It's the fiber, not the fat: significant effects of dietary challenge on the gut microbiome

Abstract

Background: Dietary effects on the gut microbiome play key roles in the pathophysiology of inflammatory disorders, metabolic syndrome, obesity, and behavioral dysregulation. Often overlooked in such studies is the consideration that experimental diets vary significantly in the proportion and source of their dietary fiber. Commonly, treatment comparisons are made between animals fed a purchased refined diet that lacks soluble fiber and animals fed a standard vivarium-provided chow diet that contains a rich source of soluble fiber. Despite the well-established critical role of soluble fiber as the source of short chain fatty acid production via the gut microbiome, the extent to which measured outcomes are driven by differences in dietary fiber is unclear. Further, the interaction between sex and age in response to dietary transition is likely important and should also be considered.

Results: We compared the impact of transitioning young adult and 1-year aged male and female mice from their standard chow diet to a refined low soluble fiber diet on gut microbiota community composition. Then, to determine the contribution of dietary fat, we also examined the impact of transitioning a subset of animals from refined low-fat to refined high-fat diet. We used a serial sampling strategy coupled with 16S rRNA marker gene sequencing to examine consequences of recurrent dietary switching on gut microbiota community dynamics. Analysis revealed that the transition from a chow diet to a refined diet that lacks soluble fiber accounted for most of the variance in community structure, diversity, and composition across all groups. This dietary transition was characterized by a loss of taxa within the phylum Bacteroidetes and expansion of Clostridia and Proteobacteria in a sex- and age-specific manner. Most notably, no changes to gut microbiota community structure and composition were observed between mice consuming either refined low- or high-fat diet, suggesting that transition to the refined diet that lacks soluble fiber is the primary driver of gut microbiota alterations, with limited additional impact of dietary fat on gut microbiota.

Conclusion: Collectively, our results show that the choice of control diet has a significant impact on outcomes and interpretation related to diet effects on gut microbiota. As the reduction of soluble fiber may influence synthesis of microbial metabolites that are important for regulating metabolic, immune, behavioral, and neurobiological outcomes, additional studies are now needed to fully delineate the contribution of fat and fiber on the gut microbiome. Video Abtract.

Keywords: Aging; Diet; Metabolism; Microbiome; Sex differences.

Fig. 1

Lack of soluble fiber and increased fat in diet formulations influence weight gain in mice in an age- and sex-specific manner. a Schematic of the experimental study design. Young adult (17 weeks old) and 1-year aged (60 weeks old) C57Bl/6:I129 males and females consuming a chow diet were switched to a refined low-fat diet (rLFD) for 1 week to acclimate. Following acclimation to a refined diet, half of the animals remained on rLFD while the other half was switched onto a 45% refined high-fat diet (rHFD). Purple circles denote times when fecal samples were collected. Animals were co-housed and therefore all analysis is conducted at the level of the cage to control for co-housing effects (N = 3 cages/age/sex/diet, total N = 92 mice). b Composition of diet nutritional composition and ingredients for the chow, rLFD, and rHFD, demonstrating differences in fiber source and quantity between chow and refined diets. c–f To determine the impact of dietary switching in young adult and aged males and females, weekly body weights were collected prior to refined diet switch, 1 week following switch to rLFD, and weekly measurements during consumption rLFD or rHFD. c Body weight was significantly changed over time in young adult females (RM ANOVA, main effect of time, F_{5, 110} = 15.39, P < 0.000, main effect of diet, F_{1, 22} = 2.920, P = 0.1016, N = 24, time × diet interaction, F_{5, 110} = 2.782, P = 0.021). d Body weight of aged females was significantly changed over time (RM ANOVA, main effect of time, F_{5, 90} = 17.43, P < 0.0001, N = 20), across diets (RM ANOVA, main effect of diet, F_{1, 18} = 6.800, P = 0.0178, N = 20), and their interaction (RM ANOVA, time × diet, F_{5, 90} = 12.02, P = < 0.0001, N = 20). Post hoc analysis revealed aged females fed rHFD weighed more at 2 (t₁₀₈ = 3.499, P = 0.0041), 3 (t₁₀₈ = 3.748, P = 0.0017), and 4 (t₁₀₈ = 4.781, P < 0.0001) weeks compared with rLFD-fed aged females. e Body weight was significantly changed over time in young adult males (RM ANOVA, main effect of time, F_{5, 105} = 88.146, P < 0.0001, main effect of diet, F_{1, 21} = 0.4240, P = 0.522, N = 23). f Body weight of aged males was significantly changed over time (RM ANOVA, main effect of time, F_{5, 100} = 67.034, P < 0.0001, N = 22). Data represented as mean ± SEM. Repeated measures ANOVA followed by Sidak correction for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 2

