Insulin secretory deficiency and glucose intolerance in Rab3A null mice.

Insulin secretory dysfunction of the pancreatic beta-cell in type-2 diabetes is thought to be due to defective nutrient sensing and/or deficiencies in the mechanism of insulin exocytosis. Previous studies have indicated that the GTP-binding protein, Rab3A, plays a mechanistic role in insulin exocytosis. Here, we report that Rab3A(-/-) mice develop fasting hyperglycemia and upon a glucose challenge show significant glucose intolerance coupled to ablated first-phase insulin release and consequential insufficient insulin secretion in vivo, without insulin resistance. The in vivo insulin secretory response to arginine was similar in Rab3A(-/-) mice as Rab3A(+/+) control animals, indicating a phenotype reminiscent of insulin secretory dysfunction found in type-2 diabetes. However, when a second arginine dose was given 10 min after, there was a negligible insulin secretory response in Rab3A(-/-) mice, compared with that in Rab3A(+/+) animals, that was markedly increased above that to the first arginine stimulus. There was no difference in beta-cell mass or insulin production between Rab3A(-/-) and Rab3A(+/+) mice. However, in isolated islets, secretagogue-induced insulin release (by glucose, GLP-1, glyburide, or fatty acid) was approximately 60-70% lower in Rab3A(-/-) islets compared with Rab3A(+/+) controls. Nonetheless, there was a similar rate of glucose oxidation and glucose-induced rise in cytosolic [Ca(2+)](i) flux between Rab3A(-/-) and Rab3A(+/+) islet beta-cells, indicating the mechanistic role of Rab3A lies downstream of generating secondary signals that trigger insulin release, at the level of secretory granule transport and/or exocytosis. Thus, Rab3A plays an important in vivo role facilitating the efficiency of insulin exocytosis, most likely at the level of replenishing the ready releasable pool of beta-granules. Also, this study indicates, for the first time, that the in vivo insulin secretory dysfunction found in type-2 diabetes can lie solely at the level of defective insulin exocytosis.

Insulin is the major anabolic hormone controlling metabolic homeostasis, and without an effective supply of insulin diabetes mellitus ensues. Type-1 diabetes occurs as a result of autoimmune destruction of pancreatic ␤-cells that produce insulin, and type-2 diabetes develops as a result of insulin secretory dysfunction, as well as insufficient ␤-cell mass, that no longer compensates for peripheral insulin resistance (1,2). The insulin secretory dysfunction in type-2 diabetes is derived from ␤-cell secretory abnormalities, proposed to be either at the level irregular glucose metabolism required for generating secondary signals necessary to trigger insulin exocytosis and/or deficiencies in the exocytotic mechanism itself (1,3). Insulin secretion from the ␤-cell is highly regulated and only occurs in response to certain nutrients, hormones, neurotransmitters, and pharmacological reagents (4). Of these, glucose is the most physiologically relevant. The secondary signals that emanate from increased glucose metabolism to stimulate insulin release in ␤-cells have been relatively well defined, and of these a rise in cytosolic [Ca 2ϩ ] i is a prerequisite (4,5). Indeed, increased [Ca 2ϩ ] i is the necessary signal to trigger regulated exocytosis in most neuroendocrine cells (6). In comparison, the mechanism of insulin exocytosis is less well defined. Several proteins required for insulin exocytosis in ␤-cells have been indicated, including SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) proteins, analogous to the mechanism of regulated exocytosis in neurons (5). However, how such "exocytotic proteins" interact in a regulated manner to control insulin exocytosis is currently unclear (5).
