Sulfonylurea and glinide drugs display different effects on insulin granule motion in single β-cells in vitro. We therefore investigated the different effects that these drugs manifest towards insulin release in an in vivo long-term treatment model. Diabetic GK (Goto-Kakizaki) rats were treated with nateglinide, glibenclamide or insulin for 6 weeks. Insulin granule motion in single β-cells and the expression of SNARE (soluble N-ethylmaleimide-sensitive factor-attachment protein receptor) proteins were then analysed. Perifusion studies showed that decreased first-phase insulin release was partially recovered when GK rats were treated with nateglinide or insulin for 6 weeks, whereas no first-phase release occurred with glibenclamide treatment. In accord with the perifusion results, TIRF (total internal reflection fluorescence) imaging of insulin exocytosis showed restoration of the decreased number of docked insulin granules and the fusion events from them during first-phase release for nateglinide or insulin, but not glibenclamide, treatment; electron microscopy results confirmed the TIRF microscopy data. Relative to vehicle-treated GK β-cells, an increased number of SNARE clusters were evident in nateglinide- or insulin-treated cells; a lesser increase was observed in glibenclamide-treated cells. Immunostaining for insulin showed that nateglinide treatment better preserved pancreatic islet morphology than did glibenclamide treatment. However, direct exposure of GK β-cells to these drugs could not restore the decreased first-phase insulin release nor the reduced numbers of docked insulin granules. We conclude that treatment of GK rats with nateglinide and glibenclamide varies in long-term effects on β-cell functions; nateglinide treatment appears overall to be more beneficial.
Considering the prevalence in the number of Type 2 diabetes cases with a high risk of cardiovascular disease [1,2], prevention of diabetes onset and strict blood glucose control in diabetic patients are expedient. Treatments for Type 2 diabetic patients with characteristically impaired insulin release [3–5] aim to restore insulin release from β-cells, and include pharmaceutical agents targeting KATP (ATP-sensitive K+) channels [6–9]. Sulfonylurea drugs, such as glibenclamide, and glinide drugs, such as mitiglinide, inhibit KATP channel functions and are common treatments [10–14]. Although both drug types elicit insulin secretion via KATP channel closure, the mechanisms of action for these drugs differ. Our previous TIRF (total internal reflection fluorescence) imaging results showed that these drugs have disparate effects on insulin granule motion in single β-cells in vitro . The question remains of how long-term treatment in vivo with these drugs affects insulin exocytosis.
Use of sulfonylureas such as glibenclamide corresponds with a high frequency of hypoglycaemia, and long-term use of glibenclamide yields a progressive decrease in its efficacy. Although this decreased efficacy conceivably results from decreased insulin-producing capacity by the pancreatic β-cells, the mechanism of secondary failure is still unknown . Because glibenclamide binds tightly to SUR (sulfonylurea receptor) sites at the pancreatic β-cell membrane and has long-acting metabolites [16,17], glibenclamide is effective in a once-daily dose . In contrast, glinides such as nateglinide bind quickly and reversibly to SUR sites , resulting in a rapid and short insulin response resembling the physiological pattern of postprandial insulin release [20,21]. Unlike glibenclamide, nateglinide appears not to promote hypoglycaemic attack. Thus it is thought that long-term treatment with these two drug types in vivo in diabetic GK rats may display different effects on β-cell function, particularly insulin exocytosis. To address this issue, in the present study, we administered glibenclamide, nateglinide or control insulin to diabetic GK (Goto-Kakizaki) rats for 6 weeks. We then examined β-cell functions by TIRF imaging of insulin exocytosis, perifusion analysis and expression analysis of SNARE (soluble N-ethylmaleimide-sensitive factor-attachment protein receptor) proteins. Our results demonstrated that long-term nateglinide treatment in GK rats is more effective than glibenclamide in improving the expression of SNAREs, the docking status of insulin granules and fusion events from these granules.
Diabetic GK rats and non-diabetic male Wistar rats were obtained from a commercial breeder (Oriental Yeast, Tokyo, Japan). Rats were housed under a 12 h light/12 h dark cycle and given free access to food and water until the start of experiments, which were conducted with 8-week-old male rats. The body weights of GK rats were not statistically different from controls. For 6 weeks from the experimental start date, GK rats were fed twice daily (9:00 h and 16:00 h) for 1 h, and vehicle (0.5% methylcellulose), nateglinide (50 mg/kg) or glibenclamide (2 mg/kg) was administered by oral gavage just before feeding. For insulin administration, human insulin (2 units of Humulin®; Eli Lilly and Co.) was injected subcutaneously just before feeding. Blood samples were collected from tail veins at the indicated times, and blood glucose levels were measured by Glutest® (Sanwa Kagaku Kenkyusho).
