Cross-species engraftment biases and metabolic divergence in gnotobiotic mice humanized with ulcerative colitis microbiota

Ulcerative colitis (UC) is a chronic inflammatory disease of the human colon. Dysbiotic gut microbiota play a central role in its pathogenesis, and alterations in microbial composition and function are closely linked to disease activity. Humanized gnotobiotic mice are increasingly used to study how dysbiotic, human-derived microbial communities shape intestinal inflammation. However, the fidelity of microbiota engraftment and its impact on host physiology and metabolism remain incompletely understood. In this study, we performed a multiomics analysis following fecal microbiota transfer (FMT) from eight patients with active UC into germ-free C57BL/6N mice (five mice per donor). The mice were monitored over three weeks. Longitudinal analysis of microbial communities was performed using 16S rRNA (bacteria) and ITS2 (fungi) amplicon sequencing. Microbial metabolic flux was inferred via genome-scale metabolic modeling, and plasma metabolites were assessed by targeted metabolomics. We observed donor-specific physiological changes in recipient mice, including variations in body weight and adipose tissue. Spontaneous colonic inflammation occurred in one group and was subsequently linked to unintended transfer of Clostridioides difficile, which was previously clinically unrecognised in the donor. While bacterial engraftment overall was generally donor-specific and stable across mice, fungal taxa were transferred inconsistently and at low abundance. Despite similar overall plasma metabolomic profiles, select metabolites, including 3-indoleacetic acid, were differentially associated with specific microbial taxa. Moreover, metabolic modeling revealed disrupted metabolic exchange networks in the mouse microbiota compared to the original human donor communities. In conclusion, while human FMT into germ-free mice reliably transmits bacterial features, it introduces metabolic alterations and fails to fully reproduce the fungal microbiome. These findings underscore the need for cautious interpretation of microbiota-driven effects in gnotobiotic models and highlight the limitations of current approaches in replicating the full complexity of human gut ecosystems.