Modulations in the offspring gut microbiome are refractory to postnatal synbiotic supplementation among juvenile primates
We and others have previously shown that alterations in the mammalian gut microbiome are associated with diet, notably early life exposure to a maternal high fat diet (HFD). Here, we aimed to further these studies by examining alterations in the gut microbiome of juvenile Japanese macaques (Macaca fuscata) that were exposed to a maternal HFD, weaned onto a control diet, and later supplemented with a synbiotic comprised of psyllium seed and Enterococcus and Lactobacillus species.
Eighteen month old offspring (n = 7) of 36% HFD fed dams were fed a control (14% fat) diet post weaning, then were synbiotic supplemented for 75 days and longitudinal stool and serum samples were obtained. All stool samples were subjected to 16S rRNA metagenomic sequencing, and microbiome profiles and serum lipids and triglycerides were compared to untreated, healthy age matched and diet matched controls (n = 7). Overall, 16S-based metagenomic analysis revealed that supplementation exerted minimal alterations to the gut microbiome including transient increased abundance of Lactobacillus species and decreased abundance of few bacterial genera, including Faecalibacterium and Anaerovibrio. However, serum lipid analysis revealed significant decreases in triglycerides, cholesterol, and LDL (p < 0.05). Nevertheless, supplemented juveniles challenged 4 months later were not protected from HFD-induced gut dysbiosis.
Synbiotic supplementation is temporally associated with alterations in the gut microbiome and host lipid profiles of juvenile Japanese macaques that were previously exposed to a maternal HFD. Despite these presumptive temporal benefits, a protective effect against later HFD-challenge gut dysbiosis was not observed.
KeywordsProbiotic, synbiotic Non-human primate Microbiome Dysbiosis
Non-alcoholic fatty liver disease
Operational Taxonomic Unit
Short-chain fatty acid
Initial studies of the human microbiome characterized the diversity of the healthy gut microbiome, and subsequent analyses have revealed that the gut microbiome is significantly altered in association with obesity, diet, and metabolic disease [1, 2, 3, 4, 5, 6, 7, 8, 9]. Similar trends have been observed in animal model systems [10, 11, 12, 13, 14, 15, 16, 17], including Japanese macaques (Macaca fuscata) . We have previously demonstrated that the gut microbiome of the Japanese macaque is structured similarly to the human gut microbiome and is modulated primarily by diet, which provides an animal model for examining the human microbiome . However, while diet is associated with significant alterations in the gut microbiome [4, 5, 6, 7, 8, 9, 12, 13, 14, 15, 16, 17, 19], duration and modulation of the offspring gut dysbiosis associated with maternal high-fat diet (HFD) exposure is relatively unexplored. Similarly, while no human studies to date have shown causation of metabolic disease by the gut microbiome, there have been numerous studies associating dysbiosis of the gut microbiome with metabolic disease [2, 3, 11, 12, 20, 21, 22, 23]. We have previously shown that exposure to a maternal HFD during gestation and lactation is associated with impaired islet vascularization, increases in hepatic lipids, and dysbiosis of the offspring gut microbiome at 1 year of age [18, 24, 25]. Altogether, this model provides a unique opportunity to closely examine interactions between dietary exposures, metabolic disease, and the microbiome.
Probiotics are any viable microorganisms (e.g. Lactobacilli and Bifidobacteria) that when consumed or applied confer beneficial effects to the host [26, 27]. Prebiotics are non-host digestible food substances that act as an energy source for beneficial microorganisms in the gut and encourage microbial colonization, growth, and activity . Synbiotics are a combination of the two. These classes of food products are largely accepted as a non-harmful means to promote metabolic health . Previous studies have indicated that prebiotic, probiotic, and synbiotic supplemental use are associated with alterations in physiologic and metabolic measures of host and/or microbe, bowel function, mucosal contact time, and microbial abundances [28, 29, 30, 31, 32, 33, 34, 35, 36]. More specifically, a study with rats demonstrated decreases in weight gain, blood glucose and insulin, and liver triglycerides associated with synbiotic supplementation . Inherent to this notion of benefit is the assumption that the gut is colonized by these probiotic strains and species, although evidence demonstrating either transient or persistent colonization is lacking.
