The Impact of NOD2 Variants on Fecal Microbiota in Crohn's Disease and Controls Without Gastrointestinal Disease

Abstract

Background/aims: Current models of Crohn's disease (CD) describe an inappropriate immune response to gut microbiota in genetically susceptible individuals. NOD2 variants are strongly associated with development of CD, and NOD2 is part of the innate immune response to bacteria. This study aimed to identify differences in fecal microbiota in CD patients and non-IBD controls stratified by NOD2 genotype.

Methods: Patients with CD and non-IBD controls of known NOD2 genotype were identified from patients in previous UK IBD genetics studies and the Cambridge bioresource (genotyped/phenotyped volunteers). Individuals with known CD-associated NOD2 mutations were matched to those with wild-type genotype. We obtained fecal samples from patients in clinical remission with low fecal calprotectin (<250 µg/g) and controls without gastrointestinal disease. After extracting DNA, the V1-2 region of 16S rRNA genes were polymerase chain reaction (PCR)-amplified and sequenced. Analysis was undertaken using the mothur package. Volatile organic compounds (VOC) were also measured.

Results: Ninety-one individuals were in the primary analysis (37 CD, 30 bioresource controls, and 24 household controls). Comparing CD with nonIBD controls, there were reductions in bacterial diversity, Ruminococcaceae, Rikenellaceae, and Christensenellaceae and an increase in Enterobacteriaceae. No significant differences could be identified in microbiota by NOD2 genotype, but fecal butanoic acid was higher in Crohn's patients carrying NOD2 mutations.

Conclusions: In this well-controlled study of NOD2 genotype and fecal microbiota, we identified no significant genotype-microbiota associations. This suggests that the changes associated with NOD2 genotype might only be seen at the mucosal level, or that environmental factors and prior inflammation are the predominant determinant of the observed dysbiosis in gut microbiota.

FIGURE 1.

Flow diagram of recruitment.

FIGURE 2.

Inverse Simpson index of microbial diversity by NOD2 status and case type.

FIGURE 3.

Hierarchical clustering by Jaccard distance metric of the 16S rRNA gene data showing differences by study group and NOD2 status. The panel on the right shows the relative proportions of the 10 most prevalent bacterial families and the cumulative relative proportion of all other bacteria (shown in black).

FIGURE 4.

A, Relative abundance of the 12 most prevalent bacterial families in both CD patients and non-IBD controls. P values are corrected for multiple testing using Holm’s method across all 59 families seen in the sequencing data. Corrected P values < 0.05 are highlighted in bold. B, Relative abundance of the 12 most prevalent bacterial families where samples have been grouped by diagnosis and by NOD2 genotype. P values are corrected for multiple testing using Holm’s method across all 59 families seen in the sequencing data. Mutant NOD2 is defined here as the presence of 2 CD associated mutations (rs2066844, rs2066845, rs2066847); wild-type NOD2 is defined as the absence of any of these mutations.

FIGURE 5.

Concentration of butanoic acid stratified by cohort and by NOD2 status. P values shown are uncorrected and are for Mann-Whitney U tests by NOD2 status within each cohort. Mutant NOD2 is defined here as the presence of 2 CD associated mutations (rs2066844, rs2066845, rs2066847); wild-type NOD2 is defined as the absence of any of these mutations.