Sphingosine 1-phosphate Stimulates Insulin Secretion and Improves Cell Survival by Blocking Voltage-dependent K+ Channels in β Cells

Abstract

Recent studies suggest that Sphingosine 1-phosphate (S1P) plays an important role in regulating glucose metabolism in type 2 diabetes. However, its effects and mechanisms of promoting insulin secretion remain largely unknown. Here, we found that S1P treatment decreased blood glucose level and increased insulin secretion in C57BL/6 mice. Our results further showed that S1P promoted insulin secretion in a glucose-dependent manner. This stimulatory effect of S1P appeared to be irrelevant to cyclic adenosine monophosphate signaling. Voltage-clamp recordings showed that S1P did not influence voltage-dependent Ca²⁺ channels, but significantly blocked voltage-dependent potassium (Kv) channels, which could be reversed by inhibition of phospholipase C (PLC) and protein kinase C (PKC). Calcium imaging revealed that S1P increased intracellular Ca²⁺ levels, mainly by promoting Ca²⁺ influx, rather than mobilizing intracellular Ca²⁺ stores. In addition, inhibition of PLC and PKC suppressed S1P-induced insulin secretion. Collectively, these results suggest that the effects of S1P on glucose-stimulated insulin secretion (GSIS) depend on the inhibition of Kv channels via the PLC/PKC signaling pathway in pancreatic β cells. Further, S1P improved β cell survival; this effect was also associated with Kv channel inhibition. This work thus provides new insights into the mechanisms whereby S1P regulates β cell function in diabetes.

Keywords: insulin secretion; sphingosine 1-phosphate; type 2 diabetes; voltage-dependent potassium channels; β cell.

FIGURE 1

Effects of sphingosine 1-phosphate (S1P) treatment on blood glucose and plasma insulin in vivo. The 8-wk-old C57BL/6 mice were given intraperitoneal injection of S1P (80 μg/kg) or vehicle. After 10 min, the mice were fed with glucose load (2 g/kg) followed by determination of blood glucose level (A) and plasma insulin concentration (C) at the indicated time points. The AUC values of blood glucose (B) and insulin levels (D) during the OGTT were compared, respectively (n = 7). *p < 0.05, **p < 0.01 vs control group treated with the vehicle at each time point; by unpaired two-tailed Student’s t test. The results represent by the mean ± SEM.

FIGURE 2

Effects of sphingosine 1-phosphate (S1P) on glucose-stimulated insulin secretion and intracellular cAMP production in rat islets. (A) Induction of insulin secretion by S1P under low- and high-glucose conditions (2.8 and 16.7 mM, respectively). The insulin secretion of each group was normalized by that of the control group with 2.8 mM glucose (n = 7). (B) Effect of S1P (10 μM) on rat islet cAMP levels. Forskolin (10 μM) was used as positive control. Intracellular cAMP levels were normalized by the control group (n = 8). **p < 0.01, ***p < 0.001; by one-way ANOVA. The results represent by the mean ± SEM.

FIGURE 3

Effects of sphingosine 1-phosphate (S1P) on [Ca²⁺]_i and voltage-dependent Ca²⁺ channel (VDCC) activation in pancreatic β cells. (A) Photomicrograph showing changes in fluorescence of β cells after treatment with different S1P doses. Blue represents β cells stained with Fura-2 AM. Green represents the increase in fluorescence density of Fura-2 AM staining. (B) Changes in F340/F380 after treatment with different doses of S1P in 16.7 mM glucose, which reflects changes in [Ca²⁺]_i. (C) Summary of the mean F340/F380 values after treatment with different doses of S1P (n = 6). *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant; by one-way ANOVA. (D) Representative current traces of VDCC recorded in the presence or absence of S1P (10 μM). (E) Current-voltage relationship curves of VDCCs in β cells. (F) Summary of the mean current densities recorded at 0 mV depolarization (n = 8). Statistical differences between two groups (with or without S1P) were compared using an unpaired two-tailed Student’s t test. The results represent by the mean ± SEM.

