Responses of gut microbiota to diet composition and weight loss in lean and obese mice

Abstract

Maintenance of a reduced body weight is accompanied by a decrease in energy expenditure beyond that accounted for by reduced body mass and composition, as well as by an increased drive to eat. These effects appear to be due--in part--to reductions in circulating leptin concentrations due to loss of body fat. Gut microbiota have been implicated in the regulation of body weight. The effects of weight loss on qualitative aspects of gut microbiota have been studied in humans and mice, but these studies have been confounded by concurrent changes in diet composition, which influence microbial community composition. We studied the impact of 20% weight loss on the microbiota of diet-induced obese (DIO: 60% calories fat) mice on a high-fat diet (HFD). Weight-reduced DIO (DIO-WR) mice had the same body weight and composition as control (CON) ad-libitum (AL) fed mice being fed a control diet (10% calories fat), allowing a direct comparison of diet and weight-perturbation effects. Microbial community composition was assessed by pyrosequencing 16S rRNA genes derived from the ceca of sacrificed animals. There was a strong effect of diet composition on the diversity and composition of the microbiota. The relative abundance of specific members of the microbiota was correlated with circulating leptin concentrations and gene expression levels of inflammation markers in subcutaneous white adipose tissue in all mice. Together, these results suggest that both host adiposity and diet composition impact microbiota composition, possibly through leptin-mediated regulation of mucus production and/or inflammatory processes that alter the gut habitat.

Figure 1

Effects of diet and weight reduction on the gut microbiota. (a) Phylogenetic Diversity (PD) of the cecal samples from the four groups of mice (mean ± s.e.m. compared by two-way ANOVA) and (b) Principal Coordinates Analysis (PCoA) plot of the unweighted UniFrac distances. PC1 and PC3 values for each mouse sample are plotted; percent variation explained by each PC is shown in parentheses: DIO-AL, diet-induced obese mice: blue; DIO-WR, weight-reduced DIO: red; CON-AL, control diet-fed mice: purple; CON-WR, weight-reduced CON: green. (c) Relative abundances of the different phyla in each of the groups. The phylum Firmicutes was broken down into OTU 303, OTU 716 (which are both classified as Allobaculum), and all other Firmicutes that did not fall into these two OTUs.

Figure 2

Members of the microbiota that differ in abundance by diet composition and treatment (WR vs. AL). (a) Nearest shrunken centroid analysis of the 15 OTUs accounting for the differences among the four groups of mice. For each OTU listed in center, direction of the horizontal bars indicates relatively over-represented (right of vertical line) and under-represented (left of vertical line); the length of the bar indicates the strength of the effect. (b) Heat map of the “classifying” OTUs. Columns show, for each mouse, the abundance data of OTUs listed in center. The abundances of the OTUs were clustered using unsupervised hierarchical clustering (blue, low abundance; red, high abundance). The Phylum, Genus of each of the classifying OTUs is noted. AL, ad-libitum diets; OTU, relative operational taxonomic unit; WR, weight-reduced.

Figure 3

Associations between host serum leptin concentrations and gut microbiota. (a) Correlations of fat mass content (by nuclear magnetic resonance) with circulating leptin concentrations. (b, c, d, e) Correlations between leptin concentrations and the abundance of relative operational taxonomic units of interest. CON-WR, weight-reduced CON: green, CON-AL, control diet-fed mice: purple, DIO-AL, diet-induced obese mice: blue, DIO-WR, weight-reduced DIO: red.

Figure 4

Heat map describing the correlation of the abundances of different operational taxonomic units and transcription levels of inflammation-related genes in inguinal adipose tissue. The colors range from blue (negative correlation; −1) to red (positive correlation; 1). Significant correlations are noted by *P < 0.05 and **P < 0.01 (The computed false discovery rate is about 0.25 using the Benjamini Hochberg procedure (38)).

Figure 5

Schematic depicting possible inter-relationships among diet composition, gut microbiota, circulating leptin, body fat, markers of inflammation, and gut mucin. Body fat directly determines leptin production and elevated body fat increases macrophage infiltration (with associated production/release of inflammatory molecules such as tumor necrosis factor α, serum amyloid A3, and chemokine (C-C motif) ligand 2 (MCP-1) in adipose tissue). The results presented here suggest that diet composition (fractional fat content) directly affects gut microbiota independent of effects mediated by body weight and body composition. Leptin promotes proliferation, differentiation, and survival of immune cells. Leptin also stimulates mucin production in mouse and human intestinal cells (27,28). Mucin affects local bacterial “micro-niches” in the gut by favoring the growth of some bacteria (25,26). Leptin can affect intestinal barrier function by inhibiting apoptosis and promoting regeneration of intestinal epithelium (39,40). These changes in epithelial composition may in turn affect microbiota populations in the gut. The dashed line between body fat and gut microbiota suggests biologically possible connection(s) that might be mediated by adipocytokines or other molecules secreted from adipose tissue.