Insulin signaling in alpha cells modulates glucagon secretion in vivo

Abstract

Glucagon plays an important role in glucose homeostasis by regulating hepatic glucose output in both normo- and hypoglycemic conditions. In this study, we created and characterized alpha cell-specific insulin receptor knockout (alphaIRKO) mice to directly explore the role of insulin signaling in the regulation of glucagon secretion in vivo. Adult male alphaIRKO mice exhibited mild glucose intolerance, hyperglycemia, and hyperglucagonemia in the fed state and enhanced glucagon secretion in response to L-arginine stimulation. Hyperinsulinemic-hypoglycemic clamp studies revealed an enhanced glucagon secretory response and an abnormal norepinephrine response to hypoglycemia in alphaIRKO mice. The mutants also exhibited an age-dependent increase in beta cell mass. Furthermore, siRNA-mediated knockdown of insulin receptor in glucagon-secreting InR1G cells promoted enhanced glucagon secretion and complemented our in vivo findings. Together, these data indicate a significant role for intraislet insulin signaling in the regulation of alpha cell function in both normo- and hypoglycemic conditions.

Figure 1. α-cell specific recombination, fed hyperglycemia and hyperglucagonemia in αIRKO mice

(A) Two individual islets from pancreas sections of glucagon-Cre/ROSA26-LacZ mice are shown. Bar= 100 μm. Magnification: X40. (B) Recombination of insulin receptor in hypothalamus assessed by RT-PCR. Positive control, βIRKO β-cells; Negative control, LoxLox β-cells. (C) Food intake in 2 and 6 month-old male mice. n=6 in each group. (D) Body weights in 6 month-old male and female mice. n=8–12 in each group. (E) Blood glucose, (F) plasma insulin, and (G) plasma glucagon were measured after an overnight (16 h) fast or in random-fed states in 2, 5, and 12 month-old males. n=6–8 in each group. (H) Insulin/glucagon ratio. Control: empty bar; αIRKO: filled bar. Data are expressed as means ± SEM, *, p<0.05; control versus αIRKO.

Figure 2. Mild glucose intolerance and enhanced glucagon secretion in αIRKO mice

(A) Glucose tolerance in 2 and 5 month-old male mice. n=3–12 in each group. (B) Whole body insulin sensitivity (at insulin doses, 0.5 and 1 U/kg BW) in 2 and 5 month-old male mice. n=3–5 in each group. (C) Insulin and glucagon responses to L-arginine were assessed by intra-peritoneal L-arginine stimulation test in 2 and 5–6 month-old male mice. n=4–6 in each group. (D) Plasma glucagon was assessed in the random fed state before and after streptozotocin treatment. n=6–7 in each group before treatment, n=3–4 after treatment. (E, F) Insulin and (G, H) glucagon secretion were examined by ex vivo whole pancreas perfusion. Control: empty square; αIRKO: filled triangle. Data are expressed as means ± SEM, *, p<0.05; control versus αIRKO.

Figure 3. Hyperinsulinemic-hypoglycemic clamp and counter-regulatory responses in αIRKO mice

The mice were continuously infused with insulin and glucose to maintain hypoglycemia. (A) Blood glucose and (B) glucose infusion rate were measured every 10 min. At 0, 30 and 120 min of the clamp experiment, (C) glucagon, (D) norepinephrine, (E) epinephrine and (F) corticosterone response were examined. n=4–5 in each group. Control: empty circle, αIRKO: filled triangle. Data are expressed as means ± SEM. *, p<0.05; control versus αIRKO.

Figure 4. Glycemic parameters and hepatic gene expression in αIRKO mice

The mice were subjected to 36 h fasting followed with 12 h re-feeding. (A) Blood glucose and (B) body weight were measured every 12 h. n=4–8 in each group. (C) Plasma glucagon and (D) insulin were measured in the fed, 36 h fast, and 12 h re-fed conditions. n=4–8 in each group. Control: empty square, αIRKO: filled triangle. Data are expressed as means ± SEM. *, p<0.05; control versus αIRKO. Hepatic expression of (E) PEPCK, (F) G6Pase, (G) glucokinase, and (H) fatty acid synthase genes were quantified by real-time PCR and normalized to TBP. n=4 in each group. Control: empty bar, αIRKO: filled bar. Data are expressed as means ± SEM, *, p<0.05, **, p<0.01; between indicated groups.

Figure 5. Impact of α-cell specific insulin receptor disruption on insulin and glucagon gene expression

(A, C) Effect of incubating islets from low (3 mM) to high (11.1 mM) glucose; or (B, D) high (11.1 mM) to low (3 mM) glucose on glucagon (A, B) or insulin gene expression (C, D). Fold changes were calculated relative to controls. n=3 each in each group. Control: empty bar, αIRKO: filled bar. Data are expressed as means ± SEM. *, p<0.05; versus 3 mM -3 mM control group (C), versus 11 mM - 11 mM control group (D). #, p<0.05; versus 3 mM - 3 mM αIRKO group (A, C), versus 11 mM - 11 mM αIRKO group (B, D).

Figure 6. Pancreas morphometry in αIRKO mice

Pancreas sections were immunostained as indicated (A, C, D; Bar= 50 μm. Magnification: X40). (B) β- and α-cell mass were quantified at 6 and 12 month-old. n=4–5 in each group. Somatostatin, ghrelin, PDX-1 and MafB were immunostained in 12 month-old controls and mutants. Control: empty bar, αIRKO: filled bar. Data are expressed as means ± SEM, *, p<0.05; control versus αIRKO.

Figure 7. Enhanced glucagon secretion in InR1G cells with insulin receptor knockdown

(A) Western blots for insulin receptor, total and phospho-Akt and FoxO1 and normalized for actin. Representative of 3 independent experiments. (B) Glucagon secretion was assessed by static incubation for 60 min and expressed per mg of total protein. n=6 in each group. (C) Total protein content and total glucagon content. n=3 in each group. Empty bar: scrambled siRNA; Filled bar: siRNA for insulin receptor. Data are expressed as means ± SEM, *, p<0.05, **, p<0.01; scramble versus siRNA for insulin receptor.