Progress 03/01/19 to 02/28/22
Outputs
Target Audience:We primarily targeted other researchers in this field including nutritional scientists, microbiologists, chemists, scientists in vascular biology and medical doctors. Due to the pandemic, a small number of seminars were presented; these included talks at several meetings including the International Society of Microbial Ecology Latin American Meeting, Colombia (online), International Microbiome Symposium, University of Costa Rica (online), American Chemical Society Meeting (spring meeting), and academic institutions including the Microbiome centers at the Cleveland Clinic, University of South Florida, and University of North Carolina, the aging program at the University of Wisconsin- Madison, the Biochemistry department at the University of Georgia. Changes/Problems:Due to issues described in previous cycle with germ-free apoE mice, we focused the study on cardiometabolic phenotypes, but did not assess atherosclerosis. What opportunities for training and professional development has the project provided?In the last year, this grant provided opportunities to train one undergraduate student to work with germ-free mice and one graduate student, in techniques related to anaerobic microbiology, food sciences, gnotobiology, and nutrition How have the results been disseminated to communities of interest?Results have been disseminated through on-line seminars at several universities in the US and at national and international meetings. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
What was accomplished under these goals? This work seeks to assess the relationship between interpersonal differences in gut microbiome and health benefits obtained from the consumption of food containing high levels of fibers including legumes, whole-grains (WG), and vegetables. These foods represent common sources of dietary fiber in the US and have been associated with protective effects against cardiometabolic disease (CMD). However, clinical intervention studies have revealed a large degree of interpersonal variation in the effects associated with their consumption and we cannot currently predict who will benefit from their consumption. GOAL 1: 90% completed (i-ii) Major activities completed / experiments conducted/Data collected: a. Most data were collected in previous 2 cycles. In the last cycle we focused on analysis and interpretation of data with a focus on assessment of untargeted microbial metabolites (HPLC MS/MS), host phenotypes (measurements of gene expression on host tissues) and data integration. (iii) Summary statistics and discussion of results: GF B6 mice were colonized vwith human fecal communities subA, SubB or subC (n = 28-30 mice/community). Colonized mice were then divided into 3 treatments and received 1 of 3 isocaloric diets that differed on the type of fiber-rich whole food they contained (i) whole grains (WG): oats and quinoa trim; (ii) vegetables: carrots and broccoli; or (iii) legumes: black beans and pinto beans (n = 9-10 mice/ community/diet). We also maintained groups of GF mice in each diet (n =8/diet). Diets provided to the mice also contained a relatively high fat content (45% cal derived from fat) and 1% cholesterol to promote disease. Mice were maintained in these experimental diets for 12weeks. Fecal samples were collected prior to the diet intervention (i.e., baseline) and at the end of the study for 16S rRNA gene analysis. PCoA of uwUniFrac distances of fecal samples collected before dietary intervention show a clear clustering by donor community, suggesting that mice colonized with the same community harbored highly similar baseline microbiomes. The same analysis applied to samples collected after 12 weeks of diet showed similar patterns, where samples showed a clear clustering by donor and to a less extent by diet. PCoA analysis of weighted UniFrac distances of fecal microbiome from mice colonized with SubA or SubB consuming the 3 diets (WG vs. vegetable vs. legume) disclosed significant differences in composition (p < 0.05), suggesting these fiber-rich foods caused significant restructuring of these two communities. In SubC, mice consuming legume diet showed significant differences relative to WG and vegetable fed-animals (p < 0.05). However, SubC-colonized mice consuming WG and vegetable diets harbored similar microbiomes (p = 0.081). Altogether these results suggest that all three diets support colonization of similar assemblages for each community, however for each community there were taxa that varied significantly in their abundances as a function of diet, mostly in a community-specific manner. We compared metabolic phenotypes of the animals mentioned above. We found that the three diets varied on how they impacted host metabolic outcomes and that these effects differed by community. SubA-colonized animals consuming the vegetable diet had the lowest insulin levels compared with mice colonized with the same community consuming any of the other two diets. Additionally, SubA-colonized animals consuming the vegetable and legume diets showed significantly lower HOMA-IR than counterparts consuming WG diet (p < 0.05). SubB-colonized mice also showed different metabolic responses to the three diets, with the vegetable diet associated with more beneficial outcomes as well, but these differed from those described for SubA mice. For example, SubB-colonized animals showed significantly reduced serum cholesterol and triglyceride levels in the vegetable diet relative to the other two diets (p < 0.05); vegetable diet consumption was also associated with lower serum glucose levels for mice colonized with this community. SubC-colonized mice consuming the vegetable diet showed reduced body weight and fat mass relative to counterparts