We hypothesized that synbiotic supplementation may modulate the gut microbiome of offspring exposed to a maternal HFD in our non-human primate model (NHP) [18, 37, 38, 39]. To test this hypothesis, we utilized a synbiotic composed of probiotic species of Enterococcus and Lactobacillus and a prebiotic of psyllium seed. Psyllium seed (Plantago ovata), is composed mainly of a highly branched arabinoxylan , and is a prevalent prebiotic [35, 41, 42]. Psyllium seed is not absorbed into the bloodstream; instead, it passes through the gastrointestinal tract and binds with water to increase stool bulk. Thus, psyllium seed is used to treat diarrhea and diarrhea-like symptoms by absorbing and retaining water in order to make stool more formed and less fluid [43, 44]. In concordance, psyllium seed has been used clinically with diarrhea, irritable bowel syndrome, and inflammatory bowel diseases (e.g. ulcerative colitis and Crohn’s disease; reviewed in [45, 46]). Here, we aimed to determine in our NHP model if synbiotic use leads to transient alterations in the gut microbiome and serum lipid profiles [37, 38, 39]. Additionally, and in accordance with our NHP maternal model of obesity, we aimed to examine if synbiotic treatment modulated dysbiosis of the gut associated with a short term HFD challenge at 2 years of age and following 18 months of control diet feeding.
In this opportunistic study, one socially housed troop of juvenile offspring (17–20 months of age) was observed to have intermittent non-infectious loose stools (culture negative, absence of clinical disease including absence of hair and weight loss, social irritability, or appetite loss) but did not experience any weight loss during the duration of the study (Additional file 1: Figure S1). Within 7 days the animals were supplemented with synbiotic therapy. The synbiotic formulation was administered per os over the course of 75 days hidden in daily treats (consisting of banana, yogurt, pumpkin puree, oats, cinnamon, water, honey, gelatin, probiotic, and prebiotic; Fig. 1). The probiotic component of the synbiotic consisted of 5.0 × 106–1.0 × 107 CFUs of Enterococcus faecium (SF-273), Lactobacillus acidophilus (LC-322), Lactobacillus plantarum (LA-5), and Lactobacillus casei (CH6072) (Probiocin oral pet gel, Probios). The prebiotic component consisted of 1.0–1.5 g of psyllium seed powder (EquiAid Natural Psyllium Fiber, Phoenix, AZ; Fig. 1). Around 27 months of age, and post-supplementation, juveniles were challenged with a HFD for 36 days. Two juvenile cohorts that were untreated, healthy, age and dietary exposure matched were used for comparative analyses, including one sample set corresponding to the pre-supplementation time point (n = 3, two males and one female, 15 months old) and another sample set during a HFD challenge time point (n = 4, all females, 25–27 months old). Animals utilized in this study were not treated with antibiotics at any point during the course of the study.
Blood (serum lipid) analysis
For triglycerides, cholesterol, LDL, and HDL levels, blood samples were taken prior to the start of the synbiotic supplementation and at one-month post-supplementation and submitted for serum lipid profile analysis. Whole-blood samples were collected in red top Vacutainer collection tubes. Serum was separated from whole blood by centrifugation at 2400 rpm for 6 min (Eppendorf Centrifuge5810, Westbury, New York). Biochemistry and lipid profiles were performed on an ABXPentra400 (Horiba, Irvine, CA) in accordance with manufacturer’s guidelines. Quality control procedures were performed each morning prior to analysis with commercial QC materials and samples were run only when the control runs passed inspection. An aliquot of 300 μl was analyzed for cholesterol, triglyceride, high- and low-density lipoproteins. Statistical significance was determined using the Wilcoxon matched-pairs signed rank test (non-parametric).
