Disruption of the murine Glp2r impairs Paneth cell function and increases susceptibility to small bowel enteritis

Abstract

Exogenous glucagon-like peptide-2 receptor (GLP-2R) activation elicits proliferative and cytoprotective responses in the gastrointestinal mucosa and ameliorates experimental small and large bowel gut injury. Nevertheless, the essential physiological role(s) of the endogenous GLP-2R remain poorly understood. We studied the importance of the GLP-2R for gut growth, epithelial cell lineage allocation, the response to mucosal injury, and host-bacterial interactions in Glp2r(-/-) and littermate control Glp2r(+/+) mice. Glp2r(-/-) mice exhibit normal somatic growth and preserved small and large bowel responses to IGF-I and keratinocyte growth factor. However, Glp2r(-/-) mice failed to up-regulate intestinal epithelial c-fos expression in response to acute GLP-2 administration and do not exhibit changes in small bowel conductance or small or large bowel growth after administration of GLP-2R agonists. The crypt and villus compartment and the numbers and localization of Paneth, enteroendocrine, and goblet cells were comparable in Glp2r(+/+) vs. Glp2r(-/-) mice. Although the severity and extent of colonic mucosal injury in response to 3% oral dextran sulfate was similar across Glp2r genotypes, Glp2r(-/-) mice exhibited significantly increased morbidity and mortality and increased bacterial translocation after induction of enteritis with indomethacin and enhanced mucosal injury in response to irinotecan. Moreover, bacterial colonization of the small bowel was significantly increased, expression of Paneth cell antimicrobial gene products was reduced, and mucosal bactericidal activity was impaired in Glp2r(-/-) mice. Although the Glp2r is dispensable for gut development and the response to colonic injury, Glp2r(-/-) mice exhibit enhanced sensitivity to small bowel injury, and abnormal host-bacterial interactions in the small bowel.

FIG. 1

Characterization of Glp2r^−/− mice. A, Body weights of male and female Glp2r^+/+, Glp2r^+/−, and Glp2r^−/− littermates (n = 4 per group) over 24 wk. B, The c-fos-immunopositive cell number in the jejunal epithelium of Glp2r^+/+ and Glp2r^−/− male mice (n = 3– 4 per group) 1 h after GLP-2 (5 μg) administration. Small and large bowel weights (C) and jejunal crypt cell proliferation (D) after the administration of saline or GLP-2 (10 μg once daily) for 7 d in female Glp2r^+/+, Glp2r^+/−, and Glp2r^−/− littermates (n = 7– 8 per group). E, Small bowel weight in response to GLP-2 (2.5 μg twice daily), Long-R³-IGF-I (25 μg twice daily), or rHuKGF (5 mg/kg body weight once daily) administration to female Glp2r^+/+ and Glp2r^−/− mice for 10 d (n = 5– 6 per group). F, The jejunal (Jej) transmural ion conductance from female Glp2r^+/+ and Glp2r^−/− mice (n = 3– 6 per group) administered saline or GLP-2 (5 μg twice daily) for 9 d. *, P < 0.05, ** or ##, P < 0.01, *** or ###, P < 0.001, treatment vs. vehicle control.

FIG. 2

Body weight (A), large bowel weight (B) and length (C), histology (D), and percentage intact mucosa (E) from male Glp2r^+/+, Glp2r^+/−, and Glp2r^−/− mice treated with water alone (0% DS) or oral dextran sulfate (3% DS) in the drinking water for 9 d (n = 4 – 8 per group). ***, P < 0.001 0% vs. 3% DS-treated mice. Similar results were obtained in studies of female mice of the same genotypes (data not shown).

FIG. 3

Indomethacin and irinotecan produce increased small bowel injury in Glp2r^−/− mice. Proportion of moribund (A), percent survival (B), and percent positive (C) bacterial cultures in male Glp2r^+/+, Glp2r^+/−, and Glp2r^−/− mice (n = 17–28 per group) after two sc injections of vehicle or indomethacin. *, P < 0.05, Glp2r^−/− vs. Glp2r^+/+ (for A and C). Jejunal crypt depth (D) and crypt density (E) and prevalence of hemolytic activity in blood (F), a reflection of bacterial septicemia, in male Glp2r^−/− vs. Glp2r^+/+ mice (n = 5– 6 per group) 24 h after two ip injections of vehicle or irinotecan (IRT). **, P < 0.01, ***, P < 0.001 IRT vs. vehicle; #, P < 0.05, Glp2r^−/− vs. Glp2r^+/+.

FIG. 4

Defective gene expression in response to irinotecan (IRT) in the small bowel of Glp2r^−/− mice. Relative jejunal mRNA levels of lgr5 (leucine rich repeat-containing G protein coupled receptor 5) (A), epgn (epigen) (B), egf (epidermal growth factor) (C), erbB1 (epidermal growth factor receptor) (D), igf-1r (IGF-I receptor), (E) and il-6 (IL-6) (F) from Glp2r^+/+ and Glp2r^−/− male mice (n = 4 – 6 per group) treated with IRT. *, P < 0.05, IRT vs. vehicle; #, P < 0.05, Glp2r^−/− vs. Glp2r^+/+.

FIG. 5

Intestinal bacterial load, gene expression profiles, and intestinal bactericidal activity in Glp2r^−/− mice. A, Small intestinal aerobic bacterial load from Glp2r^+/+, Glp2r^+/−, and Glp2r^−/− male mice (n = 11–13 per group). Each data point corresponds to one mouse. *, P < 0.05, and #, P < 0.01 for Glp2r^−/− vs. Glp2r^+/+ and Glp2r^+/−, respectively. Relative mRNA levels of Paneth cell markers (B) and inflammation-related genes (C) in jejunum and ileum of male Glp2r^−/− mice and wild-type littermate controls (n = 7–9 per group). *, P < 0.05, **, P < 0.01 Glp2r^−/− vs. Glp2r^+/+. D, Small intestinal crypts isolated from Glp2r^+/+ and Glp2r^−/− male mice (n = 4 –10 per group) were stimulated with carbamylcholine (CCH) or LPS and the supernatants assayed for bactericidal activity. *, P < 0.05, Glp2r^−/− vs. Glp2r^+/+.

FIG. 6

16S rRNA sequencing reveals differential expression of the gut microbiome in the feces of Glp2r^−/− mice. Relative abundance of Bacteroidetes (A), Firmicutes (B), Actinobacteria (C), Proteobacteria (D), and Clostridiales (E) in the cecum and feces of male Glp2r^−/− mice and wild-type littermates (n = 5– 6 per group). *, P < 0.05, **, P < 0.01, ***, P < 0.001, Glp2r^−/− mice vs. wild-type littermates.