Abstract
Dietary fibers are increasingly appreciated as beneficial nutritional components. However, a requisite role of gut microbiota in fiber function and the overall impact of fibers on metabolomic flux remain unclear. We herein showed enhancing effects of a soluble resistant maltodextrin (RM) on glucose homeostasis in mouse metabolic disease models. Remarkably, fecal microbiota transplantation (FMT) caused pronounced and time-dependent improvement in glucose tolerance in RM recipient mice, indicating a causal relationship between microbial remodeling and metabolic efficacy. Microbial 16S sequencing revealed transmissible taxonomic changes correlated with improved metabolism, notably enrichment of probiotics and reduction of Alistipes and Bacteroides known to associate with high fat/protein diets. Metabolomic profiling further illustrated broad changes, including enrichment of phenylpropionates and decreases in key intermediates of glucose utilization, cholesterol biosynthesis and amino acid fermentation. These studies elucidate beneficial roles of RM-dependent microbial remodeling in metabolic homeostasis and showcase prevalent health-promoting potentials of dietary fibers.
Similar content being viewed by others
Introduction
Dietary management and intervention is increasingly appreciated as a vital strategy to combat the worldwide epidemic of metabolic syndrome. One important class of beneficial food components is dietary fibers, known as plant-derived complex polysaccharides resistant to digestion by amylases and glycoamylases in the small intestine1,2,3. Whereas insoluble fibers promote colonic regularity and gastrointestinal (GI) function mainly through physical bulking effects, soluble dietary fibers, such as inulin, oligofructosaccharide and resistant maltodextrin, have shown diverse health benefits both locally in the gastrointestinal (GI) tract and systemically throughout the body1,4,5. In particular, a large body of studies using both animal models and human subjects highlight an important role of fibers in energy metabolism, serving to blunt body weight gain and improve glucose and lipid homeostasis6,7,8. Detailed knowledge of the underlying functional mechanisms is thus important to fully exploit the health benefits of dietary fibers.
Accumulating evidence underscores a functional relationship between soluble dietary fibers and gut microbiota in the large intestine9,10,11. Soluble fibers are recognized prebiotics able to enrich probiotic bacteria, most notably Lactobacillus and Bifidobacterium, that are beneficial for digestive function, mucosal integrity and immune response12,13,14. On the other hand, dietary fibers are fermented by gut bacteria, producing metabolites for energy and signaling needs15,16. The primary fermentation products of dietary fibers are short-chain fatty acids (SCFAs), mainly produced by the predominant phyla Firmicutes and Bacteriodetes3,17,18. SCFAs can provide energy for distinct tissues13,19 and several recent studies have also revealed novel mechanisms whereby SCFAs act on membrane receptors and nutrient sensors to regulate physiological processes including glucose homeostasis8,20,21,22. Together, these studies highlight important roles of gut microbiota in the metabolic regulation by fibers.
A number of fundamental questions remain concerning the functional relationship between microbiota and fiber. First, although microbiome profiling has revealed extensive correlation between fiber intake and microbial shift, a requisite role of gut microbiota for metabolic regulation by fibers has not been unequivocally established16. Second, whereas SCFAs are well established metabolites of fiber, global changes in host metabolic network remains poorly understood, thus hampering identification of additional key metabolic pathways and in-depth mechanistic understanding of fiber function2,23,24,25,26. Answering these important questions in metabolic disease models can further lead to identification of bacterial and molecular markers associated with functional fiber.
In the current study, we employed a resistant maltodextrin (RM) to address the important questions whether gut microbiota are responsible for conferring the health benefits of RM and what microbial and metabolic signatures are associated with RM-mediated improvement in mouse metabolic disease models. Combining physiological, 16S sequencing and metabolomic approaches, our work demonstrates that the beneficial effects of RM are mediated through the metabolic function of gut microbiota and reveals important microbial and metabolic markers for RM in metabolic disease.
Results
Resistant maltodextrin (RM) improves glycemic control
To address the question whether resistant maltodextrin (RM) can improve metabolism in metabolic disease models, we employed a previously described soluble RM (Fibersol®-2) manufactured through a proprietary method involving controlled enzymatic treatment of corn starch5,27. Specifically, db/db mice (Jackson Laboratory), a commonly used type 2 diabetes model lacking a functional leptin receptor and consequently developing severe obesity and glucose intolerance, were fed with regular chow diet (Purina 5001) and either regular (Ctrl) or 1% (w/v) RM drinking water. Mice treated with RM over an 8-week period showed a modest trend toward a blunted weight gain and food intake during in RM-fed mice (Supplementary Fig. 1). Importantly, RM treatment led to significantly decreased fasting glucose levels (Fig. 1a). Furthermore, oral glucose tolerance test (OGTT) showed an improved glucose control in the RM group over Ctrl (Fig. 1b). Indicative of a sensitized insulin response, insulin tolerance test (ITT) also revealed improved glucose clearance in response to insulin injection in the RM group (Fig. 1c). To further substantiate the efficacy of RM in glycemic control, we utilized diet-induced obesity (DIO) mice as a second metabolic disease model. Wild-type C57B/6J mice (Jackson Laboratory) were treated with high-fat diet, in conjunction with regular drinking water (Ctrl) or 1% RM water. Consistent with results from db/db mice, we again observed reduced weight gain over the 8-week treatment period (Supplementary Fig. 2a). Likewise, glucose control was also improved in RM-treated DIO mice (Supplementary Fig. 2, b and c). Together, in accordance with previous findings4,28, our results from two complementary mouse metabolic disease models showed that RM exerts a significant role in glucose control.
