Prebiotic inulin-type fructans induce specific changes in the human gut microbiota

Abstract

Objective: Contrary to the long-standing prerequisite of inducing selective (ie, bifidogenic) effects, recent findings suggest that prebiotic interventions lead to ecosystem-wide microbiota shifts. Yet, a comprehensive characterisation of this process is still lacking. Here, we apply 16S rDNA microbiota profiling and matching (gas chromatography mass spectrometry) metabolomics to assess the consequences of inulin fermentation both on the composition of the colon bacterial ecosystem and faecal metabolites profiles.

Design: Faecal samples collected during a double-blind, randomised, cross-over intervention study set up to assess the effect of inulin consumption on stool frequency in healthy adults with mild constipation were analysed. Faecal microbiota composition and metabolite profiles were linked to the study's clinical outcome as well as to quality-of-life measurements recorded.

Results: While faecal metabolite profiles were not significantly altered by inulin consumption, our analyses did detect a modest effect on global microbiota composition and specific inulin-induced changes in relative abundances of Anaerostipes, Bilophila and Bifidobacterium were identified. The observed decrease in Bilophila abundances following inulin consumption was associated with both softer stools and a favourable change in constipation-specific quality-of-life measures.

Conclusions: Ecosystem-wide analysis of the effect of a dietary intervention with prebiotic inulin-type fructans on the colon microbiota revealed that this effect is specifically associated with three genera, one of which (Bilophila) representing a promising novel target for mechanistic research.

Trial registration number: NCT02548247.

Keywords: COLONIC MICROFLORA; CONSTIPATION; PREBIOTIC.

Figure 1

Randomised, double-blind, placebo-controlled, cross-over trial. Participants were randomly distributed over two study arms, with the order of placebo and treatment interventions switching between arms. After each run-in and intervention period, faecal samples for microbiota and metabolite analyses as well as metadata were collected (visits V2–V5).

Figure 2

Prebiotic treatment accounts for 0.8% of between-sample microbiome variation. Prebiotic treatment was used as a constraining factor in a Principal Coordinates Analysis with Bray Curtis distance (CAP) on samples taken at the end of intervention (o) as well as run-in periods (x). Samples taken after inulin supplementation (blue) account for 0.8% of between-sample microbiome variation (permutation test, p<0.01). Genera responsive to prebiotic treatment (Bifidobacterium, Anaerostipes and Bilophila, Wilcoxon test, q-value <10⁻⁴, <10⁻³ and <0.01, respectively) were plotted on the ordination.

Figure 3

Inulin consumption increases Bifidobacterium and Anaerostipes abundances, while Bilophila numbers drop. Genus-level cross-over analysis of amplicon profiles shows increase in relative abundances of Bifidobacterium and Anaerostipes spp. on prebiotic intervention (Wilcoxon test, q-value <10⁻⁴ and <10⁻³, respectively) and reveals a decrease in the Bilophila population (q-value <0.01).

Figure 4

Overview of inulin-induced compositional changes in the gut microbiota complemented with mechanistic hypotheses. Radar plot shows scaled variation in relative genus abundances (end riI vs end I) per individual, with decreased abundance of the genus over the inulin intervention situated within and increased abundance outside the solid circle (zero). Variation in abundances of genera identified as inulin-responsive are highlighted.