Differences in Systemic IgA Reactivity and Circulating Th Subsets in Healthy Volunteers With Specific Microbiota Enterotypes

Abstract

Changes in the intestinal microbiota have been associated with the development of immune-mediated diseases in humans. Additionally, the introduction of defined bacterial species into the mouse intestinal microbiota has been shown to impact on the adaptive immune response. However, how much impact the intestinal microbiota composition actually has on regulating adaptive immunity remains poorly understood. Therefore, we studied aspects of the adaptive immunity in healthy adults possessing distinct intestinal microbiota profiles. The intestinal microbiota composition was determined via Illumina sequencing of bacterial 16S rRNA genes extracted from the feces of 35 individuals. Blood B-cell and T-cell subsets from the same individuals were studied using flow cytometry. Finally, the binding of fecal and plasma Immunoglobulin A (IgA) to intestinal bacteria (associated with health and disease) Bacteroides fragilis, Prevotella copri, Bifidobacterium longum, Clostridium difficile, and Escherichia coli was analyzed using ELISA. Unsupervised clustering of microbiota composition revealed the presence of three clusters within the cohort. Cluster 1 and 2 were similar to previously-described enterotypes with a predominance of Bacteroides in Cluster 1 and Prevotella in Cluster 2. The bacterial diversity (Shannon index) and bacterial richness of Cluster 3 was significantly higher than observed in Clusters 1 and 2, with the Ruminococacceae tending to predominate. Within circulating B- and T-cell subsets, only Th subsets were significantly different between groups of distinct intestinal microbiota. Individuals of Cluster 3 have significantly fewer Th17 and Th22 circulating cells, while Th17.1 cell numbers were increased in individuals of Cluster 1. IgA reactivity to intestinal bacteria was higher in plasma than feces, and individuals of Cluster 1 had significant higher plasma IgA reactivity against B. longum than individuals of Cluster 2. In conclusion, we identified three distinct fecal microbiota clusters, of which two clusters resembled previously-described "enterotypes". Global T-cell and B-cell immunity seemed unaffected, however, circulating Th subsets and plasma IgA reactivity were significantly different between Clusters. Hence, the impact of intestinal bacteria composition on human B cells, T cells and IgA reactivity appears limited in genetically-diverse and environmentally-exposed humans, but can skew antibody reactivity and Th cell subsets.

Keywords: 16S sequencing; IgA reactivity; IgA+ B cells; enterotypes; helper-T cells; intestinal microbiota; γδT cells.

Figure 1

Clustering of individuals based on their microbiota composition. (A) Unsupervised hierarchical clustering based on Bray Curtis dissimilarity of 35 individuals using all detected bacterial genera (n = 228). Relative abundance of the most abundant genera are shown in the heatmap. When genus name was not assigned, the family name is given. A cluster of related microbiota profiles was determined as a cut-off value at height 0.7 of the dissimilarity index. Sample numbers given on the right # indicate vegetarian individuals. (B) Relative abundance of Bacteroides, Prevotella, and Ruminococcus for individuals within each of the three main clusters defined in A. Each dot represents one individual, red lines are median values. Statistics were calculated with Mann-Whitney U-test. ^***P < 0.001; ^****P < 0.0001.

Figure 2

Bacterial diversity differs between the three specified clusters. (A) Relative abundance per cluster of the 20 most abundant bacteria present in our cohort. Each column shows the median plus interquartile range of the relative abundance of bacterial genera (or if not identified bacterial family), relative abundance of 1.0 = 100%. The numbers below indicate corresponding genus, colors show the bacterial family of the various genera. Ruminococcaceae I, II, and III and Lachnospiraceae I and II could not be assigned to different genera, but differ in OTU clusters. (B) Shannon diversity index for each of the three clusters to determine bacterial diversity. (C) Bacterial richness - defined as the total number of observed unique species per cluster. Bars represent median values with interqartile range. Statistics were calculated with Mann-Whitney U test. ^**P < 0.01; ^***P < 0.001.

Figure 3

Alterations in Th cell subsets in individuals with distinct intestinal microbiota. (A) Absolute numbers of total T cells and T-cell subsets for Clusters 1 (n = 7), Cluster 2 (n = 7), and Cluster 3 (n = 20). Individuals per cluster differ for cell subsets marked with a “#” T cells: C1 = 7, C2 = 8, C3 = 20/CD4 and CD8 T cells: C1 = 7, C2 = 7, C3 = 20; γδT cells: C1 = 5, C2 = 8, C3 = 20 and Tfh: C1 = 4, C2 = 4, C3 = 11. Top right - Ratio of Th1 to Th2 cells. (B) B-cell counts for total B-cells and B-cell subsets for Clusters 1 (n = 7), Cluster 2 (n = 7) and Cluster 3 (n = 19). (C) Ratio of IgA1+/ IgA2+ B-cells for Cluster 1 (n = 5), Cluster 2 (n = 6), and Cluster 3 (n = 18). Bars in A-C represent median values with interquartile ranges. Statistics were calculated using the Mann-Whitney U test: ^*P < 0.05; ^**P < 0.01; ^****P < 0.0001.

Figure 4

Higher IgA reactivity in plasma compared to feces against indicator microbiota organisms, but lack of a significant association to microbiota Cluster. (A) Fecal and plasma IgA binding against the indicator intestinal microbiota microorganisms—Bacteroides fragilis, Prevotella copri, Bifidobacterium longum, Escherichia coli, and Clostridium difficile. (B) Fecal IgA reactivity of all individuals, or (C) between Cluster 1–3 separately. (D) Plasma IgA reactivity of all individuals, or (E) between Clusters 1–3 separately. (F) Plasma and fecal IgA binding to Prevotella copri for individuals containing Prevotella (Yes) or not (No) in their intestinal microbiota. IgA binding for all experiments shown was measured as absorbance at 450 nm relative to a positive control (25 ng/ml IgA isolated from human colostrum, Sigma). Significance was determined using the Wilcoxon matched-paired signed rank test (A) or Mann-Whitney U test (B,C). ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001. Data represent all 35 included volunteers, which are only separated into cluster in panels (C,E) (cluster 1, n = 7; cluster 2, n = 8; cluster 3, n = 20).