Gut microbial remodeling by RM
Dietary fibers are generally known as prebiotics, promoting gut content of probiotics such as Lactobacillus and Bifidobacterium. Previous studies have also shown remodeling of gut microbiota by RM and RM can be fermented by gut microbiota for energy and signaling functions5. To begin to investigate whether RM improves glycemic control owing to gut microbial remodeling, we conducted Illumina-based 16S rRNA sequencing of fecal DNAs from Ctrl and RM treated db/db mice shown in Fig. 1. RM treatment did not significantly alter diversity29,30 (Fig. 1d) or phylum abundance (Fig. 1e, left panel) of gut microbiota in these severely obese and diabetic mice. However, a slight increase in the Firmicutes/Bacteroidetes ratio by RM treatment (Supplementary Table 1) correlates with improved glycemic control, consistent with previous human diabetes studies31 and distinct dietary association of these phyla32 (see also below).
In comparison, a number of notable changes were observed at the genus level (Fig. 1e, right). For example, a prebiotic role of RM was substantiated as both Lactobacillus and Bifidobacterium were found to be moderately elevated in abundance (Fig. 1e, right panel; Supplementary Table 2). Importantly, Alistipes, previously shown to be closely correlated with an animal fat/protein diet32, was markedly reduced in the RM group (8.86% vs. 16.10% in Ctrl; Supplementary Table 2). Similarly, the abundance of Bacteroides, a predominant genus in Bacteroidetes, was also diminished by RM, albeit to a lesser degree (20% reduction). These bacteria are bile compatible32, thus under negative selection pressure in fiber-enriched diets. Together, decreases in Alistipes and Bacteroides and concomitant increases in Lactobacillus and Bifidobacterium indicated that RM is a functional fiber with broad microbial remodeling activities.
A causal role of gut microbiota in RM efficacy in glucose control
To directly address a causal role of gut microbial remodeling in mediating the beneficial effects of RM in glucose control, we conducted fecal microbiota transplantation (FMT) experiments33,34. Recipient db/db mice were first treated with an antibiotics cocktail to remove a vast majority of intrinsic gut bacteria (Fig. 2a, Supplementary Fig. 3 and Tables 3 & 4). Subsequently, fecal suspensions from either Ctrl or RM donor mice (Fig. 1) were administered to respective recipient mice and glucose homeostasis was monitored throughout the experimental period. No significant effects on body weight were found (Supplementary Fig. 4a). In contrast, recipient mice that received fecal materials from RM donors showed a robust and progressive improvement in glucose tolerance up to 2 months post-FMT (Fig. 2, b & c). Interestingly, the peak glucose tolerance in recipient mice appeared to be more pronounced than that exhibited by donor mice (Fig. 1). By the end of 3 months post-FMT, however, the effects have subsided. We found improved insulin tolerance also at 2 months post-FMT (Supplementary Fig. 4b), coincident with the peak OGTT improvement. Of note, the recipient mice were never exposed to RM, indicating that gut microbiota in RM donors were sufficient to confer the efficacy of RM in glucose homeostasis in recipient mice.
Transmissible microbial remodeling by RM
To gain insight into the significant improvement in glycemic control following FMT, we conducted 16S rRNA sequencing using fecal DNAs from recipient mice. Consistent with donors, RM recipients showed moderate reduction in diversity (Fig. 3a) and similar slight shifts in the abundance of Firmicutes, Bacteroidetes and Actinobacteria (Fig. 3b, left panel; Supplementary Table 5). Transmissible microbial remodeling can be more clearly characterized at the genus level (Fig. 3b, right panel; Supplementary Table 6). Specifically, among the 22 altered OTUs at the genus level, 14 were found to retain similar up- or down-regulation responses to RM as in donors, most notably Alistipes, Bacteroides and Lactobacillus (Supplementary Tables 2 & 6). Among them, 12 showed changes that were significant or approaching significant (p values between 0.05 and 0.1) in both donors and recipients, whereas the other 2 were significant in either donors or recipients (Supplementary Table 6).
We further analyzed the dynamic microbial remodeling in recipient mice during the 3-month post-FMT period (Fig. 3c, Supplementary Fig. 5 and Tables 7 & 8). Interestingly, hierarchical clustering showed that the overall microbial patterns of RM recipients segregated from those of control samples and that 1 and 2 months post-FMT were more similar than 3 months post-FMT (Fig. 3c). This pattern mirrors that of glucose tolerance as shown in Fig. 2c. Furthermore, quantitative comparison between RM and Ctrl samples in a time-resolved manner uncovered several genera, including Lactobacillus and Bacteroides, showing pronounced changes in parallel with the relative improvement in glucose tolerance over the 3-month period (Supplementary Table 8). Together, our analyses identified RM-associated persistent and transmissible microbial changes at both phylum and genus levels.