Sequencing of the microbiome
When animals were individually caged for procedures, samples were collected both as stool and anal collections via cotton swabbing at several time points, including prior to the start of the synbiotic supplementation (17–20 months old, anus), midway through supplementation (stool), and post-supplementation [corresponding to one (stool), two (anus, stool), and 4 months (stool)] and during a HFD challenge that occurred 5 months after treatment (stool, 26–28 months old; Fig. 1, Additional file 2: Table S1). Additionally, samples were collected from the healthy, age and dietary matched, offspring at separate time points: 15 months of age (anal and stool swabs) and during a HFD challenge (anal swabs, 25–27 months old; Fig. 1, Additional file 2: Table S1). Microbial DNA was isolated via the PowerSoil DNA Isolation Kit (MOBIO, Carlsbad, CA). After DNA extraction, microbial 16SrDNA was amplified using V3 V5 primers with barcodes attached for sample identification before Roche454 sequencing [18, 51]. The V3 V5 primer sequences used include, with barcodes in lower case letters: 5’-cctatcccctgtgtgccttggcagtctcaGCCTACGGGAGGCAGCAG-3′ (B-354F forward primer) and 5’-ccatctcatccctgcgtgtctccgactcagNNNNNCCGTCAATTCMTTTRAGT-3′ (A-926R reverse primer).
16S data analysis
Sample reads were analyzed using quantitative insights into microbial ecology (QIIME, version 1.9.1) . Briefly, reads lacking a barcode or primer sequence, reads with length less than 200 nucleotides, and reads with a minimum average quality score less than 25 were removed. De-noised sequences were clustered into operational taxonomic units (OTUs) at the 97% similarity level through closed-reference OTU picking using the Greengenes13.5 database , resulting in 73% of sequences (316,470/432,779) remaining for subsequent analysis (Additional file 2: Table S1). Microbial diversity was evaluated within samples (alpha diversity) or between samples (beta diversity) on rarefied OTU tables. Rarefaction resulted in 2969 and 3125 sequences from each sample used in the analyses of the intervals containing the synbiotic supplementation and HFD challenge, respectively. Alpha diversity was evaluated using Shannon Diversity Index and chao1. Beta diversity was evaluated using phylogenetic distance (weighted and unweighted UniFrac) . The resulting distance matrices served as inputs for principal coordinates analysis (PCoA) and significance of sample clustering by PERMANOVA with 999 permutations. Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt)  was used to infer relative abundance of bacterial encoded metabolic pathways, as defined by the Kyoto Encyclopedia of Genes and Genomes (KEGG;http://www.genome.jp/kegg/). Significant differences in bacterial taxa and metabolic pathways between groups was identified via STAMP (White’s non-parametric t-test for two groups comparisons)  and feature selection through linear discriminant analysis (LDA) effect size (LEfSe) . For LEfSe, the alpha value for the factorial Kruskal–Wallis test was 0.05 and the threshold on the logarithmic linear discriminant analysis (LDA) score for discriminative features was 2.0. Weight and lipid data were analyzed using Prism 6 (GraphPad Software Inc La Jolla, CA). The 16S sequence data generated from this analysis have been deposited in the Sequence Read Archive (SRA) under accession bioproject ID PRJNA317339.
Probiotic species-specific PCR
Primers were selected from previously reported studies [58, 59] or designed using Primer-BLAST from the National Center for Biotechnology Information (NCBI) . Primer sequences and annealing conditions are included in Additional file 3: Table S2. Resultant amplicons were run on a 2% agarose gel and visualized using a GelDoc-it  Imager (UVP, LLC Upland, CA). For quantitative real-time PCR (qPCR), the FastSYBRGreen kit (Applied Biosystems ABI, Waltham, MA) was utilized, and the reaction was run on a StepOnePlus PCR System (ABI). The following strains of the probiotic species were used as positive controls for PCR - L. casei SD5868, L. acidophilus ATCC521, and L. plantarum ATCC14917.
