Pharmacological regulation of insulin secretion in MIN6 cells through the fatty acid receptor GPR40: identification of agonist and antagonist small molecules

Abstract

1. Long chain fatty acids have recently been identified as agonists for the G protein-coupled receptors GPR40 and GPR120. Here, we present the first description of GW9508, a small-molecule agonist of the fatty acid receptors GPR40 and GPR120. In addition, we also describe the pharmacology of GW1100, a selective GPR40 antagonist. These molecules were used to further investigate the role of GPR40 in glucose-stimulated insulin secretion in the MIN6 mouse pancreatic beta-cell line. 2. GW9508 and linoleic acid both stimulated intracellular Ca2+ mobilization in human embryonic kidney (HEK)293 cells expressing GPR40 (pEC50 values of 7.32+/-0.03 and 5.65+/-0.06, respectively) or GPR120 (pEC50 values of 5.46+/-0.09 and 5.89+/-0.04, respectively), but not in the parent HEK-293 cell line. 3. GW1100 dose dependently inhibited GPR40-mediated Ca2+ elevations stimulated by GW9508 and linoleic acid (pIC50 values of 5.99+/-0.03 and 5.99+/-0.06, respectively). GW1100 had no effect on the GPR120-mediated stimulation of intracellular Ca2+ release produced by either GW9508 or linoleic acid. 4. GW9508 dose dependently potentiated glucose-stimulated insulin secretion in MIN6 cells, but not in primary rat or mouse islets. Furthermore, GW9508 was able to potentiate the KCl-mediated increase in insulin secretion in MIN6 cells. The effects of GW9508 on insulin secretion were reversed by GW1100, while linoleic acid-stimulated insulin secretion was partially attenuated by GW1100. 5. These results add further evidence to a link between GPR40 and the ability of fatty acids to acutely potentiate insulin secretion and demonstrate that small-molecule GPR40 agonists are glucose-sensitive insulin secretagogues.

Figure 1

Chemical structures of the GPR40/GPR120 agonist GW9805 and selective GPR40 antagonist GW1100.

Figure 2

Pharmacological profile of the GW9508 and GW1100 at the GPR40 receptor. (a) The effect of GW9508, linoleic acid, α-linolenic acid, palmitoleic acid and cis-5,8,11,14,17-eicosapentaenoic acid on intracellular Ca²⁺ levels was examined in HEK-293 cells expressing the human GPR40 receptor. pEC₅₀ values obtained were GW9508 (pEC₅₀=7.32±0.03; n=4), linoleic acid (pEC₅₀=5.65±0.06; n=4), α-linolenic acid (pEC₅₀=5.94±0.03; n=4), palmitoleic acid (pEC₅₀=5.86±0.05; n=4) and cis-5,8,11,14,17-eicosapentaenoic acid (pEC₅₀=5.73±0.04; n=4). (b) Inhibition of GW9508 and linoleic acid induced intracellular Ca²⁺ increase by GW1100, a selective GPR40 receptor antagonist. GW1100 produced a concentration-dependent decrease in intracellular Ca²⁺ response to an EC₈₀ concentration of GW9508 (pIC₅₀=5.99±0.03; n=4) and linoleic acid (pIC₅₀₌5.99±0.06; n=4). (c) The effect of increasing concentrations of GW1100 on GW9508-induced luciferase activity in a human GPR40-5 × Gal/Gal4-Elk1-luc⁺ CHO cell line reporter assay. Cells were preincubated with GW1100 for 15 min before addition of GW9508. GW1100 produced a concentration-dependent rightward shift in the GW9508 curve followed by a depression of the signal. This is consistent with a non-competitive mode of antagonism in the presence of a significant GPR40 receptor reserve. (d) The selectivity of GW1100 in the human GPR40-5 × Gal/Gal4-Elk1-luc⁺ CHO cell line reporter assay was confirmed using Thrombin as a control stimulus. Data represent the mean±s.e.m.