Pronounced metabolomic shift in response to RM
To extend the above microbiome analysis and obtain functional insight into the glycemic effects of RM, we conducted global metabolomic profiling, using fecal samples from both donors and recipients at 2 months post-FMT, corresponding to peak metabolic efficacy (Fig. 2). A total of 727 compounds of known identity (named biochemicals) were identified (Supplementary Table 9) and large numbers (272 and 320 for donors and recipients respectively) of metabolites were found to be significantly altered in abundance as a result of RM treatment (Fig. 4a, upper panel). Venn diagrams further demonstrated significant overlaps of such changed metabolites between donors and recipients (Fig. 4a, lower panel). In accordance, principal component analysis (PCA; Fig. 4b) and hierarchical clustering (Fig. 4c) together illustrated both similarity and variance between donor vs. recipient samples with the same treatment. For example, whereas PC1 and PC2, explaining 23.85% and 20.48% of differences respectively, mainly distinguished Ctrl vs. RM treatment, PC3 (10.71%) appeared to correlate with donor vs. recipient identity (Fig. 4b).
Next, random forest analysis was conducted to pinpoint the group of metabolites most associated with RM and also preserved in both donors and recipients (Fig. 4d). Interestingly, 4 out of the 5 top metabolites identified were phenylpropionates and hydroxyphenylpropionates, products from catabolism of the aromatic amino acids phenylalanine and tyrosine with various physiological effects2,35,36. In accordance, phenylalanine and tyrosine metabolism was among the highly ranked metabolic pathways in both donor and recipient RM mice (Supplementary Fig. 6). However, note that phenylpropionates (enriched) and hydroxyphenylpropionates (depleted) were differentially affected in RM samples relative to Ctrl (Supplementary Table 10). Furthermore, levels of several other bacterial metabolites from aromatic amino acid breakdown such as p-cresol sulfate, phenol sulfate and 3-indoxyl sulfate were also reduced in RM donors and/or recipients (Supplementary Fig. 7). This latter group of phenolic and indoxyl acids have been linked with increased risk of cardiovascular disease, inflammation and oxidative damage2,18. Our results thus indicated a role of RM in disease prevention and also suggested other sources than protein catabolism for the enrichment of phenylpropionates.
In light of the improved energy metabolism in RM mice, we next examined metabolites related to sugar and lipid metabolism. Strikingly, glucose levels were found to be markedly attenuated in both RM donors and recipients compared with controls (Fig. 5). Furthermore, several intermediate metabolites for glucose metabolism, including pyruvate (glycolysis), citrate, α-ketoglutarate, aconitate and malate (TCA cycles) were also significantly reduced, strongly suggesting diminished glucose flux and improved energy homeostasis in the diabetic db/db mice.
In addition to glucose metabolism, cholesterol metabolism was also found to be considerably improved by RM (Fig. 6). Levels of two intermediate metabolites, mevolonate and mevalonolactone, were significantly decreased in RM donor and recipient mice. In accordance with their decreases, the bacterial metabolic derivative coprostanol was also reduced in RM donors. Consistent with previous studies6, these results strongly indicated a beneficial role of RM in cholesterol control. Interestingly, coprostanol, squalene and mevalonolactone showed diminished levels in rCtrl relative to dCtrl, perhaps reflecting beneficial effects of antibiotics treatment on controlling cholesterol biosynthesis37.
Together, the above metabolomic observations revealed profound metabolic benefits of RM or RM-derived microbiota.
Discussion
Average fiber intake among Americans reaches only half of the recommended amounts1. Therefore, detailed functional and causal relationship studies will provide concrete scientific basis to raise fiber awareness and consumption. Resistant maltodextrin (RM) described herein is a soluble, non-viscous and fermentable dietary fiber, with minimal GI disturbance side effects5. In studies involving healthy human subjects, RM has been shown to improve colonic motility, fecal characteristics and probiotic (Bifidobacterium) population5.
In the current study, we investigated the metabolic efficacy of RM using both diet-induced obesity (DIO) and genetic diabetic db/db mice. Whereas its effects on body weight were moderate, with slightly more pronounced efficacy in DIO mice, we observed significant ameliorative effects of RM on glucose homeostasis in these mouse models. It is worth noting that such effects require prolonged treatment, 8 weeks in our study. In accordance with an improved glucose control as revealed by these physiological assays, metabolomic profiling showed much reduced levels in both glucose and several intermediate metabolites from glycolysis and the TCA cycle in RM donor and recipient mice. Furthermore, levels of cholesterol and several metabolites for cholesterol metabolism were also strongly reduced by RM or RM-derived microbiota, consistent with a reported hypocholesterolemic effect6,38. These complementary physiological and metabolomic studies provide strong evidence for a metabolic function of RM in pathophysiological settings and highlight the importance of sustained exposure of fiber or fiber-derived microbiota.
