The human intestinal tract is populated by a complex microbial community, providing essential metabolic and immunological functions to the host, influencing normal physiology as well as disease susceptibility [1
]. Using metagenomic techniques, more than 2000 bacterial species have been detected in the intestinal tract and approximately 100–1000 different bacterial taxa are present in a given individual [2
]. The majority of these bacteria are members of the phyla Bacteroidetes and Firmicutes, while others belong to Proteobacteria, Actinobacteria and Verrucomicrobia [4
]. The composition of the microbiota in the steady state is characterized by high diversity and marked inter-individual variability. On a functional level, however, different species share metabolic properties, resulting in set of core metabolic functions that are shared by the healthy microbiota. Bacterial communities are stabilized by microbe–microbe interactions and exhibit a remarkable resilience to perturbations. In spite of these mechanisms, alterations of the microbiota have been observed in numerous pathologic conditions [6
Accordingly, there is considerable interest in approaches aimed at altering microbial composition and functions counteracting disease-driving factors and promoting health-beneficial properties. Various approaches including dietary and non-dietary interventions have been described and probiotic and prebiotic modalities are most widely used in clinical practice. Following the definition of probiotics, ‘live microorganisms which when administered in adequate amounts confer a health benefit’ [12
], a prebiotic is considered ‘a substrate that is selectively utilized by host microorganisms conferring a health benefit’ [13
]. These features are thought to demarcate a prebiotic from dietary fibers. Accordingly, even if dietary fibers such as pectins [14
] and xylans [15
] may cause alterations of gut commensal compositions, either the criterion of selective utilization or a health benefit have not yet been sufficiently demonstrated [13
Partially hydrolyzed guar gum (PHGG) is a soluble dietary fiber with a linear backbone of β-1,4-linked D-mannose residues, with α-D-galactose residues 1,6-linked to every second mannose. This galactomannan completely dissolves in water, does not form a gel and demonstrates prebiotic properties as previously defined by increasing the abundance of Lactobacilli and Bifidobacteria as well as colonic SCFA contents [16
]. In functional gastrointestinal disease such as irritable bowel syndrome (IBS), PHGG showed efficacy in improving bloating [18
] and improving IBS-related symptoms and quality of life [19
] compared to placebo. Fermentation of PHGG by the microbiota increases the abundance of short-chain fatty acids (SCFAs) including acetate, propionate and butyrate. SCFAs are an important energy source for colonic epithelial cells and have a variety of regulatory functions on gut physiology, metabolism and immunity [20
Despite some evidence for its efficacy and its clinical and nutritional use [21
], the prebiotic effects of PHGG, in light of the prebiotic definition and methodological progresses that have been made in studying the microbiota in recent years, are inadequately defined. The aim of this study was to decipher and define the overall prebiotic properties of PHGG in healthy volunteers focusing on taxonomic and functional (i.e., metabolites) properties using state-of-the-art methodologies.
Microbiota research provides us with fascinating insights into how an altered microbiota may be linked to human disease [39
]. Prebiotics look back upon a long tradition in clinical nutrition and their microbiota-modulating properties along with their favorable safety fertilize a growing prebiotic market [40
]. This prompted the International Scientific Association for Probiotics and Prebiotics (ISAPP) to revitalize the definition of a prebiotic as ‘a substrate that is selectively utilized by host micro-organisms conferring a health benefit’ [13
]. Herein, we sought to verify and re-define prebiotic effects of the widely used soluble fiber, partially hydrolyzed guar gum (PHGG), applying state-of-the-art methodology in a controlled cohort trial named PAGODA.
In day-to-day clinical practice, prebiotics have been shown to exert beneficial effects in different clinical contexts such as gastrointestinal, cardiometabolic, mental health and bone diseases [40
]. Accordingly, PHGG clinical studies have also demonstrated beneficial effects particularly in functional gastrointestinal disorders such as irritable bowel syndrome or constipation [41
]. The specific study design of the PAGODA trial, involving three serial study periods, allowed measuring specific clinical effects in a time- and exposure-resolved manner. Confirmatory of previously published data [17
], in our study, PHGG exposure was associated with an increase in stool frequency and a reduction in stool consistency according to the BSS score.
