Multiple omics uncovers host-gut microbial mutualism during prebiotic fructooligosaccharide supplementation

Abstract

Fructooligosaccharide (FOS), a prebiotic well known for its health-promoting properties, can improve the human gut ecosystem most likely through changes in its microbial composition. However, the detailed mechanism(s) of action of FOS in the modulation of the gut ecosystem remain(s) obscure. Traditional methods of profiling microbes and metabolites could barely show any significant features due to the existence of large interindividual differences, but our novel microbe-metabolite correlation approach, combined with faecal immunoglobulin A (IgA) measurements, has revealed that the induction of mucosal IgA by FOS supplementation correlated with the presence of specific bacteria. Furthermore, the metabolic dynamics of butyrate, L-phenylalanine, L-lysine and tyramine were positively correlated with that of these bacteria and IgA production, whereas p-cresol was negatively correlated. Taken together, our focused intraindividual analysis with omics approaches is a powerful strategy for uncovering the gut molecular network and could provide a new vista for understanding the human gut ecosystem.

Keywords: commensal microbiota; correlation analysis; gut ecosystem; metabolite; prebiotics.

Figure 1.

Effect of FOS supplementation on total faecal IgA production in human volunteers. (A) The average levels of faecal IgA in all seven volunteers before, during and after FOS supplementation. NS, no significant difference. (B) The individual average amounts of faecal IgA before, during and after FOS supplementation. Faecal sampling was performed at least twice during each period. P-values between FOS supplementation period and other periods were determined using the non-parametric Mann–Whitney U-test. *P < 0.1; **P < 0.01.

Figure 2.

Effect of FOS supplementation on the human microbiome. (A) Heatmap profiles of individual gut microbiome were analysed at the microbial family level. These profiles were normalized to a relative value between 0 and 1. (B) and (C) PCoA on UniFrac distance matrix from all volunteers before, during and after FOS supplementation. PCoA plots are coloured by individuals (B) and shaded by diet periods (C). (D) The comparison of Unifrac distance between intra- and interindividuals. P-values were determined using the Mann–Whitney U-test. **P < 0.01.

Figure 3.

Effect of FOS supplementation on faecal metabolites. (A) Heatmap profiles of individual metabolomes in this trial. Faecal metabolome profiling was performed by ¹H NMR and the profiles were normalized to a relative value between 0 and 1. (B) and (C) Principal component analysis on faecal metabolome data from all volunteers before, during and after FOS supplementation. Score plots are coloured by individuals (B) and shaded by diet periods (C).

Figure 4.

Individual ecological dynamics in the gut during FOS supplementation. (A) Correlation coefficients between faecal IgA–microbiome profile and (B) faecal IgA–faecal metabolome were calculated by Spearman's rank correlation coefficient using the R software (http://www.r-project.org/). A correlation matrix is represented by negative correlation (R < −0.65; blue) and positive correlation (R > 0.65; red). The correlation similarity among individuals was further clustered by HCA.

Figure 5.

Identification of the gut microbes correlated with faecal IgA induction by FOS supplementation. (A) Correlation coefficients between gut microbiome–faecal metabolome were calculated by Spearman's rank correlation coefficient. The 11 clusters based on the correlation similarity among microbes (the cut-off height of 30, which was determined so that the number of leaf nodes in a cluster is over 10, shown as a dotted line) were further clustered by HCA. (B) Composition of the individually derived microbes in each cluster.

Figure 6.

Molecular network of the gut ecosystem supplemented with FOS. This network was constructed based on the multiple correlation data from ID4 and ID5 using the Cytoscape software (http://www.cytoscape.org/). Green rectangles, blue spheres and purple hexagon indicate metabolites, microbes and IgA, respectively. Sphere size of microbial nodes corresponds to their average relative abundance. Positive and negative correlations are shown as green and red lines, respectively. Purple and blue circles indicate metabolites and microbes, respectively, which highly correlate with faecal IgA abundance. Metabolite a: l-lysine, b: butyrate, c: l-phenylalanine, d: tyramine, e: p-cresol. Microbe 1: ID4 Rikenellaceae, 2: ID5 Bifidobacteriaceae, S24-7, 3: ID5 Peptostreptococcaceae, 4: ID4 Barnesiellaceae, Mogibacteriaceae, Pasteurellaceae and ID5 Clostridiaceae, 5: ID4 Veillonelaceae, Coriobacteriaceae, 6: ID5 Pasteurellaceae, 7:ID5 Desulfovibrionaceae.