Fecal microbiota transplantation (FMT) has been a highly effective clinical treatment for bowl diseases and more recently a powerful method to investigate a functional relationship between gut microbiota and physiological changes17,39. Adapting the procedure to diabetic db/db mice, we showed that antibiotics-treated recipient mice displayed profound improvement in glucose tolerance following FMT in a time-dependent manner, peaking at 2 months post-FMT. In conjunction with antibiotics treatment, the transplanted gut microbiota were fully capable of mediating RM metabolic efficiency, strongly suggesting a causal relationship between transmissible microbial remodeling and glycemic control. The subsequent attenuation of glucose tolerance at 3 months post-FMT suggests dynamic microbial changes associated with antibiotics and/or transplantation procedures40, underscoring the aforementioned requirement for sustained exposure to fiber or fiber-derived microbiota.
Dietary fibers as prebiotics have been well-documented12. In particular, RM has previously been shown to be moderately bifidogenic in healthy subjects5. Likewise, we observed increased levels of Lactobacillus and Bifidobacterium in RM donor and/or recipients, further substantiating a prebiotic function of RM. Besides probiotics, we also identified numerous other changes in microbial landscape as a result of RM. At the phylum level, RM led to a slight shift toward Firmicutes and Actinobacteria at the expense of Bacteriodetes. More significant changes at the genus level were determined. Notably, Alistipes was found to be markedly repressed. In a previous study comparing effects of animal vs plant derived diets on microbiota, Alistipes was shown to be enriched in the former, protein-rich diet, consistent with a purported role in protein fermentation. The depletion of Alistipes by RM in the current study is thus in accordance with the observed effects of plant-derived fiber-rich diet32, suggesting a beneficial role of RM to suppress putrefactive protein breakdown. While consistent with previous studies showing correlation between improved glycemic control and abundance shift of these major phyla31,41, these results are on the other hand at odds with several other observations42,43,44,45,46, suggesting a highly context-dependent and holistic nature of microbial remodeling in association with physiological changes. Functional assessment, such as metabolic flux, is needed for physiological interpretation of microbial remodeling47,48,49.
Our global metabolomic profiling revealed extensive metabolic flux, as indicated by the large number of metabolites (~300) showing RM-induced changes in abundance. Among the top-ranked metabolites associated with RM were phenylpropionates and hydroxyphenylpropionates. It is unlikely that the striking enrichment of phenylpropionates (>50 fold, RM/Ctrl) was a result of protein breakdown as levels of hydroxyphenylpropionates and several other amino acid metabolites (see below) were found to be strongly reduced. Two other metabolic pathways may contribute to phenylpropionate enrichment. First, dietary polyphenols are known to be catabolized to generate phenylpropionates2,50. The corn and oats in mouse chow diet contain rich polyphenols and RM may enhance microbial degradation of the dietary polyphenols. Alternatively, propionate may also promote phenylpropionate synthesis. Propionate can be converted to phosphoenolpyruvate, which in turn serves as substrate for the Shikimate pathway in commensal bacteria to produce phenylalanine and thus ultimately phenylpropionates51,52,53. The untargeted metabolomic platform used in this study was not able to identify SCFAs due to their polarity and volatility. However, previous studies have demonstrated robust propionate/SCFA production from this RM5,54. Interestingly, propionate was recently shown to link central energy regulation with intestinal gluconeogenesis (IGN) to regulate energy metabolism8. Future studies will be important to further investigate these possibilities.
Beyond SCFAs, there are evidently extensive host-gut metabolic interactions26,55. To address the important questions regarding other metabolic changes associated with dietary fibers and their functional roles, we showed here that levels of several amino acid fermentation products including phenolic and indoxyl acids were significantly diminished in RM donor and recipient mice. This pattern is consistent with the depletion of Alistipes, previously shown to be associated with a high fat/protein diet32. These gut metabolic products are known to associate with higher disease risks18; thus, reduced levels of these bacterial byproducts suggest important health benefits of this fiber.
Combining physiological assays and metagenomic/metabolomic profiling, the current study reveals a key role of RM in glucose and cholesterol homeostasis and highlights the underlying profound microbial and metabolomic remodeling. Our study uncovers important changes in the abundance of microbial OTUs (e.g., probiotics and Alistipes) and fecal metabolites (phenylpropionates and glycolysis-TCA/cholesterol/amino acid intermediates). Future studies will focus on their functional mechanisms in RM-mediated metabolic regulation, which may ultimately lead to improved understanding and application of functional dietary fibers.
Materials and Methods
Mice and resistant maltodextrin
Animal husbandry for all the studies was carried out under IACUC guidelines and the procedures were conducted as described in an animal protocol approved by the University of Texas Health Science Center at Houston (UTHSC-H) and the University of Wisconsin at Parkside. Wild-type (WT) and db/db mice, on the C57BL/6J genetic background, were obtained from the Jackson Laboratory (#664 and #697, respectively). Verification genotyping was carried out according to Jackson Laboratory protocols by using 2x PCR master mix (GenDEPOT). Mice were routinely group-housed (2/cage for db/db mice and 2-4/cage for WT mice) in standard animal facility under a 12h/12h cycles. The resistant maltodextrin (Fibersol®-2), manufactured by Matsutani and ADM, has been previously described5,27.