Lack of soluble fiber, not fat, significantly alters gut microbiota composition in young adult and 1-year aged female mice. a Principal coordinates analysis comparing fecal microbiota community structure between females consuming chow and refined diet, demonstrating the significant effect of chow and refined diet on community structure (PERMANOVA, F = 26.284, r² = 0.614, P < 0.0001), accounting for 67.4% of variance. No difference in community structure between rLFD and rHFD females was observed (PERMANOVA, F = 1.75, r² = 0.074, P = 0.132). b Additional principal coordinates analysis comparing fecal microbiota community structure between young adult and aged females consuming chow, rLFD, and rHFD, demonstrating significant interactions of age, rLFD, and rHFD on community structure (PERMANOVA, F = 10.122, r² = 0.614, P < 0.001). c, d Comparison of community diversity in young adult and aged females consuming chow, rLFD, and rHFD. c The alpha diversity measure, observed species, plotted against sampling time point demonstrating no impact of diet transition and consumption of refined diets on the number of unique taxa (Kruskal-Wallis, H = 9.45, P = 0.22). d The alpha diversity measure, Shannon diversity index, plotted against sampling time point shows significant differences in community richness and evenness in young adult and aged females (Kruskal-Wallis, H = 26.647, P = 0.00038). e Stacked barplot showing the average relative abundance of taxa in chow, rLFD, and rHFD young adult and aged females, characterized by a rapid and lasting loss of taxa within the Bacteroidetes phylum and concomitant bloom of taxa within the Firmicutes and Proteobacteria phyla. Taxa key is truncated to display the 18 most abundant taxa. f Heatmap depicting 34 significantly different taxa by age and diet in females identified by linear discriminant analysis (FDR < 0.05). Columns represent taxa within each cage. N = 3 female cages/age/diet sampling time point. Data represented as individual data points average per cage ± SEM

Fig. 3

Lack of soluble fiber, not fat, significantly alters gut microbiota composition in young adult and 1-year aged male mice. a Principal coordinates analysis comparing fecal microbiota community structure between males consuming chow and refined diet, demonstrating the significant effect of chow and refined diet on community structure (PERMANOVA, F = 26.577, r² = 0.617, P < 0.0001), accounting for 69.9% of variance. No difference in community structure between rLFD and rHFD males was observed (PERMANOVA, F = 1.264, r² = 0.054, P = 0.261). b Additional principal coordinates analysis comparing fecal microbiota community structure between young adult and aged males consuming chow, rLFD, and rHFD, demonstrating significant interactions of age, rLFD, and rHFD on community structure (PERMANOVA, F = 8.84, r² = 0.688, P < 0.001). c, d Comparison of community diversity in young adult and aged males consuming chow, rLFD, and rHFD. c The alpha diversity measure, observed species, plotted against sampling time point demonstrating significant differences in the number of unique taxa following dietary transition and consumption of refined diets (Kruskal-Wallis, H = 17.32, P = 0.015). d The alpha diversity measure, Shannon diversity index, plotted against sampling time point shows significant differences in community richness and evenness in young adult and aged males (Kruskal-Wallis, H = 26.943, P = 0.00034). e Stacked barplot showing the average relative abundance of taxa in chow, rLFD and rHFD young adult and aged males, characterized by a rapid and lasting loss of taxa within the Bacteroidetes phylum and concurrent bloom of taxa within the Firmicutes and Proteobacteria phyla. Taxa key is truncated at the 18 most abundant taxa. f Heatmap depicting 32 significantly different taxa by age and diet in females identified by linear discriminant analysis (FDR < 0.05). Columns represent taxa within each cage. Data represented as individual data points averaged per cage ± SEM