Leptin acts independently of food intake to modulate gut microbial composition in male mice

Abstract

Shifts in the composition of gut bacterial populations can alter host metabolism and may contribute to the pathogenesis of metabolic disorders, including obesity. Mice deficient in leptin action are obese with altered microbiota and increased susceptibility to certain intestinal pathogens. Because antimicrobial peptides (AMPs) secreted by Paneth cells represent a major mechanism by which the host influences the gut microbiome, we examined the mRNA expression of gut AMPs, several of which were decreased in leptin receptor (LepR)-deficient db/db mice, suggesting a potential role for AMP modulation of microbiota composition. To address the extent to which the alterations in gut microbiota and AMP mRNA expression in db/db mice result from increased food intake vs other defects in leptin action, we examined the effects of pair feeding and gut epithelial LepRb ablation on AMP mRNA expression and microbiota composition. We found that the phylum-level changes in fecal microbial content and AMP gene expression persist in pair-fed db/db mice, suggesting that these differences do not stem from hyperphagia alone. In addition, despite recent evidence to support a role for intestinal epithelial LepRb signaling in pathogen susceptibility, ablation of LepRb from the intestinal epithelium fails to alter body weight, composition of the microbiota, or AMP expression, suggesting a role for LepRb elsewhere for this regulation. Indeed, gut LepRb cells are not epithelial but rather constitute a previously uncharacterized population of perivascular cells within the intestinal submucosa. Overall, our data reveal a role for LepRb signaling extrinsic to the intestinal epithelium and independent of food intake in the control of the gut microbiome.

Figure 1.

Quantitative expression analysis of Paneth cell–specific antimicrobial peptides in ad libitum–fed db/db mice compared with that in wild-type (WT) animals. A, Average body weight (grams) of wild-type and db/db mice at 8 weeks of age. B, Percent mRNA expression of antimicrobial peptides in the distal ileum relative to the control (GAPDH) for both wild-type and db/db animals. Graphed data represent average values ± SEM. n = 13 for wild-type and n = 10 for db/db mice. Statistical analysis with the Student t test: *, P < .05

Figure 2.

Cumulative food intake (grams) from weeks 4 to 8 (A) and average body weight (grams) at 8 weeks of age (B) did not differ between wild-type (WT) and pair-fed db/db mice. n = 13 for wild-type and n = 20 for pair-fed db/db mice.

Figure 3.

Altered microbiota composition and expression of antimicrobial peptides in pair-fed db/db animals compared with those in wild-type (WT) mice. A, Comparison of fecal microbiota populations by 16S rRNA-encoding gene sequencing. B, Percent relative mRNA expression of antimicrobial peptides in the distal ileum (normalized to GAPDH). Graphed data represent average values ± SEM. n = 13 for wild-type and n = 20 for pair fed db/db mice. Statistical analysis with two-way ANOVA (A) and the Student t test (B): *, P < .05

Figure 4.

IEC-LeprKO mice. A, Genomic deletion of LepRb by Villin-Cre is specific for intestinal epithelium. B and C, Cumulative food intake (grams) from weeks 4 to 8 (B) and average body weight (grams) at 8 weeks of age (C) did not differ between IEC-LepRb-KO mice and the control cohort. n = 11 for IEC-lepRb-KO and n = 26 for controls. Not significant.

Figure 5.

Microbiota composition and expression of antimicrobial peptides do not differ between IEC-Lepr-KO mice and control. A, Percent relative mRNA expression of antimicrobial peptides in the distal ileum (normalized to GAPDH). B, Comparison of fecal microbiota populations (% total phyla) by 16S rRNA-encoding gene sequencing. Graphed data represent average values ± SEM. n = 10 for IEC-LepRb-KO mice and n = 25 for the control group. Statistical analysis with the Student t test (A) and two-way ANOVA (B): not significant.

Figure 6.

Immunohistochemical analysis of ileum (A) from LepRb-Cre × Rosa26-Tomato reporter mice reveals labeling of a submucosal cell population and absence of epithelial labeling. Costained ileal sections (B) from these mice with antibodies against endothelial (CD31, green) and lymphatic (LyVe-1, blue) markers confirmed perivascular localization (B, inset).

Figure 7.

Quantitative expression analysis of LepRb from 1-cm intestinal segments from LepRb-tomato reporter mice reveals a marked increase in LepRb expression relative to that of GAPDH from proximal to distal bowel (A, graph). Concomitantly, the presence of LepRb⁺-tomato fluorescent-labeled cells increases in frequency within the submucosa (A, immunohistochemical analysis). Quantitative analysis of LepRb relative to GAPDH in epithelium vs the remaining deepithelialized ileum (B) revealed lower expression within the epithelium. Graphed data represent average values (A and B) ± SEM (A). n = 3 for each.

Figure 8.

A, FACS of dissociated ileum from LepRb-tomato reporter mice revealed these cells to be 0.25% to 0.5% of the total cell population within the ileum. B, Quantitative analysis of LepRb expression relative to GAPDH in unsorted but dissociated cells, tomato fluorescent “positive” cells and tomato fluorescent “negative” cells confirm that the tomato-positive cells are the predominant source of LepRb mRNA within the intestine. pos, positive; neg, negative.