Blockade of sphingosine 1-phosphate receptor 2 signaling attenuates streptozotocin-induced apoptosis of pancreatic β-cells

Abstract

Sphingosine 1-phosphate (S1P) is a potent sphingolipid mediator that acts through five cognate G protein-coupled receptors (S1P₁–S1P₅) and regulates many critical biological processes. Recent studies indicated that S1P at nanomolar concentrations significantly reduces cytokine-induced apoptosis of pancreatic β-cells in which genes for S1P₁–S1P₄ are co-expressed. However, the S1P receptor subtype(s) involved in this effect remains to be clarified. In this study, we investigated the potential role of S1P₂ in streptozotocin (STZ)-induced apoptosis of pancreatic β-cells and progression of diabetes. S1P₂-deficient (S1P₂^−/−) mice displayed a greater survive ability, lower blood glucose levels, and smaller numbers of TUNEL-positive apoptotic β-cells to administration of a high dose of STZ than wild-type (WT) mice. S1P₂^−/− mice showed higher insulin/glucose ratios (an index of relative insulin deficiency) and larger insulin-positive islet areas to administration of a low dose of STZ than WT mice. Moreover, administration of JTE-013, a S1P₂-specific antagonist, to WT mice ameliorated STZ-induced blood glucose elevation and reduced the incidence of diabetes. Our findings indicate that blockade of S1P₂ signaling attenuates STZ-induced apoptosis of pancreatic β-cells and decreases the incidence of diabetes.

Introduction

Type 1 diabetes is an autoimmune disease that results in the destruction of pancreatic β-cells, whereas type 2 diabetes is a much more common disorder caused by insulin resistance and relative insulin deficiency. Insulin resistance in type 2 diabetes is initially managed by enhanced secretion of insulin from β-cells; however, this may gradually lead to a decrease in β-cell mass and deterioration of key β-cell functions such as glucose-stimulated insulin secretion. Therefore, β-cell failure is a key pathogenic process in both type 1 and 2 diabetes [1,2].

Sphingosine 1-phosphate (S1P) is a bioactive lipid mediator that exhibits diverse biological functions in most cell types and regulates many pathological processes, acting through five cognate high-affinity receptors (S1P₁–S1P₅) [3,4]. S1P was found to ameliorate cytokine-induced apoptosis of β-cells [5,6] in which genes for S1P₁–S1P₄ are co-expressed [7]. Previous studies demonstrated that S1P₂ signaling induces intracellular calcium mobilization leading to increased glucose uptake in myoblasts [8], and RNA-based screening for genes that confer insulin resistance to 3T3-L1 adipocytes identified S1pr2 as a candidate [9], suggesting that S1P₂ plays an important role in the pathogenesis of diabetes. In the present study, we examined the role of S1P₂ in streptozotocin (STZ)-induced apoptosis of β-cells and progression of diabetes using S1P₂-deficient (S1P₂^−/−) mice as well as the S1P₂-specific antagonist JTE-013.

Materials and methods

Animals

S1P₂^−/− mice were generated and genotyped as described previously [10]. S1P₂^−/− mice were backcrossed with C57BL/6N (Clea Japan, Tokyo, Japan) for seven generations, and thus, littermate wild-type (WT) mice or age-matched (8-week-old) C57BL/6N were used as controls. All mice were fed ad libitum with standard chow/water and kept under a 12-hour light-dark cycle in an air-conditioned room. All animal protocols were approved by the animal care and use committee of Chiba-East National Hospital.

Induction of diabetes by STZ injection

Streptozotocin (STZ, Sigma) was freshly dissolved in 20 mM citrate buffer (pH4.5) and intraperitoneally administered under various conditions in each experiment: 50 mg/kg body weight for 5 consecutive days (50 mg/kg for 5 days), 100 mg/kg for 1 day, or 100 mg/kg for 2 days. Control mice received injections of the citrate buffer. JTE-013 (Calbiochem), a specific S1P₂ antagonist [11], was freshly dissolved in saline and intraperitoneally administered at 4 mg/kg for 6 days (one shot prior to STZ and five shots with STZ). Control mice received injections of saline. Blood was collected from the retro-orbital sinus of anesthetized mice and blood glucose levels were measured using the Accu-Chek Aviva system (Roche). Mice were diagnosed with diabetes mellitus (DM) when their blood glucose levels were ≥ 300 mg/dl on two consecutive days [12]. Serum insulin levels were measured using an insulin RIA kit (Millipore) in accordance with the manufacturer’s instructions.

Immunohistochemistry

Pancreata were quickly removed from anesthetized mice, fixed with 3% formalin in phosphate-buffered saline, and embedded in paraffin. To count islet cells, deparaffinized pancreatic sections were immunostained with guinea pig polyclonal anti-insulin antibody (Cell Marque, Rocklin, CA) using a NexES IHC system (Ventana Medical Systems, Tucson, AZ). Full area sizes (mm²) of pancreatic sections (single section per mouse) were measured and the numbers of insulin-positive islets in each section were counted. Apoptotic cells were detected using a terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay (Apotag Kit; Chemicon) in accordance with the manufacturer’s recommendations. Apoptotic cells per nm² of islet area were counted in ≥ 10 islets per section.