Alterations of the juvenile gut microbiome are associated with synbiotic supplementation
Synbiotic supplementation is temporally associated with improvements in serum lipids and associated bacterial metabolic pathways
Synbiotic treatment does not confer protection from HFD-associated gut dysbiosis following later challenge
Overall, we found that there were observable and a few significant alterations in the gut microbial community structure and function (e.g., gut ecology) in temporal association with synbiotic use. These microbiome rearrangements associated with synbiotic supplementation included several shifts in taxonomic relative abundances during synbiotic supplementation which have been associated with host biology. For example, we saw a decrease in Faecalibacterium, and F. prausnitzii has been associated with anti-inflammatory properties in Crohn’s disease and in animal models of colitis [63, 64]. Similarly, we observed a decrease in the relative abundance of Phascolarctobacterium, which has been shown increase in relative abundance in rats that received a HFD compared those on control diet . Lastly, we saw a decrease in Anaerovibrio, which is in agreement with a previous study in Surti buffalo associating gut ecology and dietary fiber [61, 65].
Based on the limited but observable and significant taxonomic shifts observed during synbiotic supplementation, it is not surprising that we observed statistically significant concurrent decreases in cholesterol, triglyceride, and LDL levels (Fig. 4). The results of our lipid profiles prompted us to examine the inferred metabolic pathways of the gut microbiome in greater detail, and provide microbiome community function data which correlates well with our serum lipid measurements. Consistent with both our microbiota and host lipid data, we found a decrease in fatty acid metabolism pathway gene content in association with synbiotic exposure (Fig. 5). These data are in concordance with prior studies in rats and humans that have found synbiotic consumption leads to decreased levels of serum cholesterol and triglycerides [66, 67]. However, while we were unable to determine if there was a higher abundance of catabolic versus anabolic pathways, our prior work in this model has shown that lipid metabolism pathways are positively correlated to bacteria, including Campylobacter and Runimococcus . In the current study, we observed enrichment for sphingolipid metabolism and glycosphingolipid biosynthesis-globo series pathways during synbiotic supplementation. Interestingly, this metabolic pathway data is consistent with previous work of others demonstrating that treatment with a Bacteroides derived glycosphingolipid in gnotobiotic mice regulates colonic invariant natural killer T cell (iNKT) homeostasis to protect against oxazolone induced colitis . Taken together, our findings and those of others suggest that synbiotic supplementation may act as an ecological modulator of the gut microbial community, rather than via direct xenobiosis.
Despite these presumptive beneficial findings during and immediately post synbiotic supplementation, we failed to observe persistent significant metabolic protection from gut dysbiosis following later HFD challenge. However, we did observe several changes in bacterial taxa at the genus level between synbiotic treated animals to untreated animals during HFD challenge which are worth commenting on. The lower relative abundance of Fibrobacter is likely related to the HFD, as F. succinogenes is a well-known degrader of cellulose , and cellulose is a principal component of the Japanese macaque diet. The remaining genera that were observed to have a lower relative abundance in synbiotic supplemented animals have been previously associated with various forms of gastrointestinal distress. Brachyspira are associated with intestinal spirochetosis and diarrhea , and Sphaerochaeta, related to Brachyspira, have been shown to have highly chimeric genomes, sharing more than 40 % of the total number of its genes with Clostridia . Flexispira, reclassified as members of Helicobacter , are associated with colitis in humans . The bacteria with increased relative abundance in the gut microbiome of synbiotic-supplemented juveniles include Bulleidia, Eubacterium, and Streptococcus. Bulleidia extructa is an obligate anaerobe that has been described in association with periodontal disease . While B. extructa is described as an oral pathogen, it is related to Holdemania filiformis and Erysipelothrix rhusiopathiae, which are both found in the stool . Therefore, it’s possible that a species of Bulleidia also resides in the gut, which is supported by a study examining the microbiome of the appendix . In this study, the relative abundance of Bulleidia was increased in subjects with appendicitis . While Bulleidia has been associated with disease, it was found in the appendix and rectum of healthy subjects although at a low relative abundance . Eubacterium is also associated with host disease and has been found to be increased in coronary heart disease subjects with metabolic syndrome . While Eubacterium are one of the most abundant group of human colonic butyrate producers , we did not observe a significant difference in the predicted pathway for butyrate metabolism during the HFD challenge. Streptococcus species have been implicated in both detrimental and beneficial aspects of non-alcoholic fatty liver disease (NAFLD) [67, 77], and we have previously extensively described that both juveniles and fetuses exposed to a maternal HFD have NAFLD in our Japanese macaque model [25, 38, 47, 78]. While a strain of S. mutans has been shown to aggravate NAFLD in mice , S. thermophilus in a probiotic cocktail mixture that included Lactobacillus and Bifidobacterium species was shown to ameliorate NAFLD induced by a HFD in rats . While our synbiotic cocktail did not include Streptococcus species, we speculate that our findings are consistent with the host having multiple body site-specific niches where one microbe supplementation (or loss) subsequently disrupts the relative abundance of another [1, 51, 79, 80, 81].