TIRFM (TIRF microscopy)
At the end of the 6-week treatments, pancreatic islets of Langerhans were isolated by collagenase digestion as described by Nagamatsu et al. , with some modifications. Isolated islets were dissociated into single cells by incubation in Ca2+-free KRB (Krebs–Ringer buffer) containing 1 mM EGTA, and cultured on fibronectin-coated (KOKEN Co.) high-refractive-index glass (Olympus) in RPMI 1640 medium (Gibco BRL) supplemented with 10% (v/v) fetal bovine serum (Gibco BRL), 200 units/ml penicillin and 200 μg/ml streptomycin as described in . Cells were maintained at 37 °C in an atmosphere of 5% CO2. To label the insulin secretory granules, pancreatic β-cells were infected with the recombinant adenovirus Adex1CA insulin–GFP (green fluorescent protein) . The Olympus total internal reflection system was used with a high-aperture objective lens [Apo 100× OHR; NA (numerical aperture) 1.65; Olympus]. To observe GFP and Alexa Fluor® 488 (Molecular Probes) labels, we used a 488 nm laser line for excitation and a 515 nm long-pass filter for the barrier. Images were then projected on to a cooled CCD (charge-coupled device) camera (DV887DCSBV; ANDOR) operated with Metamorph version 6.3 (Universal Imaging). Images were acquired at 300 ms intervals. The space constant for the exponential decay of the evanescent field was approx. 43 nm. For real-time images of GFP-tagged insulin granule motion, β-cells were placed on the high-refractive-index glass, mounted in an open chamber, and incubated for 30 min at 37 °C in KRB containing 110 mM NaCl, 4.4 mM KCl, 1.45 mM KH2PO4, 1.2 mM MgSO4, 2.3 mM calcium gluconate, 4.8 mM NaHCO3, 2.2 mM glucose, 10 mM Hepes (pH 7.4) and 0.3% BSA. Cells were then transferred to the thermostat-controlled stage (37 °C) of TIRFM, and stimulation with glucose was achieved by the addition of 52 mM glucose/KRB into the chamber (22 mM final concentration of glucose). Most analyses, including tracking (single projection of differing images) and area calculations, were performed using Metamorph software. To analyse the data, fusion events were selected manually, and the average fluorescence intensity of individual granules in a 1 μm×1 μm square placed over the granule centre was calculated. The number of fusion events was counted manually while looping approx. 5000 frame time-lapses.
At the end of the 6-week treatments, pancreatic islets were isolated and β-cells were housed in a small chamber (approx. 5×105 cells/chamber) and perifused with KRB (2.2 mM glucose) for 60 min at a flow rate of 0.5 ml/min at 37 °C before collecting fractions. Insulin release was stimulated by 16.7 mM glucose. Fractions were collected at 1 min intervals. Insulin release in 5 μl aliquots of medium was measured by an insulin ELISA kit (Morinaga).
EM (electron microscopy)
EM was carried out by conventional methods as described previously . Tissues were fixed in phosphate-buffered 2.5% glutaraldehyde (pH 7.4), post-osmicated, dehydrated with graded alcohols and embedded in Epon 812. After staining with uranyl acetate and lead citrate, ultrathin sections were examined using a transmission electron microscope TEM-1010C (JEOL). In EM, granules at their shortest distance of <10 nm from the plasma membrane were qualified as morphologically docked granules .
For TIRF analysis of syntaxin 1A (also known as HPC-1) and SNAP-25 (synaptosome-associated protein of 25 kDa) clusters at the plasma membrane, pancreatic β-cells were fixed, permeabilized with 2% (w/v) paraformaldehyde and 0.1% Triton X-100, and processed for immunocytochemistry as described previously . Cells were labelled with monoclonal anti-insulin antibodies (Sigma–Aldrich), anti-HPC-1 antibodies (Sigma–Aldrich) and anti-SNAP-25 antibodies (Wako Co.), then processed with Alexa Fluor 488-conjugated anti-mouse IgG (Molecular Probes). Immunofluorescence staining was detected by TIRFM. For morphometric analysis of islets, paraffin-embedded pancreas sections (10 μm) were labelled with anti-insulin antibody and detected using an avidin–biotin–peroxidase technique (Vector Laboratories). Sections were collected at 500 μm intervals from tissue blocks, and all islets in the sections were analysed as islet area over total pancreatic area. Images were acquired using an Olympus microscope IX70, equipped with a CCD camera, and analysed using Metamorph software.
