmonary fibrosis.