Accompanying these alterations in several taxa occurring in association with the HFD challenge were modulations in their inferred bacterial metabolic pathways. For example, expression of a glyoxylate shunt in mice resulted in resistance to the associated effects of HFD consumption, such as improved fat mass, leptin, and plasma triglycerides . While the pathway for glyoxylate metabolism was not enriched in the supplemented animals compared to non-supplemented animals, it was increased in response to the HFD challenge in both groups (Fig. 7c). In further agreement, a prior study with rats provided a high prebiotic fiber diet (derived from inulin and oligofructose) post-weaning were shown to have a dampened response to the negative effects associated with HFD consumption (e.g., lower ratio of Firmicutes to Bacteroidetes and decreased cholesterol) . These results were considered to be due to the fiber affecting alterations in the gut microbiota and hepatic gene expression . Taken together, our observations lead us to speculate that the altered carbohydrate metabolic pathways are most likely attributable to the establishment of a microbial community during the synbiotic supplementation that more readily produces short chain fatty acids (SCFAs) . Although we did not measure SCFA production directly, it is well-documented that psyllium alters SCFA production  and thereby lipid metabolism , leading to an altered response of the microbiome in association with the HFD challenge in the animals that received prior synbiotic supplementation. While we are able to speculate on potential synbiotic and HFD gut taxonomic and functional pathway associations as identified by metagenomic sequencing, future in-depth studies are needed.
There are several limitations to our study. First, it was a convenience study precipitated by the use of probiotics for loose and non-infectious stools. As such, it is limited by small sample size but benefits from pair-wise study design. We observed significant differences in clinically relevant measures during synbiotic supplementation, suggesting our small sample size was nevertheless adequately powered. Similarly, following high fat diet challenge, we observed significant differences in chao 1 but not Shannon diversity measures. Given observed significance in some but not all measures, our findings most likely reflect the underlying biology and support the notion that synbiotic supplementation is not associated with persistent changes in the gut ecology. Second, we could not detect the probiotic strains in post-supplementation stool samples. However, this inability to detect the bacterial strains in the synbiotic per se are both consistent with and in contrast to previous human studies [29, 31, 34, 35, 41], and may be due to several reasons, such as synbiotic dose, probiotic species, and model system. For example, prior studies in humans have utilized Bifidobacterium [29, 31, 34, 35, 41], which differs from the probiotics used in this study. Thus, the inability to detect strains or species may be due to the absence of Bifidobacteria in the synbiotic supplement. However, we were able to detect alterations in Lactobacillus at the genus level in supplemented juveniles (Additional file 1: Figures S4 and S5). This lack of detection at the species level is consistent with other studies, and may be due to the complexity of the sample, the probiotic dosage, and/or model system. A prior study in canines where bacterial DNA from a probiotic was detected by quantitative real-time PCR (qPCR) provided daily supplementation at > 108 CFU . In contrast and consistent with our findings, a human study employing culturable strains similarly could only detect later stool output by qPCR at the genus level . More recently, a study in preterm infants was able to detect probiotic strain DNA by qPCR but only in neonatal stool , which is non-complex and harbors similar strain microbes found within the NICU . Therefore, the ability to detect probiotic bacterial DNA in the stool of preterm infants may be due a lower complexity of the sample or shared built environments.