Results are means±S.E.M. for at least three different experiments performed independently. Statistical analysis was performed by ANOVA followed by Dunnett's test or Tukey–Kramer's test using the Statview software for Windows version 5.0 (SAS Institute, Cary, NC, U.S.A.).
Effects of nateglinide and glibenclamide in vivo in GK rats
We first examined how short-term administration of nateglinide and glibenclamide before feeding affects GK rat blood glucose levels. The concentration of each drug to be used was chosen based on the total area under the curve (AUC) generated by administering drug to fasted rats, feeding them for 1 h beginning at 09:00 h, then tracking glucose levels from 09:00 to 15:00 h (glucose AUC09:00-15:00). Because there was no significant difference in the total glucose AUC09:00-15:00 for 50 mg/kg nateglinide and 2 mg/kg glibenclamide (Figure 1A), we used these concentrations of drugs in the present study. Therefore the dosage of nateglinide used was nearly 25-fold higher than that of glibenclamide. In fact, in an ex vivo study, the maximal insulinotropic effects of nateglinide and glibenclamide on isolated rat pancreatic islets were observed at concentration of approx. 10 μM and 300 nM respectively , suggesting that the effective concentration of nateglinide was approx. 30-fold higher than that of glibenclamide. As shown in Figure 1(A), nateglinide administration reduced blood glucose 1 h after feeding. In contrast, glibenclamide administration reduced the blood glucose levels 2 and 3 h after feeding, but not 1 h after feeding. To examine long-term effects, GK rats were fed twice daily, and either nateglinide or glibenclamide was administered just before feeding for 6 weeks. We studied how long-term treatment of GK rats with nateglinide, glibenclamide or insulin injection affected blood glucose levels. During the period of treatments, there was no difference in fasting blood glucose levels among vehicle-treated, nateglinide-treated, glibenclamide-treated and insulin-treated GK rats (Figure 1B). As shown in Figure 1(C), nateglinide treatment decreased 1 h postprandial levels of blood glucose [182.6±5.8 mg/dl in vehicle treatment compared with 149.9±3.6 mg/dl in nateglinide treatment (n=10 each), P<0.01 compared with vehicle treatment, at 3 weeks after treatment; 198.2±6.1 mg/dl in vehicle treatment compared with 160.5±4.0 mg/dl in nateglinide treatment (n=10 each), P<0.01 compared with vehicle treatment, at 6 weeks after treatment]. Insulin treatment resulted in a similar decrease in blood glucose [182.6±5.8 mg/dl in vehicle treatment compared with 138.6±6.9 mg/dl in insulin treatment (n=10 each), P<0.01 compared with vehicle treatment, at 3 weeks after treatment; 198.2±6.1 mg/dl in vehicle treatment compared with 141.2±2.5 mg/dl in insulin treatment (n=10 each), P<0.01 compared with vehicle treatment, at 6 weeks after treatment]. However, glibenclamide treatment did not decrease 1 h postprandial blood glucose levels (Figure 1C).
Comparison of blood glucose levels in GK rats following one-shot administration (A) and during long-term treatment (B, C) of vehicle, nateglinide or glibenclamide
Nateglinide and insulin treatments partially recover first-phase glucose-induced insulin release
After the 6-week treatment with drug or insulin, we isolated pancreatic β-cells from treated GK rats. Cells were perifused, and insulin release was subsequently measured by ELISA. ELISA data showed partial recovery of first-phase release in insulin-treated β-cells relative to control (Wistar rat) cells. Second-phase release was similar in insulin-treated and control levels (Figure 2A). Because the levels of insulin release are low in nateglinide- and glibenclamide-treated β-cells, the scale in the graph was enlarged as shown in Figure 2(B). Nateglinide-treated GK β-cells showed a biphasic glucose-induced insulin release pattern. In contrast, there was no obvious first phase in glibenclamide-treated β-cells.
