Transformation of postingestive glucose responses after deletion of sweet taste receptor subunits or gastric bypass surgery

Abstract

The glucose-dependent secretion of the insulinotropic hormone glucagon-like peptide-1 (GLP-1) is a critical step in the regulation of glucose homeostasis. Two molecular mechanisms have separately been suggested as the primary mediator of intestinal glucose-stimulated GLP-1 secretion (GSGS): one is a metabotropic mechanism requiring the sweet taste receptor type 2 (T1R2) + type 3 (T1R3) while the second is a metabolic mechanism requiring ATP-sensitive K(+) (K(ATP)) channels. By quantifying sugar-stimulated hormone secretion in receptor knockout mice and in rats receiving Roux-en-Y gastric bypass (RYGB), we found that both of these mechanisms contribute to GSGS; however, the mechanisms exhibit different selectivity, regulation, and localization. T1R3(-/-) mice showed impaired glucose and insulin homeostasis during an oral glucose challenge as well as slowed insulin granule exocytosis from isolated pancreatic islets. Glucose, fructose, and sucralose evoked GLP-1 secretion from T1R3(+/+), but not T1R3(-/-), ileum explants; this secretion was not mimicked by the K(ATP) channel blocker glibenclamide. T1R2(-/-) mice showed normal glycemic control and partial small intestine GSGS, suggesting that T1R3 can mediate GSGS without T1R2. Robust GSGS that was K(ATP) channel-dependent and glucose-specific emerged in the large intestine of T1R3(-/-) mice and RYGB rats in association with elevated fecal carbohydrate throughout the distal gut. Our results demonstrate that the small and large intestines utilize distinct mechanisms for GSGS and suggest novel large intestine targets that could mimic the improved glycemic control seen after RYGB.

Fig. 1.

Impaired glucose tolerance and reduced insulin responses in type 1 taste receptor 3 (T1R3)-null (−/−), but not type 1 taste receptor 2 (T1R2)-null, mice. A: blood glucose during an oral glucose tolerance test (OGTT) for T1R3^+/+ (n = 7; black, solid line), T1R3^+/− (n = 10; gray, solid line), and T1R3^−/− (n = 11; black, broken line) mice. Repeated-measures ANOVA: P = 0.01 (glucose); ANOVA (40 min glucose): P = 0.02; ANOVA (60 min glucose): P = 0.0005; ANOVA (90 min glucose): P = 0.0007; Scheffé's post hoc: *P < 0.05 (−/− vs. +/+ or +/−); #P < 0.05 (−/− vs. +/−). B: glucose area under the curve (AUC) for A; ANOVA: P = 0.002; Scheffé's post hoc: *P < 0.05. C: incremental plasma insulin levels during the same OGTT as in A. Repeated-measures ANOVA: P = 0.01; ANOVA (10 min insulin): P = 0.001; ANOVA (20 min insulin): P = 0.0007; ANOVA (90 min insulin): P = 0.01; Scheffé's post hoc: *P < 0.05 (−/− vs. +/+ or +/−); #P < 0.05 (−/− vs. +/−). D: insulin AUC for C; ANOVA: P = 0.0005; Scheffé's post hoc: *P < 0.05. E: blood glucose during an ip glucose tolerance test (IPGTT) for T1R3^+/+ (n = 4), T1R3^+/− (n = 11), and T1R3^−/− (n = 6) mice. ANOVA: P = 0.19. F: glucose AUC for E; repeated-measures ANOVA: P = 0.08. G: incremental plasma insulin levels during the same IPGTT as in E; repeated-measures ANOVA: P = 0.47. H: insulin AUC for G; ANOVA: P < 0.01; Scheffé's post hoc: *P < 0.05. I: blood glucose levels during an OGTT for T1R2^+/+ (n = 4; solid line) and T1R2^−/− (n = 4; broken line) mice; repeated-measures ANOVA: P = 0.77. J: glucose AUC for I; ANOVA: P = 0.18. K: incremental plasma insulin levels during the same OGTT as I; repeated-measures ANOVA: P = 0.25. L: insulin AUC for K; ANOVA: P = 0.45. All values are means ± SE. Mice aged 12–20 wk.

Fig. 2.

Disruption of glucose-stimulated glucagon-like peptide-1 (GLP-1) secretion from small intestine of T1R3^−/− and T1R2^−/− mice. A–C: ELISA-based measurements of GLP-1 secretion from intestinal explants of duodenum (A), jejunum (B), and ileum (C) of T1R3^+/+ (black) or T1R3^−/− (white) mice upon stimulation with buffer (B), 250 mM glucose (G), 100 mM sucralose (S), or 500 mM fructose (F). D–F: ELISA-based measurements of GLP-1 secretion from intestinal explants of duodenum (D), jejunum (E), and ileum (F) of T1R2^+/+ (black) or T1R2^−/− (white) mice upon stimulation with buffer or 250 mM glucose. Each bar, n = 4 mice. Data are presented as means ± SE. *P < 0.01 vs. buffer for that genotype, except for comparison of buffer and glucose treatments in T1R2^−/− ileum (F; P = 0.01).