consuming WG diet (p < 0.05). For mice colonized with this community levels of cholesterol were also lower in animals consuming vegetable and legume diets compared to WG diet (p < 0.05), whereas vegetable diet consumption was associated with lower serum triglycerides levels relative to WG and legume (p < 0.05). Altogether these results suggested that vegetable diet was associated with overall better health outcomes across all communities while legume diet was associated with better health outcomes only in SubA and SubC colonized mice. Furthermore, the phenotypes impacted by these two diets varied by community. We performed untargeted metabolomics analysis at experimental endpoint. We focused on metabolites modulated by the microbiome, indicated as elevation or depletion compared to GF controls. Few of these metabolite features were unique to a given donor or a given diet, though the greatest proportion of uniquely elicited metabolites for a given donor were observed under the WG diet (p<0.05 for all pairwise comparisons between WG diet and other diets, for a given donor, except for legume vs WG in subB). The donor community with the greatest number of unique metabolites was variable depending on the diet: subC for vegetables and WG, and subB for legume diet (p<0.05). Across all diets, subA had the fewest unique metabolite features (p<0.05 for all pairwise comparisons, except for WG diet subA vs subB). In contrast, we observed substantial variation in metabolite abundance between donors and diets (BC distance; PERMANOVA p=0.001 R2=0.196 for diet; p=0.001 R2=0.174 for donor; p=0.001 R2=0.0783 for the interaction between diet and donor). The vegetable diet was the most divergent (pairwise PERMANOVA pseudo-F of 10.0 for legumes vs vegetable and 12.0 for WG vs vegetable; pseudo-F of 8.4 for legumes vs WG; all q-values=0.001). SubB was the most divergent overall, with subA closer to subC (pairwise PERMANOVA pseudo-F of 9.4 for subB vs. subA and of 13.6 for subB vs subC, both q-value=0.0015; pseudo-F of 3.2 for subA vs subC, q-value=0.003). Within each donor, all diets differed significantly from each other (pairwise PERMANOVA q-value<0.05), with the most divergent diet different for each donor (WG for subA and subC, legume for subB). In accordance with the considerable beta diversity changes observed, 59% of microbiota-modulated metabolite features significantly differed between donor-diet combinations (1768/3016, Kruskal-Wallis p<0.05, FDR-corrected). These metabolites are mainly lipids (glycerophosphocholines, acylcarnitines, linoleic acid derivatives), small peptides, and plant-derived metabolites including triterpenoids, hydroxycinnamic acids. Of particular interest is commendamide, which is a known GPCR agonist produced by bacteria that was elevated across diets in subA (p<0.05, FDR-corrected). Bile acids differed in abundance between diets and between donors. We are currently exploring association of these metabolites with host phenotypes and assessing metabolites in serum (all samples have been analyzed by MS). iv) Key outcomes or other accomplishments realized. The results presented here and from the published work from last year (Murga-Garrido 2021) suggest that the gut microbiome represents an important and differential factor that modulates the benefits associated with consumption of purified fibers as well as vegetables and legumes on metabolic disease. A paper on the work summarized above is in progress with an estimated submission date for July 2022.
Publications
|
Progress 03/01/20 to 02/28/21
Outputs
Target Audience:We primarily targeted other researchers in this field including nutritional scientists, microbiologists, chemists, scientists in vascular biology and medical doctors. Due to the pandemic, a small number of seminars were presented; these included talks at the- The Allied Genetics Conference (Genetic Society of America) meeting, the Nebraska Food For Health Center, University of Nebraska; the Microbiome Center at Johns Hopkins University, Department Pharmacology at University of Toledo, Department of Chemical Biological Chemistry at the University of California, Irvine; Department of Nutritional Sciences, Cornell Unniversity; University of Wisconsin Metabolism Symposium (all lectures were given online). Changes/Problems:In this phase of the project, we focused on metabolic phenotypes and have not measured atherosclerotic lesions in the mice. The reason for this is that our germ-free facility got contaminated during the 2020 covid-related shutdown. We lost the germ-free apoE ko colony, which are crucial for atherosclerosis measurements, Nonetheless, we are measuring risk factors for atherosclerosis including cholesterol. Additionally, we have now re-derived ApoE ko mice germ-free again and we plan to perform atherosclerosis studies with the diets/communities described above. What opportunities for training and professional development has the project provided?In the last year, this grant provided opportunities to train one undergraduate student to work with germ-free mice and one graduate student (MS), in techniques related to anaerobic microbiology, food sciences, gnotobiology, and nutrition. How have the results been disseminated to communities of interest?Results have been disseminated through on-line seminars at several universities in the US and at national and international meetings. What do you plan to do during the next reporting period to accomplish the goals?In the next year, we plan to test causal links of microbes and microbe-derived metabolites modulated by diet that are associated with metabolic phenotypes. We plan to use gntobiotic mouse models of disease, the diets described above and defined communities containing/lacking the microbes of interest (e.g., Prevotella).