Insulin release in response to 16.7 mM glucose in perifused β-cells from control Wistar rat and 6-week-treated GK rat β-cells
TIRF imaging analysis of insulin exocytosis
We explored the effects of nateglinide, glibenclamide or insulin treatment on the dynamic motion of single insulin granules by TIRF imaging. As shown in our previous study [23,28], in rat β-cells fusion of insulin granules during first-phase release mainly involved previously docked granules, whereas fusion from newcomer granules was observed during second-phase release (Supplementary Movie S1 at http://www.BiochemJ.org/bj/412/bj4120093add.htm and Figure 3, Wistar control). There was no fusion from previously docked granules in vehicle-treated GK β-cells (Supplementary Movie S2 at http://www.BiochemJ.org/bj/412/bj4120093add.htm). However, nateglinide-treated (Supplementary Movie S3 at http://www.BiochemJ.org/bj/412/bj4120093add.htm) and insulin-treated (Supplementary Movie S4 at http://www.BiochemJ.org/bj/412/bj4120093add.htm) GK β-cells showed fusion events from previously docked granules under 22 mM glucose stimulation (Figure 3). There was little fusion observed from previously docked granules in glibenclamide-treated GK β-cells (Supplementary Movie S5 at http://www.BiochemJ.org/bj/412/bj4120093add.htm and Figure 3). We then counted the number of fusion events manually, and performed statistical analysis. The total number of fusion events during first-phase release (5 min after stimulation, per 200 μm2) was increased in both insulin-treated and nateglinide-treated GK β-cells [4.8±0.9 in vehicle-treated cells (n=5); 19.5±2.2 in insulin-treated cells (n=6), P<0.01 compared with GK-vehicle; 13.7±1.3 in nateglinide-treated cells (n=8), P<0.01 compared with GK-vehicle], but the number of fusion events in glibenclamide-treated GK β-cells was not increased (9.6±2.7 in glibenclamide treated cells, no significant difference from GK-vehicle).
TIRF imaging of insulin exocytosis in single β-cells
We then examined the docking status of insulin granules using TIRFM with immunostaining for insulin (Figure 4). In agreement with our previous findings , we rarely observed morphologically docked granules in diabetic GK β-cells. The number of docked insulin granules in vehicle-treated GK rat β-cells markedly decreased to 31.2±1.9% of control numbers (Figure 4B). In contrast, nateglinide (66.0±3.1% of control, P<0.01 compared with GK-vehicle) and insulin (80.4±3.5% of control, P<0.01 compared with GK-vehicle) caused a considerable increase in the number of docked insulin granules, whereas glibenclamide appeared mildly effective relative to nateglinide (45.5±5.0% of control; P<0.05 compared with GK-vehicle; P<0.01, GK-nateglinide compared with GK-glibenclamide) (Figure 4B). In order to confirm the TIRFM data, we used EM to examine the status of docked insulin granules. As shown in Figure 5(A), the number of docked insulin granules was reduced in vehicle-treated GK β-cells, but nateglinide and insulin treatments increased the number of docked granules. We found that the number of insulin granules within 10 μm of the plasma membrane (10.0±1.2 and 1.8±0.7 granules per 10 μm of plasma membrane in Wistar control and GK-vehicle treatment samples respectively) was increased by nateglinide and insulin treatments (GK-nateglinide 5.0±0.8, P<0.01 compared with GK-vehicle; GK-insulin 4.5±0.4, P<0.01 compared with GK-vehicle), whereas the effect of glibenclamide was minor (GK-glibenclamide 2.6±0.2, no significant difference compared with GK-vehicle; P<0.05, GK-nateglinide compared with GK-glibenclamide) (Figure 5B).