Mouse treatment and body weight measurements
Six-week-old db/db mice, fed with regular chow diet (Purina 5001), were randomly grouped to receive regular drinking water (Ctrl) or 1% RM ad libitum during the experimental period. Body weight was measured weekly. For diet-induced obesity, WT mice at 6 weeks of age were fed with HFD (Research Diets D12492) until the end of the experimental protocol. Mice were randomly divided into the control (Ctrl) group fed with regular drinking water and the RM group fed with 1% RM drinking water ad libitum during the experimental period.
Food and drinking water intake
Food intake was determined by calculating the difference in food weight during 24 hr intervals. Three independent experiments were carried out to calculate the average food intake. For drinking water measurement, the volume of drinking water was measured once every two weeks. The drinking water intake was calculated from averaged volumes of drinking from four independent experiments.
Oral glucose tolerance tests
Overnight fasted db/db and DIO mice were oral gavaged with 1 g/kg glucose and glucose were measured from tail blood before and 15, 30, 60, or 120 min by using the ONETOUCH UltraMini blood glucose monitoring system (LifeScan).
Insulin tolerance tests
Following 5 hr fasting, mice were injected intraperitoneally with 1.0 U/kg insulin (Sigma) and glucose levels were measured from tail blood before at 15, 30, 60, or 120 min by using the ONETOUCH UltraMini blood glucose monitoring system (LifeScan).
Antibiotics treatment and fecal microbiota transplantation
Six-week-old db/db mice were treated with a cocktail of broad spectrum antibiotics (1 g/L ampicillin, neomycin and metronidazole and 0.5 g/L vancomycin) in drinking water for 3-4 weeks33,34. The mice were allowed 3-4 days to recover before fecal microbiota transplantation started. Fresh fecal pellets were collected from donor db/db mice after two hours in collection cages with paper liner. Subsequently, 200 mg of pellets were weighed and resuspended and homogenzied at 1:10 (w/v) in transfer buffer (0.1 M phosphate buffered saline, pH 7.0, pre-reduced with 0.05% cysteine HCL. To each recipient mice, 100 μl of homogenates were used for oral gavage. The transplantation procedure was carried out every three days, four times total for each experiment. Throughout the entire experimental period, the mice were maintained on the regular chow diet (Purina 5001).
16S rRNA gene sequencing
Bacterial 16S rRNA gene profiling was conducted by the Alkek Center for Metagenomics and Microbiome Research at Baylor College of Medicine using Illumina MiSeq platform as previously described56. Briefly, fresh fecal pellets were collected from each cage on 15-20 min intervals and immediately frozen on dry ice. Pooled aliquots were stored at −80 °C at the end of 2 hr collection periods. Microbial DNA was extracted with PowerSoil® DNA Isolation kit (MoBio) following the manufacturer’s guidelines. The 16S rDNA V4 region amplicons (single index) were produced by PCR and sequenced on the MiSeq platform (Illumina) using the 2 × 250 bp protocol yielding pair-end reads that overlap by ~247 bps. Following sequencing, raw BCL files were retrieved from the MiSeq platform and called into fastqs by Casava v1.8.3 (Illumina). The read pairs were demultiplexed based on unique molecular barcodes, filtered for PhiX using Bowtie2 and reconstituted into two fastq files for each read using standard BASH. A barcodes file was generated from a raw fastq base called previously to preserve the original barcode qualities associated per read cluster. Sequencing reads were merged (allowing 4 mismatches per ≥50 bases) and processed using USEARCH v7.0.1001 (maximum error method)57. Sequences were demultiplexed using QIIME v1.8.0 and then clustered using the UPARSE pipeline57. Operational taxonomic unit (OTU) classification was achieved by mapping the UPARSE OTU table to the SILVA database. Abundances were recovered by mapping the demultiplexed reads to the UPARSE OTUs. A custom script constructed an OTU table from the output files generated in the previous two steps. The OTU table was used to calculate alpha-diversity, beta-diversity, provide taxonomic summaries and in a variety of other analyses built into QIIME that allowed for the characterization of individual and group of samples based on alpha and beta diversity indices.
Fecal metabolomic analysis
Global, untargeted metabolomic analysis was conducted by using Metabolon UPLC-MS/MS and GC-MS platform. Briefly, fresh fecal pellets were collected as above for 16S sequencing from either donor db/db mice after 2 months of RM treatment or recipient db/db mice 2 months after transplantation and stored at -80 °C. Frozen feces, six experimental samples from each group (dCtrl, dRM, rCtrl and rRM), were lyophilized and weighed. Weight equivalents were then subjected to non-targeted metabolomic analysis platform including UPLC-MS/MS and GC/MS at Metabolon Inc.58. Identification and quantification of named metabolites were conducted based on previously published methods59. The metabolomic data were then analyzed by unsupervised principal component analysis to identify sets of patterns corresponding to uncorrelated variables. Such patterns, called principle components, can reveal metabolic distinction and similarity as a function of donor/recipient status and treatment. In the random forest analysis60 to identify biochemicals made the largest contribution to the classification of RM vs. Ctrl samples (donor and recipient combined), Mean Decrease Accuracy (MDA) was determined by randomly permuting a metabolite, running the observed values through a series of decision trees and then reassessing the prediction accuracy. A predictive accuracy of 50% would be expected by chance and a greater MDA score indicates stronger differentiating power.