One protein implicated to play a role in control of insulin exocytosis is the small GTP-binding protein, Rab3A (7). In ␤-cells, Rab3A is located on the cytosolic face of ␤-granules where it is probably involved in control of ␤-granule transport and/or exocytosis (8,9). Members of the Rab protein family are key to directing vesicular transport in eukaryotic cells (10). In Rab3A Ϫ/Ϫ mice, there is a defect in recruiting synaptic vesicles for exocytosis in hippocampal neurons (11), and it has been indicated that Rab3A plays a role in the later stages of synaptic vesicle exocytosis controlling the efficiency of neurotransmitter release (12). However, although there are similarities, the mechanism of synaptic vesicle exocytosis is distinct from that for large dense-core granules (13). Nonetheless, given the proposed key role that Rab3A plays in controlling insulin exocytosis, we examined whether there was a deficient insulin secretory phenotype of Rab3A Ϫ/Ϫ mice that would not only better characterize the mechanism of exocytosis in ␤-cells, but also reveal novel insight into insulin secretory dysfunction found in type-2 diabetes.

EXPERIMENTAL PROCEDURES
Materials-The EasyTag TM Expre 35 S 35 S protein labeling mix from PerkinElmer Life Sciences, containing 73% of L-[ 35 S]methionine, was used for islet protein synthesis radiolabeling. Uridine 5Ј-[␣-(King of Prussia, PA). Rab3A polyclonal antibody was from Santa Cruz Biotechnology Inc. (Santa Cruz, CA), and VAMP-2 antibody was from Calbiochem. The anti-rabbit IgG-horseradish peroxidase conjugate was from Jackson ImmunoResearch (West Grove, PA). All other reagents were of analytical grade and obtained from either Sigma or Fisher.
Animals-The Rab3A ϩ/ϩ on a B6 background (B6129SF2/J) and Rab3A Ϫ/Ϫ mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Mice were housed on a 12-h light/dark cycle and were allowed free access to standard mouse food and water. Mice were used at 12-20 weeks of age.
Glucose, Arginine, and Insulin Tolerance Tests-Glucose (1 mg/g), arginine (1 mg/g), and insulin (0.75 milliunits/g) tolerance tests were performed on 15-17-week-old Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice after an overnight fast by intraperitoneal injection dose relative to body weight as described (14). Blood samples were obtained from the tail vein at the times indicated after the glucose injection. Blood glucose concentrations were measured with a HemoCue blood glucose analyzer (HemoCue AB, Ä ngelholm, Sweden), and plasma insulin levels measured by enzymelinked immunosorbent assay (Crystal Chem, Chicago, IL).
Islet Isolation and in Vitro Insulin Secretion Analysis-Pancreatic mouse islets were isolated by collagenase digestion, and insulin secretory activity examined in static or perifusion incubation studies of isolated islets as described previously (15), in response to various concentrations glucose, 1 nM GLP-1, 5 M glyburide, or 125 M oleate complexed to 1% (w/v) BSA. 1 Histological Analyses-Pancreata from 16-week Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice were removed and fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, rinsed in ethanol, and embedded in paraffin. Serial sections (10 m) were stained with hematoxyline and eosin or immunostained with either anti-insulin or a combination of anti-glucagon/ anti-somatostatin antibodies for visualization of islet ␤-cells and ␣-/␦cells using a Leica confocal microscope as described previously (16). For each section the number and cross-sectional islet area, and the number and size of ␤-cells and ␣-/␦-cells per islet were assessed (2).
Measurement of Glucose Oxidation-Glucose oxidation assays were performed as described previously (17). Briefly, groups of 40 -50 islets were suspended in 200 l of Krebs-Ringer bicarbonate buffer, 16 mM HEPES, and 0.1% (w/v) BSA containing 2.8 mM glucose or 16.7 mM glucose in rounded bottom cups sealed with a rubber-sleeved stopper. Then, ϳ2 ϫ 10 5 cpm of [U-14 C]glucose in samples containing glucose as a substrate was added to the islet suspensions. The cups were then sealed within a 20-ml borosilicate glass scintillation vial using a rubbersleeved stopper and incubated for 2 h at 37°C. Then islet-cell oxidation was halted by the addition of 20 l of 100 mM sodium phosphate buffer (pH 6.0) containing 100 M rotenone to the islets in the cups via the sleeved stoppers. Afterward, 100 l of HClO 4 was added to the islets in the cup, and 300 l of 1 M benzethonium was added to the bottom of the scintillation vials via the appropriate sleeved stoppers. The samples then were incubated for an additional 2 h at 37°C in a shaking water bath. Then the seals were removed, the cups were discarded, and 10 ml of scintillation mixture was added to the vials. The vials then were kept at 25°C overnight before scintillation counting.