TIRF imaging of insulin granules morphologically docked at the plasma membrane
Status of docked insulin granules observed by EM
Decreased number of t-SNARE (target membrane SNARE) clusters partially restored by nateglinide and insulin treatments
Since the expression and number of syntaxin 1A clusters on the plasma membrane regulate first-phase insulin release , we examined the effect of 6-week treatments with these drugs on SNARE clusters. To quantify the number of t-SNARE clusters on the plasma membrane, pancreatic β-cells were immunostained and observed by TIRFM. The number of syntaxin 1A clusters (26.5±1.0% of control) and SNAP-25 clusters (22.5±1.0% of control) markedly decreased in vehicle-treated GK rat islets (Figure 6). Both nateglinide and insulin treatments caused recovery of the decreased number of syntaxin 1A and SNAP-25 clusters (for syntaxin 1A, GK-nateglinide showed 61.4±2.6% of control, P<0.01 compared with GK-vehicle; GK-insulin 72.9±2.9% of control, P<0.01 compared with GK-vehicle; for SNAP-25, GK-nateglinide showed 47.4±1.6% of control, P<0.01 compared with GK-vehicle; GK-insulin 51.2±1.8% of control, P<0.01 compared with GK-vehicle). However, glibenclamide treatment only slightly increased the number of clusters relative to that in the GK-vehicle and GK-nateglinide samples (syntaxin 1A, 34.4±4.3% of control, no significant difference compared with GK-vehicle; SNAP-25, 28.9±1.0% of control, P<0.05 compared with GK-vehicle), indicating that glibenclamide was less effective than nateglinide.
TIRF images and the quantification of syntaxin 1A clusters and SNAP-25 clusters on the plasma membrane of control or treated GK β-cells
Morphology of the pancreatic islets
We examined the morphology of the pancreatic islets in GK rats by immunostaining for insulin at the end of the 6-week treatment with drug or insulin. As shown in Figure 7, GK rat islets displayed an irregular structure, with cells separated from each other. Nateglinide and insulin treatments restored such irregularity to the structure similarly to Wistar control samples. Whereas insulin immunoreactivity was much weaker in vehicle-treated GK rats, nateglinide and insulin treatments caused strong insulin immunoreactivities that were not observed in glibenclamide-treated islets.
Morphological comparison of islets from control or treated rat pancreatic tissue by immunostaining for insulin
Direct exposure of GK β-cells to drug failed to recover the decreased first-phase insulin release or the reduced numbers of docked insulin granules
Finally, we examined the direct effect of nateglinide and glibenclamide on pancreatic β-cells isolated from GK rats to determine whether acute exposure to these drugs is sufficient to restore the decreased first-phase insulin release or the reduced numbers of docked insulin granules. GK rat β-cells were pre-incubated for 1 h with 10 μM nateglinide or 300 nM glibenclamide, concentrations which, as described above, should have maximal effects on isolated rat pancreatic β-cells. After these drugs were removed, treated cells were then stimulated with 22 mM glucose. As shown in Figure 8(A) using TIRFM, neither nateglinide nor glibenclamide could restore the decreased first-phase insulin release. We also conducted TIRFM with immunostaining for insulin to observe the docking status of insulin granules in these treated cells (Figure 8B). The number of docked granules in vehicle-pre-incubated GK rat β-cells was reduced to 27.9±3.2% of control numbers, in agreement with results shown in Figure 4(B). Similarly, the pre-incubation with nateglinide (29.8±3.4% of control) and glibenclamide (25.1±3.3% of control) had no effect on the number of docked granules. These results indicate that direct exposure of GK β-cells to these drugs cannot restore the decreased first-phase insulin release nor the reduced numbers of docked insulin granules. Thus the beneficial effects of nateglinide on restoring the decreased glucose-induced first-phase insulin release and docked insulin granules might be due to an indirect effect as a consequence of its protection against pancreatic islet damage.
Effects of acute nateglinide and glibenclamide treatment on the decreased first-phase insulin release (A) and the reduced numbers of docked insulin granules (B) in GK β-cells
We have reported previously that sites of action for glinides in pancreatic β-cells are different from those for sulfonylureas, on the basis of evidence that only mitiglinide, but not glibenclamide, causes the fusion of insulin granules in SURl-KO mice β-cells . Thus different mechanisms underlying glinide and sulfonylurea drug activities are expected to variably affect diabetic β-cell functions when these drugs are administered long-term in vivo. We found that nateglinide, but not glibenclamide, treatment in GK rat β-cells caused a restoration of both decreased t-SNARE clusters and the number of docked insulin granules, increasing fusion events from docked insulin granules under glucose stimulation. However, direct exposure of GK β-cells to these drugs could not restore the decreased first-phase insulin release nor the reduced numbers of docked insulin granules. Notably, although both nateglinide and glibenclamide could equally reduce the blood glucose levels during a 6 h period after feeding, only nateglinide treatment could restore the impaired insulin exocytosis. Thus our data indicate that the beneficial effect of nateglinide treatment may be due not only to release of glucotoxicity, but also to an indirect effect as a consequence of its protection against pancreatic islet damage. The finding that nateglinide, but not glibenclamide, could rapidly reduce blood glucose levels after feeding indicates distinct effects on β-cell functions. Insulin treatment also suppressed the blood glucose levels promptly after feeding, permitting effective recovery of first phase of insulin release and fusion from previously docked granules. It has been proposed that the restoration of both the early phase of insulin release and postprandial hyperglycaemia have potentially significant implications in improving metabolic control and reducing macrovascular complications . Therefore blood glucose levels just after feeding play an important role not only in the recovery of impaired insulin exocytosis, but also in rationalizing the therapeutic approach. Of course, it remains possible that alteration of fatty acid or other metabolite levels elicited by glinide treatment is affecting insulin exocytosis.
