The glucagon-like peptide-1 receptor regulates endogenous glucose production and muscle glucose uptake independent of its incretin action

Abstract

Glucagon-like peptide-1 (GLP-1) diminishes postmeal glucose excursions by enhancing insulin secretion via activation of the beta-cell GLP-1 receptor (Glp1r). GLP-1 may also control glucose levels through mechanisms that are independent of this incretin effect. The hyperinsulinemic-euglycemic clamp (insulin clamp) and exercise were used to examine the incretin-independent glucoregulatory properties of the Glp1r because both perturbations stimulate glucose flux independent of insulin secretion. Chow-fed mice with a functional disruption of the Glp1r (Glp1r(-/-)) were compared with wild-type littermates (Glp1r(+/+)). Studies were performed on 5-h-fasted mice implanted with arterial and venous catheters for sampling and infusions, respectively. During insulin clamps, [3-(3)H]glucose and 2[(14)C]deoxyglucose were used to determine whole-body glucose turnover and glucose metabolic index (R(g)), an indicator of glucose uptake. R(g) in sedentary and treadmill exercised mice was determined using 2[(3)H]deoxyglucose. Glp1r(-/-) mice exhibited increased glucose disappearance, muscle R(g), and muscle glycogen levels during insulin clamps. This was not associated with enhanced muscle insulin signaling. Glp1r(-/-) mice exhibited impaired suppression of endogenous glucose production and hepatic glycogen accumulation during insulin clamps. This was associated with impaired liver insulin signaling. Glp1r(-/-) mice became significantly hyperglycemic during exercise. Muscle R(g) was normal in exercised Glp1r(-/-) mice, suggesting that hyperglycemia resulted from an added drive to stimulate glucose production. Muscle AMP-activated protein kinase phosphorylation was higher in exercised Glp1r(-/-) mice. This was associated with increased relative exercise intensity and decreased exercise endurance. In conclusion, these results show that the endogenous Glp1r regulates hepatic and muscle glucose flux independent of its ability to enhance insulin secretion.

Figure 1

Insulin clamp experiments in 5-h-fasted mice. Arterial glucose (A) and GIR (B) during insulin clamps are presented for Glp1r^+/+ (black squares) and Glp1r^−/− (white squares) mice. R_d (C), muscle (gastrocnemius) glycogen (D), endoR_a (E), and liver glycogen (F) from 5-h-fasted (basal, black bars) mice and mice from clamp experiments (insulin clamp, white bars) are presented. Data are mean ± sem for 11–12 mice/genotype. *, P < 0.05 vs. Glp1r^+/+; †, P < 0.05 vs. basal within the same genotype.

Figure 2

R_g after insulin clamp experiments in 5-h-fasted mice. R_g in soleus (A), gastrocnemius (B), SVL (C), diaphragm (D), and heart (E) are presented for Glp1r^+/+ (black bars) and Glp1r^−/− (white bars) mice. Data are mean ± sem for 11–12 mice/genotype. *, P < 0.05 vs. Glp1r^+/+.

Figure 3

Immunoblots for insulin signaling proteins after insulin clamp experiments in 5-h-fasted mice. Representative immunoblots from gastrocnemius muscle (A) and liver (D) extracts are shown. Protein content and protein activation is presented for Glp1r^+/+ (black bars) and Glp1r^−/− (white bars) mice. Protein content of Ser⁴⁷³-phosphorylated Akt (P-Akt), total Akt, Ser⁹-phosphorylated GSK-3β (P-GSK-3β), and total GSK-3β (GSK-3β) is shown from muscle (B) and liver (E) extracts. Protein content is normalized to the GAPDH loading control. Akt activation (P-Akt/Akt) and GSK-3β inactivation (P-GSK-3β/GSK-3β) are also shown from muscle (C) and liver (F) extracts. Inactivation of GSK-3β promotes glycogen synthesis. Data are mean ± sem for 11–12 mice/genotype. *, P < 0.05 vs. Glp1r^+/+.

Figure 4

Exercise experiments in 5-h-fasted mice. Arterial glucose (A) levels are presented for Glp1r^+/+ (black diamonds and squares) and Glp1r^−/− (white diamonds and squares) mice that remained sedentary (diamonds) or underwent treadmill exercise (squares) for 30 min. Insulin (B), NEFA (C), muscle (gastrocnemius) glycogen (D), and liver glycogen (E) levels in sedentary (black bars) and exercised (white bars) Glp1r^+/+ and Glp1r^−/− mice are shown. Data are mean ± sem for 14–16 mice/genotype. *, P < 0.05 vs. Glp1r^+/+ exercise; †, P < 0.05 vs. Glp1r^−/− sedentary.

Figure 5

R_g and K_g in 5-h-fasted sedentary (black bars) and exercised (white bars) Glp1r^+/+ and Glp1r^−/− mice. R_g in soleus, gastrocnemius, SVL, diaphragm, and heart (A–E) is shown. K_g is shown for the same tissues (F–J). Data are mean ± sem for 14–16 mice/genotype. *, P < 0.05 vs. Glp1r^+/+ within the same experimental condition; †, P < 0.05 vs. sedentary within the same genotype.

Figure 6

Immunoblots for AMPK in 5-h-fasted sedentary and exercised Glp1r. Representative immunoblots are shown from gastrocnemius muscle (A) and liver (D) extracts. Protein content and protein activation is presented for Glp1r^+/+ (black bars) and Glp1r^−/− (white bars) mice. Protein content of Thr¹⁷²-phosphorylated AMPK (P-AMPK), and α1/α2 AMPK (AMPK) is shown from muscle (B) and liver (E) extracts from exercised mice. Protein content is normalized to the GAPDH loading control. AMPK activation (P-AMPK/AMPK) is also shown from muscle (C) and liver (F) extracts from exercised mice. Data are mean ± sem for eight to 10 mice/genotype. *, P < 0.05 vs. Glp1r^+/+.