Fig. 3.

Upregulation of glucose-stimulated GLP-1 secretion from large intestine explants of T1R3^−/− but not T1R2^−/− mice. A and B: ELISA-based measurements of GLP-1 secretion from intestinal explants of colon (A) or rectum (B) of T1R3^+/+ (black) or T1R3^−/− (white) mice upon stimulation with buffer (B), 250 mM glucose (G), 100 mM sucralose (S), or 500 mM fructose (F). C and D: ELISA-based measurements of GLP-1 secretion from intestinal explants of colon (C) or rectum (D) of T1R2^+/+ (black) or T1R2^−/− (white) mice upon stimulation with buffer or 250 mM glucose. Each bar, n = 4 mice. Data are presented as means ± SE. *P < 0.01 vs. buffer for that genotype. #P < 0.01 vs. same stimulus in T1R3^+/+ mice.

Fig. 4.

ATP-dependent K⁺ (K_ATP) channel closure is required for glucose-stimulated GLP-1 secretion in colon but not ileum. A and B: ELISA-based measurements of GLP-1 secretion from intestinal explants of ileum (A) and colon (B) of T1R3^+/+ (black) and T1R3^−/− (white) mice upon treatment with buffer (B), 250 mM glucose (G), or 50 μM glibenclamide (Gb). Each bar, n = 4 mice. Data are presented as means ± SE. *P < 0.01 vs. buffer for that genotype. C: ELISA-based measurements of GLP-1 secretion from colon explants from T1R3^+/+ (black) and T1R3^−/− (white) mice upon stimulation with buffer, 250 mM glucose, or 250 mM glucose in combination with 50 μM glibenclamide (G + Gb), 100 μM diazoxide (G + Dz), or 10 μM cromokalim (G + Cr). Each bar, n = 4 mice. Data are presented as means ± SE. *P < 0.01 vs. buffer for that genotype, except for comparison between buffer and glucose + cromokalim treatments in T1R3^+/+ colon (P = 0.03, B). #P < 0.01 vs. glucose treatment for that genotype.

Fig. 5.

Carbohydrate malabsorption in T1R3^−/−, but not T1R2^−/−, mice. A: carbohydrate content of feces from ileum, colon, and rectum of T1R3^+/+ (black) and T1R3^−/− (white) mice. Each bar, n = 10 mice. *P < 0.01. B: carbohydrate content of feces from ileum, colon, and rectum of T1R2^+/+ (black) and T1R2^−/− (white) mice. Each bar, n = 8 mice. No significant differences were found.

Fig. 6.

Altered insulin secretion kinetics in pancreatic islets of T1R3^−/− mice. A and B: total internal reflection fluorescence (TIRF) images of yellow fluorescent protein (YFP)-labeled secretory granules in pancreatic islets isolated from T1R3^+/+ (A) and T1R3^−/− (B) mice 24 h after gene delivery by adenovirus. Scale bar, 5 μm. C: sequential images of representative membrane fusions of individual YFP-VAMP2-labeled secretory granules in T1R3^+/+ (top) and T1R3^−/− (bottom) islets upon glucose (10 mM) stimulation. Scale bar, 1 μm. D: representative change in fluorescence intensity of single granules from T1R3^+/+ (black) and T1R3^−/− (gray) islets upon glucose stimulation. The fusion rates, as determined by the slope of the fluorescence increase upon stimulation, are near the mean for the entire data set, shown in F. E: representative change in fluorescence intensity of single granules from T1R2^+/+ (black) and T1R2^−/− (gray) islets upon glucose stimulation. The fusion rates, as determined by the slope of the fluorescence increase upon stimulation, are near the mean for the entire data set (F). F: mean fusion rate (±SE) for individual granules upon islet stimulation with glucose (10 mM; n ≥ 50 granules, taken from at least 10 cells) or KCl (30 mM; n = 25 granules). ANOVA: P < 0.001. *P < 0.05 (Scheffé's post hoc test).

Fig. 7.

Robust glucose-stimulated GLP-1 secretion in colon of Roux-en-Y gastric bypass (RYGB) rats. A: carbohydrate content of feces from ileum, colon, and rectum of sham (black) and RYGB (white) surgery rats. Each bar, n = 5 rats. *P < 0.01. B and C: ELISA-based measurements of GLP-1 secretion from intestinal explants of ileum (B) or colon (C) of sham (black) or RYGB (white) rats upon treatment with buffer or 250 mM glucose. Each bar, n = 5 rats. *P < 0.01 vs. buffer for that same surgical manipulation. #P < 0.01 vs. same stimulus in sham controls.

Fig. 8.

Distinct mechanisms of glucose-stimulated GLP-1 secretion in small and large intestine. Enteroendocrine L cells from ileum (pink) or colon (yellow) can both secrete GLP-1 but show different selectivities for sweet stimuli and utilize different mechanisms to couple glucose detection to hormone secretion. K_ATP, ATP-sensitive K⁺ channel.