Abstract

INTRODUCTION. Pathogenesis of inflammatory bowel disease (IBD) is likely to be influenced by differences in the function of gut microbiota. Bile acid (BA) metabolism is a bacterial function carried out in the gut lumen mainly by three enzymes: bile salt hydrolase (BSH), hydroxysteroid dehydrogenase (HSDH), and alphadehydoxylase (ADH). BA metabolism starts with BA deconjugation by BSH producing unconjugated BAs which in turn are substrates to HSDH and ADH. BA ­metabolizing microbiota (BAMM) has demonstrated influence on host homeostasis through changes in BA composition. AIM. To investigate the association between inferred BAMM abundance, type of disease, location of mucosal sample and endoscopic inflammation in IBD patients. METHODOLOGY. We recruited Crohn’s disease (CD), ulcerative colitis (UC), IBD unclassified (IBDU) patients and healthy controls (HC) who underwent colonoscopy. A segmental simple endoscopic score >2 and a segmental Mayo endoscopic score > 0 were considered inflamed mucosa in CD and UC/IBDU, respectively. Biopsy samples from terminal ileum (TI) and sigmoid colon (SC) were collected. DNA was isolated for microbial 16s rRNA sequencing and functional content was predicted using PICRUSt. We calculated relative abundance of Clusters of Orthologous Groups COG3049, COG1902 and COG1062, representing BSH, HSDH and ADH activity, respectively. Using a mixed linear regression model, the relation between these COGs, diagnosis (CD vs. UC/IBDU vs. HC), site of biopsy (TI vs. SC) and inflammation (inflamed vs. non­inflamed mucosa) were analyzed. RESULTS. A total of 428 samples (TI=180, SC=248) were analyzed from 262 individuals (CD=98, UC/IBDU=113, HC=51). Mean age was 37±14 years and 45% of subjects were female. In IBD samples, inflamed mucosa was observed in 19.6% of TI and 50.5% of SC. Higher abundance of COG3049 was observed in SC (p<0.05) and increased independently in IBD compared to HC (p<0.05). Similarly, abundance of COG1902 was significantly increased in CD patients (p<0.005) compared to HC and UC/IBDU subjects and increased independently in TI (p<0.05). Finally, higher abundance of COG1062 was significantly associated with inflamed mucosa (p<0.05) regardless of biopsy location and diagnosis. No evidence of association between inflamed mucosa at the time of biopsy and COG1902 and COG3049 was observed. CONCLUSION. We observed significant associations between abundance of microbial BA­ metabolizing activity and IBD phenotypic traits. Increased activity of COG3049 and COG1902 observed in our IBD cohort could influence substrates available for ADH activity (COG1062), which was associated with mucosal inflammation. Simultaneous increase of studied functions could lead to major changes in BA composition affecting the local immune response. Further analysis including host gene expression, which may elucidate this interaction, is currently underway.