Statistical analyses
All data are presented as means ± SEM. Statistical significance was determined by one-way ANOVA (Dunnett’s test), two-way ANOVA repeated measures (Bonferroni’s test) and Mann-Whitney test. P < 0.05 was accepted as statistically significant. Statistical analyses were performed using the SigmaStat3.5 software (for ANOVA) and the Wilcox.test program of R (for Mann-Whitney). For metabolic studies, the N numbers refer to mouse numbers. For 16S sequencing and metabolomic profiling, the N numbers refer to independent fecal samples.
Additional Information
How to cite this article: He, B. et al. Transmissible microbial and metabolomic remodeling by soluble dietary fiber improves metabolic homeostasis. Sci. Rep. 5, 10604; doi: 10.1038/srep10604 (2015).
References
Anderson, J. W. et al. Health benefits of dietary fiber. Nutr Rev 67, 188–205, 10.1111/j.1753-4887.2009.00189.x (2009).
Nicholson, J. K. et al. Host-gut microbiota metabolic interactions. Science 336, 1262–1267, 10.1126/science.1223813 (2012).
Flint, H. J., Scott, K. P., Louis, P. & Duncan, S. H. The role of the gut microbiota in nutrition and health. Nat Rev Gastroenterol Hepatol 9, 577–589, 10.1038/nrgastro.2012.156 (2012).
Kishimoto, Y., Oga, H., Tagami, H., Okuma, K. & Gordon, D. T. Suppressive effect of resistant maltodextrin on postprandial blood triacylglycerol elevation. Eur J Nutr 46, 133–138, 10.1007/s00394-007-0643-1 (2007).
Fastinger, N. D. et al. A novel resistant maltodextrin alters gastrointestinal tolerance factors, fecal characteristics and fecal microbiota in healthy adult humans. J Am Coll Nutr 27, 356–366 (2008).
Brown, L., Rosner, B., Willett, W. W. & Sacks, F. M. Cholesterol-lowering effects of dietary fiber: a meta-analysis. Am J Clin Nutr 69, 30–42 (1999).
Chandalia, M. et al. Beneficial effects of high dietary fiber intake in patients with type 2 diabetes mellitus. N Engl J Med 342, 1392–1398, 10.1056/NEJM200005113421903 (2000).
De Vadder, F. et al. Microbiota-generated metabolites promote metabolic benefits via gut-brain neural circuits. Cell 156, 84–96, 10.1016/j.cell.2013.12.016 (2014).
Cani, P. D. et al. Gut microbiota fermentation of prebiotics increases satietogenic and incretin gut peptide production with consequences for appetite sensation and glucose response after a meal. Am J Clin Nutr 90, 1236–1243, 10.3945/ajcn.2009.28095 (2009).
Everard, A. et al. Responses of gut microbiota and glucose and lipid metabolism to prebiotics in genetic obese and diet-induced leptin-resistant mice. Diabetes 60, 2775–2786, 10.2337/db11-0227 (2011).
Walsh, C. J., Guinane, C. M., O’Toole, P. W. & Cotter, P. D. Beneficial modulation of the gut microbiota. FEBS Lett, 10.1016/j.febslet.2014.03.035 (2014).
Saulnier, D. M., Spinler, J. K., Gibson, G. R. & Versalovic, J. Mechanisms of probiosis and prebiosis: considerations for enhanced functional foods. Curr Opin Biotechnol 20, 135–141, 10.1016/j.copbio.2009.01.002 (2009).
Tuohy, K. M., Rouzaud, G. C., Bruck, W. M. & Gibson, G. R. Modulation of the human gut microflora towards improved health using prebiotics--assessment of efficacy. Curr Pharm Des 11, 75–90 (2005).
Scott, K. P., Duncan, S. H., Louis, P. & Flint, H. J. Nutritional influences on the gut microbiota and the consequences for gastrointestinal health. Biochem Soc Trans 39, 1073–1078, 10.1042/BST0391073 (2011).
Koropatkin, N. M., Cameron, E. A. & Martens, E. C. How glycan metabolism shapes the human gut microbiota. Nat Rev Microbiol 10, 323–335, 10.1038/nrmicro2746 (2012).
Shen, J., Obin, M. S. & Zhao, L. The gut microbiota, obesity and insulin resistance. Mol Aspects Med 34, 39–58, 10.1016/j.mam.2012.11.001 (2013).
Ridaura, V. K. et al. Gut microbiota from twins discordant for obesity modulate metabolism in mice. Science 341, 1241214, 10.1126/science.1241214 (2013).
Macfarlane, S. & Macfarlane, G. T. Regulation of short-chain fatty acid production. Proc Nutr Soc 62, 67–72, 10.1079/PNS2002207 (2003).
