Multi-kingdom gut microbiome for gastric cancer

Alterations in gut microbiota contribute to gastric cancer (GC) pathogenesis. We conducted deep metagenomic sequencing on fecal samples from a discovery cohort of 182 GC patients and 306 healthy controls, anlyzing microbial composition and networks, functions, predicted metabolomes, and cytokines across enterotypes. Four enterotypes were identified in the discovery cohort, among which pattern 4 emerged as a novel non-canonical type highly enriched in GC (71.4% vs. 13.8% in controls). Individuals with pattern 4 was featured by enrichment of Escherichia coli, Streptococcus spp. (e.g., S. parasanguinis, S. anginosus, S. salivarius), and Ligilactobacillus salivarius in gut microbiota, coupled with depletion of short-chain fatty acid (SCFA)-producing microbes. Functionally, it showed heightened xenobiotic and nitrogen metabolism but reduced C19-steroid biosynthesis and glycosaminoglycan degradation. Metabolome predictions revealed toxin accumulation and SCFA loss in their serum and feces, while immune profiling indicated broad immunosuppression with reduced cytokine levels. Integration of enterotyping markedly improved the case-control discrimination accuracy of random forest classifers in the discovery cohort (optimal area under the curve [AUC] = 0.971). Importantly, pattern 4 persisted in the validation cohort (40.0% of patients vs. 12.9% of controls), and the discovery-trained classifer maintained high diagnostic accuracy (AUC = 0.883). We identify a distinct GC-associated enterotype marked by multi-level dysregulation in taxonomy, function, metabolism, and immuniy, serving as a robust diagnostic biomarker and offering new insights into microbiota-mediated gastric carcinogenesis.