Sildenafil amplifies calcium influx and insulin secretion in pancreatic β cells

Abstract

Sildenafil, a phosphodiesterase-5 (PDE5) inhibitor, has been shown to improve insulin sensitivity in animal models and prediabetic patients. However, its other metabolic effects remain poorly investigated. This study examines the impact of sildenafil on insulin secretion in MIN6-K8 mouse clonal β cells. Sildenafil amplified insulin secretion by enhancing Ca²⁺ influx. These effects required other depolarizing stimuli in MIN6-K8 cells but not in K_ATP channel-deficient β cells, which were already depolarized, indicating that sildenafil-amplified insulin secretion is depolarization-dependent and K_ATP channel-independent. Interestingly, sildenafil-amplified insulin secretion was inhibited by pharmacological inhibition of R-type channels, but not of other types of voltage-dependent Ca²⁺ channels (VDCCs). Furthermore, sildenafil-amplified insulin secretion was barely affected when its effect on cyclic GMP was inhibited by PDE5 knockdown. Thus, sildenafil stimulates insulin secretion and Ca²⁺ influx through R-type VDCCs independently of the PDE5/cGMP pathway, a mechanism that differs from the known pharmacology of sildenafil and conventional insulin secretory pathways. Our results reposition sildenafil as an insulinotropic agent that can be used as a potential antidiabetic medicine and a tool to elucidate the novel mechanism of insulin secretion.

Keywords: diabetes; insulin secretion; pancreatic β cells; sildenafil; voltage‐dependent calcium channels.

FIGURE 1

Sildenafil amplifies insulin secretion from β cell lines in a depolarization‐dependent manner. (a) Dose‐dependent effects of sildenafil on insulin secretion at 11.1 mM glucose in MIN6‐K8 cells. n = 4. (b) Effects of sildenafil on insulin secretion at 2.8 mM glucose in MIN6‐K8 and Kcnj11 ^−/− β cells. n = 4. (c) Effect of sildenafil on insulin secretion in the presence of glimepiride at 2.8 mM glucose in MIN6‐K8 cells. n = 4. (d) Effect of diazoxide on sildenafil‐amplified insulin secretion in MIN6‐K8 cells. n = 4. Data were presented as mean ± standard deviation (SD). 2.8G, 2.8 mM glucose; 11.1G, 11.1 mM glucose. The reagents were added to achieve the following final concentrations unless otherwise specified: sildenafil (Sil)—100 μM, glimepiride (GLM)—1 μM, and diazoxide—100 μM. Statistical comparisons were performed using Welch's one‐way ANOVA with Dunnett's post hoc test in (a–c), and Welch's unpaired two‐tailed t‐test for (d).

FIGURE 2

Sildenafil potentiates the influx of extracellular Ca²⁺. Intracellular Ca²⁺ levels were measured using Fluo‐4. The time course of normalized fluorescence intensity at 508–579 nm is indicated in (a), (c), (e), and (g). The black arrow indicates the addition of the indicated stimulations at time = 0. The magnitude of Ca²⁺ responses was quantified as iAUC in (b), (d), and (f). (a, b). Effect of sildenafil on intracellular Ca²⁺ in MIN6‐K8 cells. 2.8G + Sil: n = 8; 11.1G: n = 12; 11.1G + Sil: n = 11. (c, d) Effect of sildenafil on intracellular Ca²⁺ in Kcnj11 ^−/− β cells. Carbachol was added at time = 600 as positive control. 11.1G + Nife: n = 13; 11.1G + Nife + Sil: n = 11. (e, f) Effect of nifedipine on sildenafil‐induced Ca²⁺ response in MIN6‐K8 cells. 11.1G + Nife: n = 13; 11.1G + Nife + Sil: n = 11. (g) Effect of sildenafil on intracellular Ca²⁺ under extracellular Ca²⁺‐free conditions in MIN6‐K8 cells. 2.8G + Sil: n = 8; 11.1G: n = 12; 11.1G + Sil: n = 11. Data are presented as mean ± SD. SD is indicated by shaded regions in (a), (c), (e), and (g), and by error bars elsewhere. 2.8G, 2.8 mM glucose; 11.1G, 11.1 mM glucose. The reagents were added to achieve the following final concentrations: sildenafil (Sil)—100 μM, nifedipine (Nife)—10 μM, carbachol (CCh)—50 μM, and EGTA—0.2 mM. Statistical comparisons were made using Welch's one‐way ANOVA with Dunnett's post hoc test in (b), and Welch's unpaired two‐tailed t‐test for (d) and (f).

FIGURE 3

Sildenafil‐amplified insulin secretion is dependent on Ca²⁺ influx through R‐type VDCCs. (a, b) Effects of nifedipine or extracellular Ca²⁺‐free conditions on sildenafil‐amplified insulin secretion. n = 4. The data are presented in their original value in (a) and as fold change over 11.1 mM glucose in (b). (c, d) Effects of SNX‐482 and conotoxin (CTx)‐MVIIC on sildenafil‐amplified insulin secretion. n = 4. The data are presented in their original value in (c) and as fold change over 11.1 mM glucose in (d). (e, f) Effect of thapsigargin on sildenafil‐amplified insulin secretion. n = 4. The data are presented in their original value in (e) and as fold change over 11.1 mM glucose in (f). All experiments were performed using MIN6‐K8 cells. Data were presented as mean ± SD. The reagents were added to achieve the following final concentrations: sildenafil (Sil)—100 μM, nifedipine (Nife)—10 μM, EGTA—0.2 mM, SNX‐482—500 nM, CTx‐MVIIC—2.5 μM, and thapsigargin—1 μM. Statistical comparisons were performed using Welch's unpaired two‐tailed t‐test for (a), (c), and (f), and Welch's one‐way ANOVA with Dunnett's post hoc test for (b) and (d).

FIGURE 4

PDE5 is not involved in sildenafil‐amplified insulin secretion. (a) Knockdown efficiency of Pde5a assessed by RT‐qPCR. mRNA levels were normalized to siNT (non‐targeting siRNA)‐treated cells. n = 3. (b, c) Knockdown efficiency of PDE5A protein assessed by immunoblotting. Two PDE5A isoforms (isoforms 1 and 2) were detected and are indicated by black triangles. The average intensity of the two PDE5A isoforms was normalized to α‐tubulin and expressed as fold change in (c). n = 6. (d). Effect of Pde5a knockdown on the regulation of intracellular cGMP by sildenafil. Whole‐cell cGMP levels were normalized to total protein. n = 7. (e, f) Effect of Pde5a knockdown on sildenafil‐amplified insulin secretion. n = 4. The data are presented in their original value in (e) and as fold change over 11.1 mM glucose in (f). All experiments were performed using MIN6‐K8 cells. Data were presented as mean ± SD. Sildenafil (Sil)—100 μM. Statistical comparisons were made using Welch's unpaired two‐tailed t‐test for (c) and (f), and Welch's one‐way ANOVA with Dunnett's post hoc test for (d).