Donohoe, D. R. et al. The microbiome and butyrate regulate energy metabolism and autophagy in the mammalian colon. Cell Metab 13, 517–526, 10.1016/j.cmet.2011.02.018 (2011).
Maslowski, K. M. et al. Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43. Nature 461, 1282–1286, 10.1038/nature08530 (2009).
Trompette, A. et al. Gut microbiota metabolism of dietary fiber influences allergic airway disease and hematopoiesis. Nat Med 20, 159–166, 10.1038/nm.3444 (2014).
Frost, G. et al. The short-chain fatty acid acetate reduces appetite via a central homeostatic mechanism. Nat Commun 5, 3611, 10.1038/ncomms4611 (2014).
Russell, W. R., Hoyles, L., Flint, H. J. & Dumas, M. E. Colonic bacterial metabolites and human health. Curr Opin Microbiol 16, 246–254, 10.1016/j.mib.2013.07.002 (2013).
Jacobs, D. M., Gaudier, E., van Duynhoven, J. & Vaughan, E. E. Non-digestible food ingredients, colonic microbiota and the impact on gut health and immunity: a role for metabolomics. Curr Drug Metab 10, 41–54 (2009).
Polakof, S. et al. Resistant starch intake partly restores metabolic and inflammatory alterations in the liver of high-fat-diet-fed rats. J Nutr Biochem 24, 1920–1930, 10.1016/j.jnutbio.2013.05.008 (2013).
Martin, F. P. et al. Dietary modulation of gut functional ecology studied by fecal metabonomics. J Proteome Res 9, 5284–5295, 10.1021/pr100554m (2010).
Baer, D. J. et al. The metabolizable energy of dietary resistant maltodextrin is variable and alters fecal microbiota composition in adult men. J Nutr 144, 1023–1029, 10.3945/jn.113.185298 (2014).
Delzenne, N. M., Cani, P. D. & Neyrinck, A. M. Modulation of glucagon-like peptide 1 and energy metabolism by inulin and oligofructose: experimental data. J Nutr 137, 2547S–2551S (2007).
Cotillard, A. et al. Dietary intervention impact on gut microbial gene richness. Nature 500, 585–588, 10.1038/nature12480 (2013).
Le Chatelier, E. et al. Richness of human gut microbiome correlates with metabolic markers. Nature 500, 541–546, 10.1038/nature12506 (2013).
Larsen, N. et al. Gut microbiota in human adults with type 2 diabetes differs from non-diabetic adults. PLoS One 5, e9085, 10.1371/journal.pone.0009085 (2010).
David, L. A. et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature 505, 559–563, 10.1038/nature12820 (2014).
Willing, B. P., Vacharaksa, A., Croxen, M., Thanachayanont, T. & Finlay, B. B. Altering host resistance to infections through microbial transplantation. PLoS One 6, e26988, 10.1371/journal.pone.0026988 (2011).
Ghosh, S. et al. Colonic microbiota alters host susceptibility to infectious colitis by modulating inflammation, redox status and ion transporter gene expression. Am J Physiol Gastrointest Liver Physiol 301, G39–49, 10.1152/ajpgi.00509.2010 (2011).
Smith, E. A. & Macfarlane, G. T. Formation of Phenolic and Indolic Compounds by Anaerobic Bacteria in the Human Large Intestine. Microbial ecology 33, 180–188 (1997).
Russell, W. R. et al. Major phenylpropanoid-derived metabolites in the human gut can arise from microbial fermentation of protein. Molecular nutrition & food research 57, 523–535, 10.1002/mnfr.201200594 (2013).
Hashimoto, S. et al. Effects of beta-lactam antibiotics on intestinal microflora and bile acid metabolism in rats. Lipids 31, 601–609 (1996).
Kishimoto, Y., Wakabayashi, S. & Takeda, H. Hypocholesterolemic effect of dietary fiber: relation to intestinal fermentation and bile acid excretion. J Nutr Sci Vitaminol (Tokyo) 41, 151–161 (1995).
Mayer, E. A., Savidge, T. & Shulman, R. J. Brain-gut microbiome interactions and functional bowel disorders. Gastroenterology 146, 1500–1512, 10.1053/j.gastro.2014.02.037 (2014).
Perez-Cobas, A. E. et al. Gut microbiota disturbance during antibiotic therapy: a multi-omic approach. Gut 62, 1591–1601, 10.1136/gutjnl-2012-303184 (2013).
Qin, J. et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature 490, 55–60, 10.1038/nature11450 (2012).
Zhang, X. et al. Human gut microbiota changes reveal the progression of glucose intolerance. PLoS One 8, e71108, 10.1371/journal.pone.0071108 (2013).
Wu, X. et al. Molecular characterisation of the faecal microbiota in patients with type II diabetes. Curr Microbiol 61, 69–78, 10.1007/s00284-010-9582-9 (2010).
Schwiertz, A. et al. Microbiota and SCFA in lean and overweight healthy subjects. Obesity (Silver Spring) 18, 190–195, 10.1038/oby.2009.167 (2010).
Turnbaugh, P. J. et al. A core gut microbiome in obese and lean twins. Nature 457, 480–484, 10.1038/nature07540 (2009).