Postprandial hyperglycaemia is determined, in large part, by peripheral glucose utilization and the severity of insulin resistance ; however, hepatic glucose output and glucose utilization are more sensitive to suppression by insulin . Insulin released during the early first phase has a more likely role in acute hepatic glucose output . Therefore the partial recovery of first-phase insulin release by nateglinide and insulin treatments may effectively suppress glucose output from the liver, rapidly decreasing postprandial hyperglycaemia. At present, the reason such an acute suppression of blood glucose levels just after feeding eventually recovers the docking status of insulin granules, as well as SNARE protein expression, is unknown. It is possible that IL-1β (interleukin 1β) production contributes to this mechanism. Hyperglycaemia may be associated with IL-1β production in β-cells [33,34], although it is still controversial . We demonstrated previously that the exposure of β-cells to IL-1β decreased SNARE expression and fusion of insulin granules . If nateglinide treatment could alter IL-1β production in β-cells, such an alteration could regulate SNARE expression and the docking status of insulin granules. Although there is no report regarding nateglinide and IL-1β to date, it appears worthwhile to investigate whether nateglinide treatment may inhibit IL-1β production.
We also reported previously that the effect of glinide on insulin exocytosis is very rapid and quickly decays; in contrast, glibenclamide acts on insulin exocytosis continuously . Mechanistic differences between these drugs may be due to their different binding affinity towards SURs. The short action of glinide may benefit β-cell survival; continuous stimulation of β-cells by glibenclamide may contribute to β-cell exhaustion. In agreement with our data, Laghmich et al.  reported that when both normal and GK rats were exposed to nateglinide and glibenclamide for 7 days, insulin in islets, the secretory response to high-glucose and basal biosynthetic activity were better in those animals which received nateglinide. Taken together, these data suggest that short-acting nateglinide preserves β-cell functions better than the long-acting glibenclamide does.
An alternative speculation is that autocrine signals induced by first-phase insulin release may regulate β-cell functions. Large amounts of insulin released in a time course resembling the first phase of glucose-stimulated insulin release may be required for binding to insulin receptors on β-cells. Autocrine signalling by insulin in islet β-cells is now well known to be important for islet mass, affecting β-cell apoptosis and proliferation [38–40]. It has typically been observed that nateglinide treatment can prevent destruction of GK rat islets [41,42]. We noted that the insulin immunostaining pattern in nateglinide-treated islets was almost similar to those of control islets; we concur that nateglinide probably inhibits apoptosis or promotes proliferation directly. It has been reported that glibenclamide, but not nateglinide, induced β-cell apoptosis in human islets , corroborating the suggestion that short-acting insulin secretagogues may be preferred to long-acting ones. In addition, it is of note that insulin treatment has the most beneficial effects on restoring biphasic insulin exocytosis. Although the precise reason is unknown, it is conceivable that insulin signalling itself makes islets survive and maintains the physiological functions of β-cells. In conclusion, long-term nateglinide treatment in GK rats is more beneficial than glibenclamide in improving insulin exocytosis.
We thank Dr Y. Kitahara (Pharmacological Research Laboratories, Ajinomoto, Japan) for generous support. This work was supported by Scientific Research on Priority Areas 18050033 and Scientific Research (C) 17590277 (to M. O.-I.) from the Japanese Ministry of Education, Culture, Sports, Science and Technology, and by a grant from the Japan Private School Promotion Foundation (to S. N.).
area under the curve
green fluorescent protein
synaptosome-associated protein of 25 kDa
soluble N-ethylmaleimide-sensitive factor-attachment protein receptor
total internal reflection fluorescence
target membrane SNARE