Statistical analysis

Results are expressed as mean ± SD. All statistical analyses were performed using Dr. SPSS II for Windows (SPSS Inc., Chicago, IL). The existence of significant differences between two groups (with an accuracy of at least 95%) was analyzed using a two-tailed unpaired t-test. Kaplan-Meier analysis was used to examine diabetes-free rates and survival rates, and the differences were determined by log-rank tests. A value of P < 0.05 was considered significant.

Results

S1P₂^−/− mice were more resistant to administration of a high dose of STZ

WT and S1P₂^−/− males were intraperitoneally injected with a high dose of STZ (100 mg/kg for 2 days), and their health status was monitored every week until the 15^th week after the final injection. Forty percent (10/25) of WT mice died at the 2^nd week, increasing to 64.0% (16/25) by the 11^th week. In contrast, S1P₂^−/− mice show lower death rates of 11.1% (2/18) and 27.7% (5/18), respectively. Kaplan–Meier analysis indicates that S1P₂^−/− mice were significantly (P = 0.0334) more resistant to STZ toxicity than WT mice (Fig. 1A). At the 15^th week after the final injection, serum glucose levels in surviving S1P₂^−/− mice were significantly lower than those in surviving WT mice (Fig. 1B).

Fig. 1.

S1P₂^−/− mice were more resistant to administration of a high dose of STZ. (A) Kaplan-Meier survival analysis of WT and S1P₂^−/− mice (n = 25 and 18, respectively) after the final injection of a high dose of STZ (100 mg/kg for 2 days). (B) Blood glucose levels of randomly fed, surviving WT and S1P₂^−/− mice at the 15^th week after the final STZ injection (n = 9 and 13, respectively). The differences were significant (*P < 0.05).

More β-cells were preserved after STZ injection in S1P₂^−/− mice

WT and S1P₂^−/− males were injected with a low dose of STZ (50 mg/kg for 5 days) so that all mice survived until at least the 30^th day after the final injection, and blood glucose levels were measured twice a week. There was no significant difference in glucose levels between WT and S1P₂^−/− mice (Fig. 2A). Eight of eleven (72.7%) WT mice were diagnosed with diabetes compared with only three of eight (37.5%) S1P₂^−/− mice; however, Kaplan-Meier analysis revealed no significant difference in diabetes-free rates between the two groups (Fig. 2B). Although blood insulin levels in S1P₂^−/− mice were comparable to those in WT mice at the 30^th day after the final STZ injection (Fig. 2C), the insulin/glucose ratios (an index of relative insulin deficiency [13]) were significantly higher in S1P₂^−/− mice than in WT mice (Fig. 2D). This indicates that insulin was more efficiently secreted in response to blood glucose elevation in S1P₂^−/− mice than in WT mice. We counted the numbers of insulin-positive islets per pancreatic area at the 30^th day after STZ injection. Insulin-positive islets were much more abundant in pancreatic sections from S1P₂^−/− mice (Fig. 2E). The numbers of insulin-positive islets per pancreatic area in S1P₂^−/− mice were comparable to those in WT mice without STZ, because more islet cells (per pancreatic area) withstood STZ-induced cytotoxicity in S1P₂^−/− mice (Fig. 2F). Taken together, these findings indicate that more islet β-cells were preserved after STZ injection in S1P₂^−/− mice than in WT mice.

Fig. 2.

More β-cells were preserved after STZ injection in S1P₂^−/− mice. WT and S1P₂^−/− mice (n = 11 and 8, respectively) were administered a low dose of STZ (50 mg/kg for 5 consecutive days intraperitoneally). (A) Blood glucose levels in randomly fed WT and S1P₂^−/− mice after STZ injection. (B) Kaplan-Meier analysis of diabetes-free rates after STZ injection. (C) Serum insulin levels (ng/ml) at the 30^th day after the final STZ injection. (D) Blood insulin/glucose ratios (× 10⁻³) as an index of relative insulin deficiency at the 30^th day after the last STZ injection. (E) Representative images of insulin-positive islets in pancreas at the 30^th day after the final STZ injection. Bars: 200 μm. (F) Numbers of insulin-positive islets per pancreatic area (mm²) without (n = 5 each for WT and S1P₂^−/− mice) or with STZ injection (at the 30^th day after the final STZ injection). *P < 0.05 and **P < 0.01.

More islet cells were protected against STZ-induced apoptosis in S1P₂^−/− mice

The administration of STZ is known to induce apoptosis of islet β-cells [13,14]. WT and S1P₂^−/− males were intraperitoneally injected with a high dose of STZ (100 mg/kg for 1 day), and apoptosis of islet cells was evaluated the following day. Histological analysis revealed that TUNEL-positive apoptotic islet cells were less abundant in S1P₂^−/− mice than in WT mice (Fig. 3A). The number of apoptotic cells per nm² of islet area in S1P₂^−/− mice was 27% of that in WT mice (Fig. 3B), indicating that the lack of S1P₂ protects islet cells from STZ-induced apoptosis.