FIGURE 4

Sphingosine 1-phosphate (S1P) affected [Ca²⁺]_i by promoting Ca²⁺ influx rather than releasing intracellular Ca²⁺ stores. (A) Changes in F340/F380 values induced by S1P in the presence and absence of EGTA (2.5 mM). EGTA concentrations were sufficient to chelate the calcium present in the KRBH buffer (2 mM). Extracellular Ca²⁺ using CaCl₂ (5 mM) was added to produce a final concentration of 5 mM of free Ca²⁺ at the end of the experiment. (B) Summary of the mean F340/F380 values with S1P (10 μM) treatments in the presence or absence of EGTA (n = 6). *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant; by one-way ANOVA. The results represent by the mean ± SEM.

FIGURE 5

Pancreatic β cells treated by sphingosine 1-phosphate (S1P) exhibited lengthened action potential duration (APD) and lower Kv currents. (A,B) Action potential waveforms of β cells treated without (A) and with (B) S1P (10 μM). (C) Summary of the mean APDs (n = 8). (D) Representative current traces recorded in the absence and presence of S1P (10 μM). (E) Current-voltage relationship curves of β cell Kv channels. (F) Summary of the mean current densities of Kv channels at 80 mV depolarization (n = 8). **p < 0.01, ***p < 0.001; by unpaired two-tailed Student’s t test. The results represent by the mean ± SEM.

FIGURE 6

Sphingosine 1-phosphate (S1P) had no direct effects on Kv2.1 channels but regulated them by modulating phospholipase C and protein kinase C (PLC/PKC) signaling pathway. (A) Representative current traces in Chinese hamster ovary (CHO)-Kv2.1 cells recorded in the absence and presence of S1P (10 μM). (B) Current-voltage relationship curves of Kv2.1 channels from CHO-Kv2.1 cells. (C) Summary of the mean current densities of Kv2.1 channels recorded at 80 mV depolarization (n = 8; ns, not significant; by unpaired two-tailed Student’s t test). (D) Representative current traces recorded after β cells were treated with S1P (10 μM) in the presence of U73122 (U73, 10 μM) or Ro 31-8220 (Ro, 10 μM). (E) Current-voltage relationship curves of Kv channels under different treatment conditions. (F) Summary of the mean current densities of Kv channels recorded at 80 mV depolarization (n = 7). *p < 0.05, **p < 0.01, ***p < 0.001; by one-way ANOVA. The results represent by the mean ± SEM.

FIGURE 7

Sphingosine 1-phosphate (S1P) potentiated insulin secretion through the phospholipase C and protein kinase C (PLC/PKC) pathway. Effect of S1P on insulin secretion under 16.7 mM glucose conditions in the presence of S1P (10 μM), U73122 (U73, 10 μM), and Ro 31-8220 (Ro, 10 μM). U73 and Ro eliminated the insulinotropic effect of S1P (n = 6). *p < 0.05, **p < 0.01, ***p < 0.001; by one-way ANOVA. The results represent by the mean ± SEM.

FIGURE 8

Effects of sphingosine 1-phosphate (S1P) on the cell viability of INS-1 cells and Chinese hamster ovary (CHO)-Kv2.1 cells. INS-1 cells or CHO-Kv2.1 cells were incubated with different treatments for 24 h, and viability was measured via the CCK8 assay. (A) S1P (10 μM) improved the viability of INS-1 cells treated with palmitic acid and high-concentration glucose (PA+HG, n = 6; **p < 0.01, ***p < 0.001; by one-way ANOVA). (B) S1P (10 μM) had no effect on the viability of CHO-Kv2.1 cells treated with palmitic acid and high-concentration glucose (PA+HG, n = 6; ***p < 0.001; by one-way ANOVA). The cell viability of each group was normalized by that of the control group. The results represent by the mean ± SEM.

FIGURE 9

Schematic of our proposed molecular mechanism by which sphingosine 1-phosphate (S1P) induces insulin secretion from pancreatic β cells. S1P inhibits voltage-dependent potassium (Kv) currents by activating the PLC/PKC pathway (rather than the adenylate cyclase (AC)/cAMP pathway), and not by direct suppression, thereby prolonging action potential duration (APD). This promotes Ca²⁺ influx and increases the intracellular Ca²⁺ concentration, finally stimulating insulin secretion.