Despite these limitations, our study has several strengths. First, it is a “real world” probiotic evaluation, whereby initial use followed common indications in both veterinary and human populations (namely, loose stools). Second, despite our inability to detect our probiotics strains (but demonstrated ability to detect same-strain genus increased abundance), we have observed significant improvement in lipid profiles concomitant with microbiome rearrangement and prokaryotic pathway enrichment unique to our synbiotic supplemented animals and only during supplementation. Thus, our data suggest that the probiotic mechanisms are likely through temporal ecological modulation rather than direct effects. Third, we leveraged an exceedingly well characterized primate model and were thus able to control for many potential confounders.
Collectively, our findings in our primate model of maternal HFD feeding indicate that synbiotic supplementation is temporally associated with minimal and a few significant alterations in the gut microbiome of juvenile offspring. Overall, the majority of microbiome alterations in both beta and alpha diversity were not significant. However, synbiotic supplementation was affiliated with significant changes in the enrichment of specific carbohydrate pathways that may be tied to microbial production of SCFAs and host lipid profiles associated with both a maternal HFD and later in life HFD challenge . Nevertheless, a protective effect against further gut dysbiosis associated with a HFD challenge at 2 years of age was not observed. Altogether, our data suggest that while synbiotic supplementation is associated with a few significant modulations of the microbiome and functional benefits to host physiology, this synbiotic-altered gut ecology does not persist and is refractory to later HFD feeding. These findings further support our prior observations that HFD exposure, be it solely maternal or as a later in life challenge, remains a significant driver of alterations of the gut microbiome. This lasting footprint of the maternal diet persists, and cannot be beneficially modified in the two years postnatally by neither synbiotic supplementation nor control diet feeding from the time of weaning onward.
The authors also wish to thank J. Burton for providing materials and reagents, and P. Ganesh and M. Seferovic, for assistance in performing experiments.
This research was carried out while R.M.P. was supported through NIH K12 GM084897, A.L.P. was supported through NICHD F32 HD082969, B.D.W.B. was supported through the Cancer Prevention Research Institute of Texas, A.S.H. was supported through the American Chemical Society and Project SEED, and K.M.A. and the work on this project was supported through NIDDK/NICHD 1RO1DK089201, R24DK090964, and NIH Director New Innovator Pioneer Award DP21DP2OD001500. The funding bodies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
Availability of data and materials
The 16S sequence data generated from this analysis have been deposited in the Sequence Read Archive (SRA) under accession bioproject ID PRJNA317339.
RMP performed experiments, analyzed the data and authored the manuscript. ALP performed experiments, analyzed the data, and authored the manuscript. J.M. analyzed data, provided computational support and authored the manuscript. KMA designed the studies, provided experimental and analytic oversight, data interpretation and analytic projection, and authored the manuscript. BDWB, ASH, MH, KB, PB, DT, TD, ELS, JEF, PK, and KG contributed to experimental design, maintenance of the animals, sample collection, and provided critical revisions of the manuscript. All authors read and approved the final manuscript.
Animals were socially housed within indoor and outdoor enclosures at the Oregon National Primate Research Center (ONPRC). All methods were carried out in accordance with IACUC guidelines and regulations, and all experimental protocols were approved by the IACUC at both ONPRC and Baylor College of Medicine.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 24.Pound LD, Comstock SM, Grove KL. Consumption of a western-style diet during pregnancy impairs offspring islet vascularization in a Japanese macaque model. Am J Phys Metab. 2014;307:E115–23.Google Scholar
- 34.Bartosch S, Woodmansey EJ, Paterson JCM, McMurdo MET, Macfarlane GT. Microbiological effects of consuming a synbiotic containing Bifidobacterium bifidum, Bifidobacterium lactis, and oligofructose in elderly persons, determined by real-time polymerase chain reaction and counting of viable bacteria. Clin Infect Dis. 2005;40:28–37.CrossRefPubMedGoogle Scholar
- 37.Cox J, Williams S, Grove K, Lane RH, Aagaard-Tillery KM. A maternal high-fat diet is accompanied by alterations in the fetal primate metabolome. Am J Obstet Gynecol. 2009;201(281):e1–9.Google Scholar
- 62.Haro C, et al. The gut microbial community in metabolic syndrome patients is modified by diet. J.Nutr. Biochemist. 2016;27:27–31.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.