Background: Imbalanced microbial composition has been linked to the pathogenesis of inflammatory bowel diseases (IBD). Hematopoietic stem cell transplantation (HSCT) proved to be successful in inducing remission in severe, highly refractory Crohn's disease (CD) patients, possibly by erasing immune responses against gut microbes. Gnotobiotic mouse models colonized with human microbiota brought insights into the mechanistic aspects of host-microbe interactions. The aim of this study was to assess the functional role of microbiota signatures associated with different disease states.
Methods: High-throughput 16S rRNA gene amplicon sequencing was performed on (n=147) fecal samples collected from (n=13) healthy donors and (n=31) HSCT-treated CD patients. Germ-free (GF) wild-type (WT) and IL10–/– mice (129 Sv/Ev; n=12 mice/group) were colonized with fecal microbiota from CD patients before and after HSCT at different disease states. Selection of CD patients for transfer into GF mice was based on clinical and endoscopic disease activity, including paired patient samples collected under remission or relapse following HSCT.
Results: Microbiota profiling showed a significantly reduced microbial diversity in patients compared with healthy controls. Patients in remission showed higher microbial diversity than patients in relapse or at active state of the disease. Patients with fistulating or ileal phenotype had the least diverse ecosystems. High level of inter-individual variation was observed. Despite an incomplete transfer of donor microbiota with a 20–40% loss of species-level taxa, humanized mice reflected the dysbiotic features of their respective human donors, indicated by richness and diversity measures. Histopathological evaluation showed moderate to severe inflammation in colon and cecum of the IL10–/– mice associated with microbiota from patients in relapse. In contrast, IL10–/– mice associated with microbiota from patients in remission remained disease-free. To validate the phenotype transfer, we gavaged the mice three times with donor microbiota during the first week of colonization. Remission-associated mice showed higher species richness but still remained disease-free, while relapse-associated mice developed enhanced inflammation measured at the level of fecal complement C3 concentrations. Endpoint microbial composition remained similar, regardless of the number of inoculations and F1 generation of mice displayed a stable engraftment of human microbiota.
Conclusions: Transfer of patient-derived fecal microbiota mimics the disease phenotype in gnotobiotic IL10–/– mice. Bacterial composition, not the number of species is responsible for disease initiation. Humanized mice represent a potential tool for recapitulating disease phenotypes in IBD.