Regulators of gut motility revealed by a gnotobiotic model of diet-microbiome interactions related to travel

Abstract

To understand how different diets, the consumers' gut microbiota, and the enteric nervous system (ENS) interact to regulate gut motility, we developed a gnotobiotic mouse model that mimics short-term dietary changes that happen when humans are traveling to places with different culinary traditions. Studying animals transplanted with the microbiota from humans representing diverse culinary traditions and fed a sequence of diets representing those of all donors, we found that correlations between bacterial species abundances and transit times are diet dependent. However, the levels of unconjugated bile acids-generated by bacterial bile salt hydrolases (BSH)-correlated with faster transit, including during consumption of a Bangladeshi diet. Mice harboring a consortium of sequenced cultured bacterial strains from the Bangladeshi donor's microbiota and fed a Bangladeshi diet revealed that the commonly used cholekinetic spice, turmeric, affects gut motility through a mechanism that reflects bacterial BSH activity and Ret signaling in the ENS. These results demonstrate how a single food ingredient interacts with a functional microbiota trait to regulate host physiology.

Figure 1. Compositions of diets used in the 6-phase travel experiment

In the foreground, word clouds convey the specific ingredients used: font sizes depict the weight-based proportional representations of ingredients within each diet. Pie charts in the background present the macromolecular compositions. See also Table S1.

Figure 2. Diet and microbiota significantly impact intestinal transit times

(A) Schematic of 6-phase travel experiment. Groups of adult germ-free C57BL/6 mice were colonized with fecal microbiota from six healthy donors and fed diets representative of those consumed by all donors in the sequence shown in panel B. (B) The central squares of this heat map represent means of transit times for each diet-microbiota combination as measured by carmine red dye assay; the frames of the squares represent s.e.m. (n=6 mice/donor microbiota). Microbiota are represented along rows, diet phases along columns. Each group of mice consumed human diets in the order shown from left to right; home diets were always consumed during the initial and final diet phases but skipped in the intervening travel diet progression (primal diet unrestricted American diet Bangladeshi diet Malawian diet Amerindian diet). (C) Histogram showing distribution of all transit times recorded throughout the 6-phase travel experiment. (D) The most contrasted diet-by-microbiota effects on transit times were observed in mice colonized with a Bangladeshi or USA_unrestricted microbiota and fed Bangladeshi versus primal diets. Results for USA_unrestricted (lean) and USA_unrestricted (obese) were aggregated and are represented together as USA_unrestricted. Results from the ‘home’ and ‘return home’ phases for mice colonized with a Bangladeshi microbiota and fed a Bangladeshi diet were also aggregated. Statistical significance was determined using a two-tailed Student’s t-test; *, p<0.05. See also Figures 1, 3, S1, S2 and Tables S1, S2, S3, S4, S5.

Figure 3. Diet-discriminatory OTUs are robust to different donor microbiota and motility phenotypes

(A) Out-of-bag estimated error rates in a Random Forests model for predicting diet, stratified by donor microbiota, as a function of numbers of diet-discriminatory OTUs. For each microbiota, 40 OTUs were sufficient to discriminate diet, yielding a total of 87 unique OTUs across six microbiota donors from five cultural/dietary traditions in the 6-phase travel experiment. (B) Evidence for the robustness of diet-discriminatory OTUs to donor microbiota and motility phenotype. Feature importance scores of 87 diet-discriminatory OTUs in each diet-microbiota context are represented in this heat map. A sparse Random Forests model built using these diet-discriminatory OTUs accurately predicted diet in the 3-phase travel experiment. See also Table S5.

Figure 4. Significant correlations between fecal bile acid metabolite concentrations and the relative abundances of bacterial 97%ID OTUs

Bile acid metabolite profiling and 16S rRNA analysis was performed on fecal samples collected from mice in the 3-phase travel experiment (Figure S1). Spearman rank correlations were calculated between bile acid concentrations and relative abundances of 97%ID OTUs. Unsupervised hierarchical clustering was applied. Significant associations (p<0.05 calculated by two-tailed Student’s t-test, with Bonferroni correction) between microbiota/diet and bile acids/OTUs are represented in the vertical and horizontal side panels. Associations between bile acids and motility were calculated by linear modeling with stepwise backward feature selection, as detailed in the text and Supplemental Experimental Procedures. See also Figures S1 and Tables S4, S6.

Figure 5. An interaction between diet, bile acid metabolism, and gut motility revealed by colonizing germ-free mice with a BSH_hi or BSH_lo consortium and feeding them a representative Bangladeshi diet with or without turmeric

(A) Turmeric consumption resulted in significantly slower motility in mice colonized with the BSH_lo but not the BSH_hi consortium. (B) While consuming the turmeric-supplemented Bangladeshi diet, BSH_hi mice had transit times comparable to mice colonized with the complete 53 strain culture collection (“CC”) and faster motility (i.e., shorter transit times) than BSH_lo mice (n=5 animals/microbiota). (C) Turmeric consumption is associated with a significant increase in total fecal unconjugated bile acid concentrations in gnotobiotic wild-type mice colonized with the BSH_hi consortium but not the BSH_lo consortium. (D) Transit times (mean ± s.e.m.) measured for two groups of gnotobiotic mice colonized with BSH_lo consortium and fed the unsupplemented or turmeric-containing Bangladeshi diet. Measurements occurred at the end of the single (monotonous) 10-day diet experiment. Statistical significance was determined using a one-tailed Student’s t-test. *, p<0.05. See also Tables S4, S6, S7, S8.