Karlsson, F. H. et al. Gut metagenome in European women with normal, impaired and diabetic glucose control. Nature 498, 99–103, 10.1038/nature12198 (2013).
Holmes, E., Li, J. V., Marchesi, J. R. & Nicholson, J. K. Gut microbiota composition and activity in relation to host metabolic phenotype and disease risk. Cell Metab 16, 559–564, 10.1016/j.cmet.2012.10.007 (2012).
Tremaroli, V. & Backhed, F. Functional interactions between the gut microbiota and host metabolism. Nature 489, 242–249, 10.1038/nature11552 (2012).
McHardy, I. H. et al. Integrative analysis of the microbiome and metabolome of the human intestinal mucosal surface reveals exquisite inter-relationships. Microbiome 1, 17, 10.1186/2049-2618-1-17 (2013).
Jenner, A. M., Rafter, J. & Halliwell, B. Human fecal water content of phenolics: the extent of colonic exposure to aromatic compounds. Free radical biology & medicine 38, 763–772, 10.1016/j.freeradbiomed.2004.11.020 (2005).
Metges, C. C. & Petzke, K. J. Utilization of essential amino acids synthesized in the intestinal microbiota of monogastric mammals. The British journal of nutrition 94, 621–622 (2005).
Ratia, K. et al. Discovery of selective inhibitors of the Clostridium difficile dehydroquinate dehydratase. PLoS One 9, e89356, 10.1371/journal.pone.0089356 (2014).
Zhou, J., Bowler, L. D. & Spratt, B. G. Interspecies recombination and phylogenetic distortions, within the glutamine synthetase and shikimate dehydrogenase genes of Neisseria meningitidis and commensal Neisseria species. Molecular microbiology 23, 799–812 (1997).
Flickinger, E. A. et al. Glucose-based oligosaccharides exhibit different in vitro fermentation patterns and affect in vivo apparent nutrient digestibility and microbial populations in dogs. J Nutr 130, 1267–1273 (2000).
Backhed, F., Ley, R. E., Sonnenburg, J. L., Peterson, D. A. & Gordon, J. I. Host-bacterial mutualism in the human intestine. Science 307, 1915–1920, 10.1126/science.1104816 (2005).
Caporaso, J. G. et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. The ISME journal 6, 1621–1624, 10.1038/ismej.2012.8 (2012).
Edgar, R. C. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nature methods 10, 996–998, 10.1038/nmeth.2604 (2013).
Evans, A. M., DeHaven, C. D., Barrett, T., Mitchell, M. & Milgram, E. Integrated, nontargeted ultrahigh performance liquid chromatography/electrospray ionization tandem mass spectrometry platform for the identification and relative quantification of the small-molecule complement of biological systems. Anal Chem 81, 6656–6667, 10.1021/ac901536h (2009).
Dehaven, C. D., Evans, A. M., Dai, H. & Lawton, K. A. Organization of GC/MS and LC/MS metabolomics data into chemical libraries. J Cheminform 2, 9, 10.1186/1758-2946-2-9 (2010).
Touw, W. G. et al. Data mining in the Life Sciences with Random Forest: a walk in the park or lost in the jungle? Briefings in bioinformatics 14, 315–326, 10.1093/bib/bbs034 (2013).
Acknowledgements
We greatly appreciate Dirk Reif (ADM Inc.), Satoshi Koikeda (Amano Enzyme Inc.), Joseph T. Bass (Northwestern U.), Edward A. Fisher (NYU) and Lenard M. Lichtenberger (UT Houston) for helpful discussions. This work is supported by the Robert A. Welch Foundation (AU-1731), AHA (11SDG7600045) and NIA (R01, AG045828) to Z.C., TMC Digestive Diseases Center (Pilot Awards from NIDDK Center Grant 2P30-DK056338) and Matsutani Chemical Industry Co. to Z.C. and S.-H.Y.
Author information
Authors and Affiliations
Contributions
Z.C., K.S. and S.-H.Y. designed research; B.H., K.N., N.J.A. and S.H. L.-O. performed research; Z.C., K.S., S.-H.Y. and B.H. wrote the paper; and all authors analyzed data and reviewed manuscript.
Electronic supplementary material
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
He, B., Nohara, K., Ajami, N. et al. Transmissible microbial and metabolomic remodeling by soluble dietary fiber improves metabolic homeostasis. Sci Rep 5, 10604 (2015). https://doi.org/10.1038/srep10604
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep10604
This article is cited by
-
Oligosaccharides from agar extends lifespan through activation of unfolded protein response via SIR-2.1 in Caenorhabditis elegans
European Journal of Nutrition (2022)
-
Influence of adding rice bran on physio-chemical and sensory properties of bread
Journal of Food Measurement and Characterization (2021)
-
Antibiotic-modulated microbiome suppresses lethal inflammation and prolongs lifespan in Treg-deficient mice
Microbiome (2019)
-
Dietary short-chain fatty acid intake improves the hepatic metabolic condition via FFAR3
Scientific Reports (2019)
-
Chitin–glucan and pomegranate polyphenols improve endothelial dysfunction
Scientific Reports (2019)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.