Fig. 3.

More islet cells were protected against STZ-induced apoptosis in S1P₂^−/− mice. Mice were injected with a high dose of STZ (100 mg/kg for 1 day) and apoptotic β-cells were identified by immunostaining (A) TUNEL-positive apoptotic β-cells in pancreatic sections obtained from WT and S1P₂^−/− mice. Bars: 50 μm. (B) Numbers of apoptotic cells per islet area (nm²) of WT and S1P₂^−/− mice (n = 10 each). The difference is significant (*P < 0.0001).

JTE-013 decreased the incidence of diabetes

We examined the effect of JTE-013, a specific S1P₂ antagonist, on the incidence of STZ-induced diabetes. JTE-013 (4 mg/kg) was intraperitoneally injected for 6 days (one shot prior to STZ and five shots with STZ (50 mg/kg)), and blood glucose elevation after the final STZ injection was examined every week. Glucose levels at the 4^th week after the final STZ injection were significantly lower in JTE-013-treated mice than JTE-013-untreated mice (Fig. 4A). Glucose levels were not indistinguishable between JTE-013-treated and JTE-013-untreated mice in the absence of STZ. Among STZ-injected mice, 71.4% (10/14) of JTE-013-untreated mice were found to be diabetic compared with only 14.3% (2/14) of JTE-013-treated mice by the 6^th week after the final STZ injection. Kaplan-Meier analysis indicates that JTE-013 treatment significantly (P = 0.0024) decreased the incidence of diabetes (Fig. 4B).

Fig. 4.

JTE-013 decreased the incidence of diabetes. JTE-013 (4 mg/kg) or saline was intraperitoneally injected for 6 days (one shot prior to STZ and five shots with a high dose of STZ (50 mg/kg)). (A) Changes in blood glucose levels of randomly fed mice with/without STZ or JTE-013 injection. Sample numbers are 14 for +STZ/−JTE-013 (closed circles), +STZ/+JTE-013 (closed squares), −STZ/−JTE-013 (open circles), and −STZ/+JTE-013 (open squares). The effects of JTE-013 were significant at the 4^th weeks after the final STZ injection (*P < 0.01). (B) Kaplan-Meier analysis of diabetes-free rates after STZ injection. Co-injection with JTE-013 significantly reduced the diabetes-free rates (*P < 0.01).

Discussion

We found that the blockade of S1P₂ signaling prevents the onset of diabetes by protecting islet β-cells from STZ-induced injury (Fig. 2 and 3). S1P₂ is known to couple with Gq, activate phospholipase C that leads to Ca²⁺ mobilization, and induce the activation of extracellular-regulated kinase, stress-activated protein kinase/c-jun N-terminal kinase, and mitogen-activated protein kinase (MAPK) p38 [15,16]. The activation of MAPK p38 is a critical event leading to β-cell apoptosis and promoting peripheral insulin resistance [17,18,19], and thus mice lacking MAPK p38 were found to be protected against pancreatic β-cell failure and insulin resistance [14]. Therefore, blockade of S1P₂ signaling may lead to deactivation of MAPK p38 and attenuation of STZ-induced β-cell failure/progression of diabetes. The present study bears analogy to our previous one, in which accelerated regeneration of hepatocytes was observed after liver injury in S1P₂^−/− mice [20], and may suggest generalization of cell protective effect by S1P₂ inactivation. Recently, it has been suggested that S1P₂ may protect against diabetes by preventing insulin resistance; S1P₂^−/− mice showed insulin resistance [9] and S1P₂ signaling increased glucose uptake in myoblasts [8]. Although further studies are necessary to solve this inconsistency with our findings, the protective effect on pancreatic β-cells resulting from S1P₂ inactivation may have surpassed the undesired insulin resistance in our system.

We further explored the possibility that the S1P₂-specific antagonists have potential for use in the treatment of diabetes. JTE-013 decreased blood glucose levels (Fig. 4A) and reduced the incidence of diabetes in STZ-injected mice (Fig. 4B). It has been shown that FTY720, a pro-drug against all the S1P receptors except S1P₂ (i.e., S1P₁ and S1P₃–₅), reduces the incidence of diabetes in mice [21]. These results suggest that the protective effects of S1P against diabetes (including blockade of immune cell migration) may be mediated by non-S1P₂-type S1P receptors. In this context, selective inactivation of pro-diabetic S1P₂ and preservation of anti-diabetic non-S1P₂-type S1P receptors by S1P₂-specific antagonists may have strong potential as a future anti-diabetic strategy; the ligand S1P is abundant in the blood stream [22].

Conclusion

We report here that selective blockade of S1P₂ signaling attenuates STZ-induced apoptosis of pancreatic β-cells and decreases the incidence of diabetes in mice. Modulation of S1P signaling may provide a new avenue for the treatment of diabetes.

Abbreviations

MAPK

mitogen-activated protein kinase

S1P

sphingosine 1-phosphate

S1P₂

sphingosine 1-phosphate receptor 2

STZ

Streptozotocin

TUNEL

terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling

WT

wild-type