Barcoded pyrosequencing reveals that consumption of galactooligosaccharides results in a highly specific bifidogenic response in humans

Abstract

Prebiotics are selectively fermented ingredients that allow specific changes in the gastrointestinal microbiota that confer health benefits to the host. However, the effects of prebiotics on the human gut microbiota are incomplete as most studies have relied on methods that fail to cover the breadth of the bacterial community. The goal of this research was to use high throughput multiplex community sequencing of 16S rDNA tags to gain a community wide perspective of the impact of prebiotic galactooligosaccharide (GOS) on the fecal microbiota of healthy human subjects. Fecal samples from eighteen healthy adults were previously obtained during a feeding trial in which each subject consumed a GOS-containing product for twelve weeks, with four increasing dosages (0, 2.5, 5, and 10 gram) of GOS. Multiplex sequencing of the 16S rDNA tags revealed that GOS induced significant compositional alterations in the fecal microbiota, principally by increasing the abundance of organisms within the Actinobacteria. Specifically, several distinct lineages of Bifidobacterium were enriched. Consumption of GOS led to five- to ten-fold increases in bifidobacteria in half of the subjects. Increases in Firmicutes were also observed, however, these changes were detectable in only a few individuals. The enrichment of bifidobacteria was generally at the expense of one group of bacteria, the Bacteroides. The responses to GOS and the magnitude of the response varied between individuals, were reversible, and were in accordance with dosage. The bifidobacteria were the only bacteria that were consistently and significantly enriched by GOS, although this substrate supported the growth of diverse colonic bacteria in mono-culture experiments. These results suggest that GOS can be used to enrich bifidobacteria in the human gastrointestinal tract with remarkable specificity, and that the bifidogenic properties of GOS that occur in vivo are caused by selective fermentation as well as by competitive interactions within the intestinal environment.

Figure 1. Characterization of the fecal microbiota in eighteen subjects that consumed increasing doses of GOS by multiplex pyrosequencing of 16S rDNA tags.

A phylogenetic tree that encompasses the phylum Actinobacteria is shown (A). The tree contains representative sequences of all OTUs detected that were significantly affected by GOS in individual subjects together with sequences of related entries in the database. The latter includes both type strains of known species and sequences from molecular studies of human fecal samples. Sequences were aligned using Muscle 3.6 and the trees were constructed using the neighbor-joining algorithm with 1,000 bootstrap replicates in MEGA 4.0. The sequences from individual subjects are labeled using open black and closed black symbols, and type strains and other sequenced human strains are indicated by grey symbols. Those OTUs that were not significantly affected in all eighteen subjects were labeled as “No significance”. Graphs to the right of the trees show the abundance of the OTUs and bacterial groups that were significantly affected by GOS. The abundances of all of the Bifidobacterium species affected by GOS consumption, for all eighteen subjects, are shown in B. These graphs show mean proportions of the three individual samples taken during the treatment periods for each subject. Baseline and washout refer to samples taken in periods where no GOS was consumed. Repeated measures ANOVA in combination with a Tukey's post-hoc test were performed to indentify differences between treatment and control periods, where * = p<0.05, ** = p<0.01, and *** = p<0.001. Baseline and washout periods were not included in the statistic analysis.

Figure 2. Bubble plots showing differences in the proportions of bacterial taxa as a percentage of the entire bacteria population detected during consumption of 5.0 g (A) and 10.0 g (B) when compared to the control period.

The size of the bubbles is representative of the percent difference. Black ovals represent increases in proportions observed during the GOS consumption period; white ovals represent decreases.

Figure 3. Temporal dynamics of the human fecal microbiota in response to the consumption of increasing doses of GOS shown in five human subjects.

Graphs on the left show proportions of the three main phyla (Actinobacteria, Firmicutes, and Bacteroidetes) and two genera (Bifidobacterium and Bacteroides) that were affected in subjects considered as “responders”. Graphs on the right show proportions of the same three phyla and two genera for subjects considered as “non-responders”.