Figure 3

Pharmacological profile of the GW9508 and GW1100 at the GPR120 receptor. (a) linoleic acid, α-linolenic acid, palmitoleic acid and cis-5,8,11,14,17-eicosapentaenoic acid on intracellular Ca²⁺ levels was examined in HEK-293 cells expressing the human GPR120 receptor. pEC₅₀ values obtained were GW9508 (pEC₅₀=5.46±0.09; n=4), linoleic acid (pEC₅₀=5.89±0.04; n=4), α-linolenic acid (pEC₅₀=6.26±0.04; n=4), palmitoleic acid (pEC₅₀=6.12±0.08; n=4) and cis-5,8,11,14,17-eicosapentaenoic acid (pEC₅₀=5.68±0.06; n=4). (b) The effect of increasing concentrations of GW1100 on GW9508 induced increases in intracellular Ca²⁺ levels. Cells were preincubated with GW1100 for 15 min before addition of GW9508. GW1100 had no effect on the GW9508 response. (c) The effect of increasing concentrations of GW1100 on linoleic acid induced increases in intracellular Ca²⁺ levels. Cells were preincubated with GW1100 for 15 min before addition of linoleic acid. GW1100 had no effect on the linoleic acid response. Data represent the mean±s.e.m.

Figure 4

Effect of GW9508 and GW1100 on insulin secretion from the MIN6 mouse insulinoma cell line. (a) GW9508 produced a concentration-dependent increase in glucose-stimulated insulin secretion in the presence of 25 mM glucose (pEC₅₀=6.14±0.03; ^**P<0.01 vs 25 mM glucose Control; n=3). No effect of GW9508 was observed in the presence of 5 mM glucose. (b) The effect of GW9508 on insulin secretion in the presence of 25 mM glucose was significantly reduced in the presence of 1 μM GW1100 (^**P<0.01 vs GW9508 alone; n=3). (c) Glucose dependency of GW9508 effect on insulin secretion. Insulin secretion was significantly enhanced in the presence of GW9508 (10 μM) at higher glucose concentrations compared with lower glucose concentrations (^**P<0.01; ^***P<0.001; n=3). (d) Linoleic acid produced a concentration-dependent increase in glucose-stimulated insulin secretion in the presence of 25 mM glucose (^*P<0.05; ^**P<0.01 vs 25 mM glucose Control; n=4). A significantly smaller increase in insulin secretion was observed for linoleic acid in the presence of 5 mM glucose. (e) The effect of linoleic acid on insulin secretion in the presence of 25 mM glucose was significantly reduced in the presence of 1 μM GW1100 (^**P<0.01; ^***P<0.001 vs linoleic acid alone; n=4). (f) GW9508 and linoleic acid effects on glucose-stimulated (25 mM) insulin secretion are additive (^**P<0.01; ^***P<0.001 vs linoleic acid alone; n=4). Data represent the mean±s.e.m.

Figure 5

Effect of KCl on GW9508-potentiatiation of insulin secretion. Potentiation of KCl-stimulated insulin secretion in the presence of GW9508 at 25 mM glucose. In the presence of 25 mM glucose, insulin release was significantly potentiated by either 10 μM GW9508 or 30 mM KCl ^**P<0.01 vs 25 mM glucose-treated MIN6 cells;n=3). The presence of GW9508 (10 μM) significantly enhanced KCl or GW9508-stimulated insulin secretion in the presence of 25 mM glucose (⁺⁺P<0.01 vs KCl or GW9508-treated MIN6 cells; n=3). Addition of the antagonist GW1100 (1 μM) prevented the effect of GW9508 on KCl or GW9508-mediated insulin release (^##P<0.01 vs GW9508 and KCl-treated MIN6 cells; n=3). Data represent mean±s.e.m. Analysis of results was by ANOVA, followed by Tukey–Kramer multiple comparison test.