Impacts
What was accomplished under these goals?
This work represents a translational approach for determining the relationship between interpersonal differences in gut microbiome and health benefits obtained from the consumption of food containing high levels of fibers including legumes, whole-grains, and vegetables. These foods represent major sources of fiber among US consumers and have been associated with protective effects against cardiovascular and metabolic disease (CMD). However, clinical intervention studies have revealed a large degree of interpersonal variation in the effects associated with their consumption and we cannot currently predict who will benefit from their consumption. The results presented in the second year of funding, suggest that host responses to these foods (i.e., benefits) are modulated by the gut microbiome and support the notion that in some cases microbe-centered personalized diets may be advantageous. GOAL 1: 60% completed (i) Major activities completed / experiments conducted: a. Publication of study focused on purified dietary fibers (research progress reported during 2020 cycle) b. Transplantation of microbiomes in germ-free mice; Assessments of phenotypes in mice; fecal DNA sequencing to characterize microbial communities in mice; Analyses of HPLC/MS of serum and ceca from mice (untargeted metabolomics), analysis of GC/MS to quantify SCFAs in cecal contents. Diets used in the study: We created four isocaloric high fat diets supplemented with different types of high-fiber foods. These included whole grains, vegetables, and legumes; diet supplemented with pure cellulose was used as a control. For the whole ingredient diets we chose high-fiber foods commonly consumed in the US and for which lyophilized ready-to-eat powders were available. We obtained lyophilized powders for broccoli, carrot (vegetables), oats, quinoa (whole grains), and black beans, pinto beans (legumes) from Futureceuticals and Van Drunen farms. These ingredients or cellulose were added into synthetic mouse diets and macro nutrients were adjusted so the four isocaloric diets contain 45% w/w Fat, 1% w/w cholesterol, 7% w/w dietary fiber. All diets were manufactured and sterilized by irradiation at Envigo. (ii) Data collected: Metabolic phenotypes from mice (body weight, fat weight, glucose, insulin, triglycerides in serum and liver, cholesterol); Fecal microbiota composition from transplanted mice (16SrRNA gene sequencing); untargeted metabolomics in plasma samples (LC-MS/MS using C8 column, positive mode detection) and cecal levels of SCFA in mice. (iii) Summary statistics and discussion of results: Three groups of 5-6 weeks-old male GF C57BL/6 mice were colonized via oral gavage with 3 fecal communities: HumA, HumB, HumC (n=40/community) and maintained for two weeks in autoclaved chow. These communities were obtained from overweight individuals with no history of type II diabetes or vascular disease (characterization of these communities was done in published 2021 study). Mice colonized with each community were then divided into four groups; each group received one of the four diets described above for 12 weeks (n=10 mice/community/diet). At the end this period phenotypes described above were measured. Groups of conventionally-raised and germ-free mice in the same diets for the same period of time were used as controls. For each community, we detected differences in metabolic phenotypes as a function of diet. Remarkably, these differences were not consistent among communities. For example, mice colonizedwith HumC community fed vegetable diet show lower body weight and epididymal fat pads compared to animals colonized in the same community consuming any of the other diets (P<0.05). Furthemore, no differences in these phenotypes were observed in mice colonized with HumA or HumB. In contrast mice colonized with HumA showed improved insulin sensitivity as determined by the homeostasis model assessment-estimated insulin resistance (HOMA-IR) index in the vegetable and legume diets compared to the whole grain and cellulose diets (P<0.01), however, we did not observed differences in HOMA-IR among diets for any of the other communities. Gut microbiome analysis identified community-specific taxa associated with the metabolic phenotypes. We found that the genus Prevotella was increased by the vegetable and legume diets, relative to cellulose and whole-grain diets in HumA colonized mice, but this genus was not detected in the other two communities. This is of interest because previous studies have shown that Prevotella plays a role in the barley kernel-based bread-induced improvement in glucose metabolism observed in certain individuals. Our results extend these to potentially legumes and vegetables. Follow-up studies will be focused on establishing causal links between Prevotella legumes/vegetables and improved glucose homeostasis using defined mouse communities. We will also analyze untargeted metabolomic data that was collected during the last year to identify potential metabolites mediating these phenotypes. Preliminary analysis of the untargeted metabolomic data suggest that diet has a greater impact on plasma metabolome than donor whereas fecal inoculum (i.e., community) has a greater impact on the cecal metabolome than diet. iv) Key outcomes or other accomplishments realized: The results presented here and from the published work (Murga-Garrido 2021) suggest that the gut microbiome represents an important and differential factor that modulates the benefits associated with consumption of purified fibers as well as vegetables and legumes on metabolic disease.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2021
