Microbiome and metabolome features in inflammatory bowel disease via multi-omics integration analyses across cohorts

作者：Lijun Ning, Yilu Zhou, Han Sun, Youwei Zhang, Chaoqin Shen, Zhenhua Wang, Baoqin Xuan, Ying Zhao, Yanru Ma, Yuqing Yan, Tianying Tong, Xiaowen Huang, Muni Hu, Xiaoqiang Zhu, Jinmei Ding, Yue Zhang, Zhe Cui, Jing‐Yuan Fang, Haoyan Chen, Jie Hong · 发表于：Nature Communications · 年份：2023 · DOI：10.1038/s41467-023-42788-0 · 被引用次数：260 · 研究领域：Gut microbiota and health、Tuberculosis Research and Epidemiology、Mycobacterium research and diagnosis

The perturbations of the gut microbiota and metabolites are closely associated with the progression of inflammatory bowel disease (IBD). However, inconsistent findings across studies impede a comprehensive understanding of their roles in IBD and their potential as reliable diagnostic biomarkers. To address this challenge, here we comprehensively analyze 9 metagenomic and 4 metabolomics cohorts of IBD from different populations. Through cross-cohort integrative analysis (CCIA), we identify a consistent characteristic of commensal gut microbiota. Especially, three bacteria, namely Asaccharobacter celatus, Gemmiger formicilis, and Erysipelatoclostridium ramosum, which are rarely reported in IBD. Metagenomic functional analysis reveals that essential gene of Two-component system pathway, linked to fecal calprotectin, are implicated in IBD. Metabolomics analysis shows 36 identified metabolites with significant differences, while the roles of these metabolites in IBD are still unknown. To further elucidate the relationship between gut microbiota and metabolites, we construct multi-omics biological correlation (MOBC) maps, which highlights gut microbial biotransformation deficiencies and significant alterations in aminoacyl-tRNA synthetases. Finally, we identify multi-omics biomarkers for IBD diagnosis, validated across multiple global cohorts (AUROC values ranging from 0.92 to 0.98). Our results offer valuable insights and a significant resource for developing mechanistic hypothese...