The prebiotic definition by the ISAPP group refers to the feature of ‘selective utilization’, which translates into utilization of the compound by specific members of the microbiota [13
], thus conferring a fitness advantage with a respective impact on microbiota composition and function. Due technical limitations in earlier days, this was often equated with an expansion of Bifidobacteria and Lactobacilli and an increase in the abundance of SCFAs [13
]. These prebiotic characteristics have also been demonstrated for PHGG [44
]. In order to consider this definition, we have chosen a study design in which each study participant can serve his or her own control allowing to compensate for effects driven by the individual microbiota. To minimize the risk of a selection bias determined by the choice of the 16S chemistry, all samples were amplified and sequenced using both V1–V3 and V3–V4 primer pairs. To obtain functional insights we performed untargeted NMR spectroscopy in order to identify PHGG-induced metabolic pathways and to control for PHGG intake and its breakdown products.
Indeed, we identified PHGG-induced alterations of microbial communities with changes in the indices of species evenness and with an increase in pairwise intra-individual Bray–Curtis dissimilarities. Notably, the observed changes were largely reproducible between the two studied regions of the bacterial 16S gene. The nuanced differences between the findings from the two regions reflect their varying sensitivity for detection of certain bacterial groups. In accordance with the suggestions by Bindels et al. [45
], we demonstrated that PHGG exerts broader effects on microbial ecology affecting 28 different taxa. In line with Bindels et al., our data provide further evidence that ‘selective utilization’ may be less dependent on specific taxa but on shared metabolic traits. Such metabolic traits may then provide secondary product that affect other groups microbes by cross-feeding highlighting the complexity of microbial ecology. Interestingly, the majority of these effects were transient and reversed during the washout period. As clinical effects and benefits along the prebiotic–microbiota–host axis are mechanistically linked by bacterial molecules arising from an altered microbial metabolic activity [39
], we sought to identify changes in bacterial metabolites induced by PHGG. As expected and in line with previous research [16
], induction of SCFAs was the most prominent effect, although additional functional changes in amino acid metabolism were identified. NMR spectroscopy made it possible to quantify fecal concentrations of the prebiotic PHGG itself along with its breakdown products. Interestingly, a daily intake of 15 g PHGG was fully metabolized by microbial metabolism and PHGG remained undetectable in the feces, whereas its cleavage products galactose and mannose were enriched. To link compositional data from 16S metagenomics with the functional metabolomic results, we employed interaction network analysis, which showed associations between OTU abundance and metabolites, particularly SCFAs, hinting at an accumulation of enzymatic functions of these bacterial species.
Another particularly interesting finding of this study was a strong inter-individual difference in response to PHGG between study participants. These differences were very pronounced with regard to clinical effects and generally stronger in male than in female participants. Nutritional data analyses revealed that women had a significantly higher baseline fiber intake, which could be a possible explanation for this discrepancy. However, differences in clinical responsiveness were not paralleled on the metagenomic level. This is consistent with findings reported by De Palma et al., who performed gnotobiotic models in animals that were conventionalized with either IBS or control stool. The authors did not find differences in the taxonomic composition between IBS- and control-conventionalized animals [46
]. Nevertheless, IBS-D but not control feces induced typical IBS alterations in gut function and behavior, underscoring the importance of bacterial metabolites mechanistically linking the microbiota with human disease [46
]. Strikingly, in our study, these gender-specific differences were reproducible on the metabolomic level, suggesting a mechanistic link between microbial function and clinical phenotype in our study. The absence of a gender difference in the effects of PHGG on a metagenomic level most likely reflects the functional convergence of individua, microbiotas, i.e., analogous biochemical functions are performed by different bacterial taxa within individual hosts.
Taken together, the highly individual responses to PHGG in our study further indicate the need for parameters enabling personalized prebiotic interventions. The complex relationships between the host, its commensal microbiota and their response to a specific prebiotic are influenced by multiple variables and makes a one-fits-all prebiotic becoming available unlikely [47
]. Using our dataset, we were able to identify a steady state microbial configuration in combination with a low butyrate concentration to predict clinical response, i.e., an above-average increase in stool frequency and reduction in stool consistency, to PHGG.