Citation:
Murga-Garrido SM, Hong Q, Cross TWL, Hutchison ER, Han J, Thomas SP, Vivas EI, Denu J, Ceschin DG, Tang ZZ, Rey FE. 2021 Gut microbiome variation modulates the effects of dietary fiber on host metabolism. Microbiome, 9:117 (PMID: 34016169)
|
Progress 03/01/19 to 02/29/20
Outputs
Target Audience:We primarily targeted other researchers in this field including nutritional scientists, microbiologists, chemists, scientists invascular biology and medical doctors. Seminars on this work were given at Yale University, University of Washington (Microbiology), The Hormell Institute (MN) and Kern Lipid Conference (Colorado). Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?This grant provided opportunities to train three undergraduate students to work with germ-free mice, one graduate student (MS), one postdoctoral fellow in techniques related to anaerobic microbiology, food sciences, gnotobiology, and cardiovascular disease. How have the results been disseminated to communities of interest?Results have been disseminated through seminars at several universities in the US (U of Chicago, Yale, U of Washington), and a presentation at the Lipid Kern conference Colorado). What do you plan to do during the next reporting period to accomplish the goals?We will continue working in Aiim1, increasing the number of microbiome transplanted in mice. We plan complete mouse data collection by the end of year 2 (experiment with whole foods).
Impacts
What was accomplished under these goals?
This work represents a translational approach for determining the relationship between interpersonal differences in gut microbiome (microbes that live in the gut) and health benefits obtained from the consumption of legumes, whole-grains, and vegetables. These foods represent major sources of fiber among US consumers and have been associated with protective effects against cardiovascular and metabolic disease (CMD). However, clinical intervention studies have revealed a large degree of interpersonal variation in the effects associated with their consumption and we cannot currently predict who will benefit from their consumption. The results presented in the first year of funding, represent the work of a graduate student and a post-doctoral fellow and suggest that protection against CMD associated with consumption of dietary fiber is modulated by the gut microbiome. We have transplanted distinct human gut microbial communities into genetically identical germ-free mice and fed them defined diets that vary on the type of fiber. We found that the benefits from each fiber differed based on the microbiome the mice harbor. Our results suggest that the gut microbiome represents an important and differential factor that modulates the benefits associated with fiber consumption. These results will be used to identify novel probiotics strains of bacteria that may be given together with these foods to enhance their beneficial effects, and to elucidate rapid screens of individuals to inform maximal nutritional efficacy. GOAL 1: 25% completed (i) Major activities completed / experiments conducted: Diet design; Transplantation of microbiomes in germ-free mice; Assessments of phenotypes in mice (measurements of body weight, glucose, quantification of atherosclerosis lesions; Fecal DNA sequencing to characterize microbial communities in mice; Analyses of HPLC/MS of plasma from mice. (ii) Data collected Metabolic phenotypes from mice (glucose, insulin, triglycerides, atherosclerosis measurements); Fecal microbiota composition from transplanted mice (16SrRNA sequencing); Plasma levels of ~700 metabolites from transplanted mice; Gene expression liver (being analyzed). (iii) Summary statistics and discussion of results Towards the central goal of the grant, we have formulated four diets that vary in the main source of fiber: i) vegetables (dried broccoli/carrots); ii) legumes (pinto/black beans); iii) whole grains (oat bran/quinoa trim); non-fermentable fiber (cellulose). All diets are isocaloric and contain 7% of dietary fiber. We have transplanted 3 human microbiomes with distinct metabolic capabilities into 12 groups of germ-free animals (3 microbiomes X 4 diets; 120 mice total). The process of selection/validation of our approach are described below. Mice are being maintained in the diets mentioned above for 12 weeks. Body weight and food consumption are monitored in a weekly