Relative abundance of the Prevotella genus within the human gut microbiota of elderly volunteers determines the inter-individual responses to dietary supplementation with wheat bran arabinoxylan-oligosaccharides

Abstract

Background: The human colon is colonised by a dense microbial community whose species composition and metabolism are linked to health and disease. The main energy sources for colonic bacteria are dietary polysaccharides and oligosaccharides. These play a major role in modulating gut microbial composition and metabolism, which in turn can impact on health outcomes.

Results: We investigated the influence of wheat bran arabinoxylan oligosaccharides (AXOS) and maltodextrin supplements in modulating the composition of the colonic microbiota and metabolites in healthy adults over the age of 60. Male and female volunteers, (n = 21, mean BMI 25.2 ± 0.7 kg/m²) participated in the double-blind, cross over supplement study. Faecal samples were collected for analysis of microbiota, short chain fatty acids levels and calprotectin. Blood samples were collected to measure glucose, cholesterol and triglycerides levels. There was no change in these markers nor in calprotectin levels in response to the supplements. Both supplements were well-tolerated by the volunteers. Microbiota analysis across the whole volunteer cohort revealed a significant increase in the proportional abundance of faecal Bifidobacterium species (P ≤ 0.01) in response to AXOS, but not maltodextrin, supplementation. There was considerable inter-individual variation in the other bacterial taxa that responded, with a clear stratification of volunteers as either Prevotella-plus (n = 8; > 0.1% proportional abundance) or Prevotella-minus (n = 13; ≤0.1% proportional abundance) subjects founded on baseline sample profiles. There was a significant increase in the proportional abundance of both faecal Bifidobacterium (P ≤ 0.01) and Prevotella species (P ≤ 0.01) in Prevotella-plus volunteers during AXOS supplementation, while Prevotella and Bacteroides relative abundances showed an inverse relationship. Proportional abundance of 26 OTUs, including bifidobacteria and Anaerostipes hadrus, differed significantly between baseline samples of Prevotella-plus compared to Prevotella-minus individuals.

Conclusions: The wheat bran AXOS supplementation was bifidogenic and resulted in changes in human gut microbiota composition that depended on the initial microbiota profile, specifically the presence or absence of Prevotella spp. as a major component of the microbiota. Our data therefore suggest that initial profiling of individuals through gut microbiota analysis should be considered important when contemplating nutritional interventions that rely on prebiotics.

Trial registration: Clinical trial registration number: NCT02693782 . Registered 29 February 2016 - Retrospectively registered, https://clinicaltrials.gov/ct2/show/NCT02693782?term=NCT02693782&rank=1.

Keywords: Arabinoxylan oligosaccharides (AXOS); Bacteroides; Bifidobacteria; Butyrate; Diversity; Gut microbiota; Prevotella; Propionate; Short chain fatty acids; Wheatbran extract.

Fig. 1

Schematic of the cross-over, randomised, double-blinded design of the human dietary supplementation study, incorporating the addition of Wheat Bran Extract arabinoxylan oligosaccharide (AXOS) and maltodextrin. Control periods, when volunteers were not receiving supplements (5 days each) were included before, in between, and after the supplementation periods (10 days per supplement). Study visits and sample collection days are indicated. Faecal samples were collected at each visit. Blood samples and blood pressure measurements were collected at the start and end of each supplementation period. Food frequency questionnaires were completed at the beginning and end of the study

Fig. 2

Habitual dietary fibre intake from different dietary sources per volunteer from data collected via food frequency questionnaires. The total intake of dietary fibre ranged from 10.6 to 38.5 g/day with an average of 20.9 g/day across the volunteers. Currently, the recommended average fibre intake for adults in the UK is 30 g/day. (https://www.gov.uk/government/publications/sacn-carbohydrates-and-health-report). # Indicates Prevotella-plus individuals

Fig. 3

Mean proportional abundance of bacteria at (a) the phylum and (b) the family level during the maltodextrin, AXOS and control (washout) periods determined by Illumina sequencing. (* = For panel A, p < 0.001; for panel B, p < 0.05)

Fig. 4

Mean proportional bacterial abundance changes at (a) the family level for the Prevotella-plus group (Prevotellaceae p = 0.002, Bifidobacteriaceae p = 0.0002, Bacteroidaceae p = 0.004) and (b) Prevotella-minus group (Bifidobacteriaceae p < 0.0001) during maltodextrin, AXOS supplementation and washout periods. Mean proportional abundance of (c) Prevotella spp. (Prevotella-plus group genus level, increased during AXOS supplementation p = 0.002), (d) Bacteroides spp. (Prevotella-plus group genus level, decreased during AXOS supplementation p = 0.004), and (e) Bifidobacterium spp. (Prevotella-plus group genus level, increased during AXOS supplementation p = 0.0001, Prevotella-minus group, AXOS p < 0.0001) during the maltodextrin, AXOS and washout periods. *p ≤ 0.01, significance from LEfSe analysis comparing across all three supplement periods

Fig. 5

Correlation of habitual fibre intake and Prevotella proportional abundance from first washout sample (baseline) in the Prevotella-plus group of volunteers (R² = 0.503, p = 0.07). One volunteer in the Prevotella-plus group (volunteer 8) did not provide a first washout sample and therefore was not included in this figure

Fig. 6

a Changes in the proportional abundance of selected Bacteroides spp. and Prevotella spp. operational taxonomic units (OTUs) per volunteer during the maltodextrin supplementation periods for the Prevotella-plus and Prevotella-minus groups and M = maltodextrin, A = AXOS supplementation periods, W = washout period. b Correlation between Bacteroides spp. and Prevotella spp. proportional abundance, for the Prevotella-plus group, during the initial washout period (R² = 0.027, p = 0.698), AXOS supplement period (R² = 0.657, p = 0.015) and maltodextrin supplementation period (R² = 0.376, p = 0.106)

Fig. 7

Correlation between the proportional abundance of Bacteroides and Prevotella spp. and propionate (%) during washout, maltodextrin and AXOS supplement periods for the (a) Prevotella-plus group and (b) Prevotella-minus group, (c) the correlation between proportional abundance of Prevotella spp. % and propionate concentration (mM) (p < 0.0001) and (d) the correlation between proportional abundance of Prevotella spp. % and total SCFA concentration (mM) (p < 0.0001)