Background: Crohn's disease (CD), ulcerative colitis (UC) and pouchitis are multifactorial and chronic inflammatory diseases of the gastrointestinal tract, termed together as inflammatory bowel diseases (IBD). Pouchitis develops in former patients with UC after total proctocolectomy and ileal pouch-anal anastomosis ("pouch surgery") and is characterized by inflammation of the previously normal small intestine comprising the pouch. It has been extensively shown that broad taxonomic and functional alterations ("dysbiosis") occur in the gut microbiome of patients with IBD. However, the extent to which microbial dysbiosis in pouchitis resembles that of CD or UC has not been investigated in-depth, and the pathogenesis of pouchitis largely remains unknown.
Results: In this study we collected 250 fecal metagenomes from 75 patients with a pouch, including both non-inflamed (normal pouch) and inflamed (pouchitis) phenotypes, and compared them to publicly available metagenomes of patients with CD (n=88), and UC (n=76), as well as healthy controls (n=56). Patients with pouchitis presented the highest level of dysbiosis compared to other IBD phenotypes based on species, metabolic pathways and enzyme profiles, and their level of dysbiosis was correlated with intestinal inflammation. In patients with pouchitis, the microbiome mucin degradation potential was lower, but was accompanied by an enrichment of Ruminococcus gnavus strains encoding specific mucin-degrading glycoside hydrolases, which might be pro-inflammatory. Butyrate and secondary bile acids producers were decreased in all IBD phenotypes and were especially low in pouchitis. Butyrate synthesis genes were positively correlated with total dietary fiber intake. Patients with pouchitis harbored more facultative anaerobic bacteria encoding enzymes involved in oxidative stress response, suggesting high oxidative stress during pouch inflammation. Finally, we have developed enzymes-based classifiers that can distinguish between patients with a normal pouch and pouchitis with an area under the curve of 0.91.
Conclusions: We propose that the non-inflamed pouch is already dysbiotic (function- and species-wise) and microbially is more similar to CD than to UC. Our study reveals microbial functions that underlie the pathogenesis of pouchitis and suggests bacterial groups and functions that could be targeted for nutritional intervention to attenuate or prevent small intestinal inflammation present in pouchitis and CD.