basis. At the end of the experiments, we will assess cardiometabolic phenotypes. Before performing these experiments (while we were formulating/producing diets containing whole food ingredients mentioned above), we validated our approach using diets of defined composition (i.e., pure ingredients). Mice were fed diets supplemented with an assortment of fibers that included resistant starch (RS) type 2 and 4, short chain fructo-oligosaccharides, inulin, and pectin Animals were placed on this diet one week prior to colonization and maintained for four additional weeks after inoculation. Unweighted UniFrac-based comparisons of bacterial 16S rRNA gene sequences generated from the input human donor microbiotas and from cecal samples collected from the transplanted mice revealed that in each case the engrafted microbiomes resembled more the fecal sample used to colonize the animals than any of the other donor samples used in the study. Importantly, the transplanted microbiomes showed distinct organismal configurations and exhibited differences in alpha-diversity. As expected from this variation, transplanted communities also differed in their capacity to produce short chain fatty acids. Remarkably there was a ~4-fold range in the levels of cecal butyrate among the eight groups despite all animals consuming the same diet. Butyrate is known to vary widely among humans and has been linked with beneficial health effects on the host Sub1 (high butyrate) and Sub8 (low butyrate), to examine how these two communities modulate host responses to different types of fiber. 8-week-old male GF B6 mice were placed on the assorted diet described above for one week and then orally inoculated with fecal slurries prepared from Sub1 and Sub8 (n=30-36 mice/community) and maintained in the same diet for two weeks. After this stabilization period, each of the two groups of animals were divided into four groups. Each of these groups received one of four isocaloric diets (n=7-10 mice/community/diet), which vary in the type of fiber they contained. Fibers used in the study were (i) cellulose (non-fermentable fiber control), (ii) inulin, (iii) pectin, or (iv) assorted fiber described above. After four weeks in these diets, we assessed metabolic phenotypes in mice colonized with the two different microbiomes. In the control diet, mice harboring the Sub8 microbiome showed decreased adiposity relative to animals colonized with Sub1 microbiome, whereas there were no statistical differences in the levels of liver triglycerides and fasting plasma glucose between these groups. In the inulin diet, mice colonized with the Sub8 microbiome showed significantly lower levels of adiposity, decreased liver triglycerides and lower levels of glucose in plasma relative to animals colonized with Sub1. In contrast, pectin fed mice colonized with the Sub8 microbiome accumulated more fat relative to animals colonized with the Sub1 microbiome. Lastly, mice colonized with the Sub8 microbiome showed significantly lower levels of adiposity than those colonized with Sub1 in the assorted fiber diet, whereas plasma glucose and liver triglycerides were comparable between the two groups. Altogether, these results suggest that the impact that different dietary fibers have on host metabolism is modulated by the composition of the gut microbiome. 16S rRNA gene sequences were generated from cecal samples collected at the end of the experiment from the mice described above. Principal coordinates analyses (PCoA) of unweighted UniFrac distances of these samples show a clear clustering by community (i.e., Sub1 and Sub8) and less by diet, supporting the notion that all four diets support colonization of similar assemblages for both communities. In all diets, each community preserves a core of common species that differentiate from the other community. Nevertheless, there are concomitant subtle richness and pronounced abundance variations within each community across the different fiber treatments. Microbes associated with the different CMD phenotypes were found. To investigate the relationship between plasma metabolome and the two communities as a function of the different dietary fibers we quantified 774 metabolites in plasma from the 8 groups of mice described above (6 samples /community/diet were analyzed). Two-way ANOVA analysis, between Sub8- and Sub1-transplanted mice identified 69 plasma metabolites in the cellulose diet, 235 in the inulin diet, 177 in the pectin diet, and 93 in the assorted fiber diet that showed significant differences between the two groups (p value< 0.05, q-value <0.10). These results suggest that fiber consumption affect plasma metabolome in a microbiome specific manner. (iv) Key outcomes or other accomplishments realized. The results presented here represent a change in knowledge and suggest that the gut microbiome represents an important and differential factor that modulates the benefits associated with fiber consumption. We are preparing a manuscript with some of the results presented above.
Publications
|
|