Somatostatin secreted by islet delta-cells fulfills multiple roles as a paracrine regulator of islet function

Abstract

Objective: Somatostatin (SST) is secreted by islet delta-cells and by extraislet neuroendocrine cells. SST receptors have been identified on alpha- and beta-cells, and exogenous SST inhibits insulin and glucagon secretion, consistent with a role for SST in regulating alpha- and beta-cell function. However, the specific intraislet function of delta-cell SST remains uncertain. We have used Sst(-/-) mice to investigate the role of delta-cell SST in the regulation of insulin and glucagon secretion in vitro and in vivo.

Research design and methods: Islet morphology was assessed by histological analysis. Hormone levels were measured by radioimmunoassay in control and Sst(-/-) mice in vivo and from isolated islets in vitro.

Results: Islet size and organization did not differ between Sst(-/-) and control islets, nor did islet glucagon or insulin content. Sst(-/-) mice showed enhanced insulin and glucagon secretory responses in vivo. In vitro stimulus-induced insulin and glucagon secretion was enhanced from perifused Sst(-/-) islets compared with control islets and was inhibited by exogenous SST in Sst(-/-) but not control islets. No difference in the switch-off rate of glucose-stimulated insulin secretion was observed between genotypes, but the cholinergic agonist carbamylcholine enhanced glucose-induced insulin secretion to a lesser extent in Sst(-/-) islets compared with controls. Glucose suppressed glucagon secretion from control but not Sst(-/-) islets.

Conclusions: We suggest that delta-cell SST exerts a tonic inhibitory influence on insulin and glucagon secretion, which may facilitate the islet response to cholinergic activation. In addition, delta-cell SST is implicated in the nutrient-induced suppression of glucagon secretion.

FIG. 1.

Intraislet organization of α- and β-cells from control and Sst^−/− islets. Consecutive sections of control (A and B) or Sst^−/− (C and D) mouse islets were stained for insulin (A and C) or glucagon (B and D), respectively, and are representative of sections from three different animals. (Please see http://dx.doi.org/10.2337/db08-0792 for a high-quality digital representation of this figure.)

FIG. 2.

Insulin and glucagon secretion in vivo. A: Plasma insulin levels in Sst^−/− mice and control mice 0, 2, 5, and 15 min after an intravenous glucose challenge (0.5 g/kg). Points show means ± SE for four to six separate animals. B: Plasma glucagon levels in Sst^−/− mice and control mice 0, 2, and 5 min after injection of intravenous arginine (0.25 g/kg). Points show means ± SE for 5 and 14 separate animals (Sst^−/− and control, respectively). C: Plasma glucose levels in Sst^−/− mice and control mice 0, 15, 30, 45, and 60 min after injection of intravenous insulin (0.4 units/kg). Points show means ± SE for five separate animals.

FIG. 3.

Hormone secretion from control and SST^−/− islets. A: Insulin secretion from control and Sst^−/− mouse islets at a substimulatory concentration of glucose (2 mmol/l G, 0–10 min) and at a stimulatory concentration of glucose (bar, 20 mmol/l G). Points show means ± SE, n = 7–8 separate perifusion channels in each experiment, typical of 10 separate experiments. In B, the same results are expressed as a percentage of total insulin content. C: Arginine-induced (20 mmol/l, bar) glucagon secretion from Sst^−/− islets and control islets. Points show means ± SE, n = 4 perifusion channels. Where no error bars are shown, they are smaller than the size of the symbols. D: Insulin secretory responses to the sulfonylurea tolbutamide (100 μmol/l, bar) in the presence of 2 mmol/l glucose from Sst^−/− islets and control islets. Points show means ± SE, n = 4 perifusion channels. E: SST secretion from control islets in response to glucose, tolbutamide, and arginine. Bars represent means ± SE, n = 8–9 in one experiment typical of three separate experiments. *P < 0.05, **P < 0.01 vs. 1 mmol/l glucose alone.

FIG. 4.

Inhibition of insulin and glucagon secretion by exogenous SST. Effect of exogenous SST (1 μmol/l) on dynamic glucose-induced insulin secretion (20 mmol/l G, bar) (A) and static arginine-induced (20 mmol/l) glucagon secretion (B) from control islets and Sst^−/− islets. Points show means ± SE, n = 4 separate perifusion channels in each experiment, typical of six separate experiments. Bars represent means ± SE, n = 6. ***P < 0.001 vs. arginine (Arg) alone.

FIG. 5.

Rate of decline in insulin secretion from control and Sst^−/− islets after removal of glucose stimulus. A: Glucose-induced (20 mmol/l G, bar) insulin secretion from Sst^−/− islets and control islets is reversed to basal levels upon removal of stimulus. B: The decline in insulin secretion from stimulated to basal levels is expressed as percent stimulated insulin secretion before removal of glucose. Points show means ± SE, n = 4 separate perifusion channels.

FIG. 6.

Parasympathetic stimulation of insulin secretion from control and Sst^−/− islets. Static insulin (A) and SST secretion (B) from control islets in response to glucose (20 mmol/l) and the cholinergic agonist CCh (500 μmol/l). Bars represent means ± SE, n = 8–9 in one experiment typical of three separate experiments. **P < 0.01 vs. 1 mmol/l glucose alone, ††P < 0.01, †††P > 0.001 vs. 20 mmol/l glucose. Data in A and B are from the same experiment. C: Effect of CCh (500 μmol/l) on dynamic glucose-induced (20 mmol/l G, bar) insulin secretion from control (○) and Sst^−/− (•) islets. Fractions were collected every 1 min between 10 and 30 min and 50 and 60 min and every 30 s between 30 and 50 min. Points show means ± SE, n = 3 perifusion channels in one experiment representative of three separate experiments. D: The effect of CCh is expressed as a percentage stimulation of the glucose-induced (20 mmol/l) secretory response (mean amplitude of secretion at time points 31–40 min expressed as a percentage of mean amplitude at 21–30 min).

FIG. 7.

Glucose-induced suppression of glucagon secretion requires SST. A: Effect of high glucose concentration (20 mmol/l) on glucagon secretion from control islets (□) and Sst^−/− islets (▪) in static incubations. Bars represent means ± SE, n = 5–7. *P < 0.05 vs. 2 mmol/l glucose. B: Glucose-induced insulin secretion from the same experiment. **P < 0.01, ***P < 0.001 vs. 2 mmol/l glucose. There was no significant increase in glucagon secretion from Sst^−/− islets in response to glucose in seven separate experiments, although in three of these, secretion appeared higher (see A). C: Glucagon secretion from control islets (□) and Sst^−/− islets (▪) in response to increasing glucose concentrations. Islets were preincubated in the absence of glucose for 1 h before exposure to the glucose concentrations shown in the figure. Bars represent means ± SE, n = 5–6. *P < 0.05 vs. 0 mmol/l glucose. D: Effect of 10 mmol/l glucose (G, bar) on dynamic glucagon secretion from control islets (○) and Sst^−/− islets (•) preincubated in the absence of glucose for 20 min. Points show means ± SE, n = 4 perifusion channels. Controls, P < 0.001 for 0 vs. 10 mmol/l glucose mean values.