Statistics from Altmetric.com
We read with interest the paper by Rehman et al 1 reporting the contribution of Nod2 genotype to the composition of gut microbiota in mice and Crohn's disease (CD) patients. This was followed by a similar description for another CD-predisposing gene, FUT2.2 To date, 163 CD- and ulcerative colitis-risk loci have been identified, and while most of the known causative genes are involved in immune functions and response to infections, their effects on the composition of the gut microbiota are mostly unknown. Studies like those mentioned above are therefore very important, since the relative abundance of specific enteric bacteria has been clearly shown to be of pathogenetic relevance in mouse models of colitis.3 By studying genotype–microbiota correlations in healthy individuals, key information could also be sought for devoid of potentially confounding effects from disease status and therapeutic treatment.
We studied the impact of 30 unequivocal CD-risk loci, each tagged by a single nucleotide polymorphism (SNP), on the mucosa-associated gut microbiota in a well-characterised group of 51 healthy individuals from the PopCol study (see online supplementary methods). We used permutation analysis of variance (ANOVA) with Bray-Curtis distances (see online supplementary methods) to determine the impact of host genotype on genera level community composition, and observed a Bonferroni-corrected statistically significant effect for the SNP rs11747270 (see online supplementary table S1). As shown in figure 1, this correlated with individuals heterozygous for the rare allele (AG; N=8) tending to cluster separately from common homozygotes (AA; N=43) in a non-metric multi-dimensional scaling plot of microbiota composition. We then compared individual genera in the two genotype groups (see online supplementary methods) and found that Bacteroides and Prevotella genera were under- and over-represented, respectively, in the gut microbial community from rs11747270 AG heterozygous individuals (see online supplementary table S2 and online supplementary figure S1; figure 1).
The association of the rs11747270 rare allele G with Prevotella and its negative correlation with Bacteroides is interesting given that these genera define two of the three enterotypes initially proposed as the main classifiers of the human microbiome.4 Clustering microbiota accordingly in our study sample resulted in one rare enterotype (e3) characterised by Prevotella and two more common enterotypes (e1 and e2) typified primarily by Bacteroides, in agreement with the original findings (see online supplementary table S3). We then stratified the three enterotypes by rs11747270 genotype, which revealed a significant correlation between the rare allele G and the Prevotella-predominant enterotype e3 (table 1 and see online supplementary figure S2). Remarkably, the latter was composed for its 66.7% (4/6) by AG heterozygous individuals, who are otherwise rare in the other two enterotype groups (11.8% (2/17) and 8.3% (2/24) in enterotypes 1 and 2, respectively). This finding is evocative of the observation that Prevotella species dominate the microbiome of African populations, where the rs11747270 allele G is much more frequent than in Europeans (respective allele frequencies 29%–53% vs 5%; http://www.hapmap.org). The immunity-related GTPase family M gene (IRGM) maps to the CD-risk locus tagged by the rs11747270 SNP, and the product of this gene is involved in the regulation of autophagy, a cell degradation process through which intracellular bacteria can be eliminated from the host cell. The increased CD-risk associated with IRGM is thought to be due to SNP-driven alterations of its expression that perturb this function. Our results suggest that similar or additional mechanisms may also contribute to the determination of healthy human gut microbiota composition through specific genotype–enterotype interactions. As for NOD2 and FUT2, this also calls for larger-scale analyses aimed at defining eventual causality in the context of inflammatory bowel disease (IBD) and other common diseases.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
CQ and EL contributed equally
Contributors MDA had the original idea, and designed the study together with LA and LE; ANA, NJT and LA collected and characterised the study subjects and acquired data; EL, ANA and LE performed sequencing of microbiota; MDA supervised genotyping; CQ, EL, ANA, AFA, JR and DG carried out statistical analyses; data analysis and interpretation was done by CQ and MDA, who drafted the manuscript with critical revision for important intellectual content from all other coauthors.
Funding Supported by funds from the Swedish Research Council (VR) to MDA and AFA, the Swedish Society of Medicine to LA, the EU project TORNADO to LE and JR and the Söderbergs Foundation to LE. CQ is funded through an EPSRC Career Acceleration Fellowship EP/H003851/1.
Competing interests None.
Ethics approval Karolinska Institutet, Stockholm, Sweden.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.