Abstract
We have previously identified NOD2 genotype and inflammatory bowel diseases (IBD) phenotype, as associated with shifts in the ileal microbiome ("dysbiosis") in a patient cohort. Here we report an integrative analysis of an expanded number of Crohn's disease (CD) related genetic defects in innate immune function (NOD2, ATG16L1, IRGM, CARD9, XBP1, ORMDL3) and composition of the ileal microbiome by combining the initial patient cohort (Batch 1, 2005-2010, n = 165) with a second consecutive patient cohort (Batch 2, 2010-2012, n = 118). These combined patient cohorts were composed of three non-overlapping phenotypes: 1.) 106 ileal CD subjects undergoing initial ileocolic resection for diseased ileum, 2.) 88 IBD colitis subjects without ileal disease (predominantly ulcerative colitis but also Crohn's colitis and indeterminate colitis, and 3.) 89 non-IBD subjects. Significant differences (FDR < 0.05) in microbiota were observed between macroscopically disease unaffected and affected regions of resected ileum in ileal CD patients. Accordingly, analysis of the effects of genetic and clinical factors were restricted to disease unaffected regions of the ileum. Beta-diversity differed across the three disease categories by PERMANOVA (p < 0.001), whereas no significant differences in alpha diversity were noted. Using negative binomial models, we confirmed significant effects of IBD phenotype, C. difficile infection, and NOD2 genotype on ileal dysbiosis in the expanded analysis. The relative abundance of the Proteobacteria phylum was positively associated with ileal CD and colitis phenotypes, but negatively associated with NOD2R genotype. Additional associations with ORMDL3 and XBP1 were detected at the phylum/subphylum level. IBD medications, such as immunomodulators and anti-TNFα agents, may have a beneficial effect on reversing dysbiosis associated with the IBD phenotype. Exploratory analysis comparing microbial composition of the disease unaffected region of the resected ileum between 27 ileal CD patients who subsequently developed endoscopic recurrence within 6-12 months versus 34 patients who did not, suggested that microbial biomarkers in the resected specimen helped stratify patients with respect to risk of post-surgical recurrence.