Proinsulin Biosynthesis and Preproinsulin mRNA Analysis-Freshly isolated islets from 16-week Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice were incubated for 1 h at 37°C in 200 l of Krebs-Ringer buffer (pH 7.4) containing a basal 2.8 mM or stimulatory 16.7 mM glucose and 0.1% (w/v) BSA. Messenger RNA levels were analyzed by the RNase protection assay, as described previously (18). Immunoprecipitaton analysis of proinsulin biosynthesis in isolated islets pulse-radiolabeled with [ 35 S]methionine was as described previously (18).
Standard Wide-field Epifluorescence Imaging-Dual-wavelength excitation microspectrophotometry was used to measure [Ca 2ϩ ] i as described previously (19). Isolated islets from 18-week Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice were loaded with Fura-2 by a 25-min incubation at 37°C in Krebs-Ringer buffer containing basal 2.8 glucose and 5 M Fura-2/AM (Molecular Probes Inc., Eugene, OR) and then placed into a temperature-controlled perfusion chamber (Medical Systems Inc.) FIG. 1. Confirmation of Rab3A gene knockout in Rab3A ؊/؊ mouse islets and brain. Equivalent total protein (25 g) containing lysates of isolated islets and brain from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice were analyzed for Rab3A, and VAMP-2 as a control, protein expression by immunoblot analysis as described previously (15).
mounted on an inverted epifluorescence microscope (Diaphot, Nikon, Inc.) and perifused by a continuous flow (rate: 2.5 ml/min) of 5% CO 2bubbled Krebs-Ringer buffer at 37°C. Groups of islets are visualized with a ϫ20 quartz objective. Fura-2 dual wavelength excitation at 340 and 380 nm, and detection of single wavelength emission at 510 nm was accomplished using the Metafluor/Metamorph system (Universal Imaging Corp.); images were collected with an intensified CCD camera.
Statistical Analysis-Where appropriate, results are expressed as a mean Ϯ S.E. Statistical analysis was performed by unpaired Student's t test or repeated measure analysis of variance, where p Ͻ 0.05 was considered significant.
In contrast to the IPGTT, an intraperitoneal arginine stimulus (1 mg/g) showed no apparent deficiency in the in vivo insulin secretory response in Rab3A Ϫ/Ϫ versus Rab3A ϩ/ϩ control mice (Fig. 3A). However, when a second intraperitoneal dose (1 mg/g) was given 10 min after the first, the subsequent insulin secretory response to arginine was negated (p Յ 0.05) in Rab3A Ϫ/Ϫ mice compared with an enhanced response in Rab3A ϩ/ϩ control animals (Fig. 3B). Blood glucose levels in Rab3A Ϫ/Ϫ or Rab3A ϩ/ϩ mice did not appreciably alter during the in vivo arginine stimulus (Fig. 3C).
The insulin secretory-deficient phenotype of the Rab3A Ϫ/Ϫ mice did not appear to be based at the level of defective (pro)insulin production. Isolated pancreatic islets from Rab3A Ϫ/Ϫ mice had similar insulin content stores to that of Rab3A ϩ/ϩ mouse islets (Fig. 4A). Likewise, preproinsulin mRNA levels were equivalent in islets from Rab3A Ϫ/Ϫ versus Rab3A ϩ/ϩ mice (Fig. 4B), and translational control of glucose-induced proinsulin biosynthesis was unaffected in isolated Rab3A Ϫ/Ϫ mouse islets (Fig. 4C). Immunohisotchemical examination of pancreata from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice indicated no discernable difference in islet architecture, with insulin-expressing ␤-cells in the central core of an islet and glucagon expressing ␣-cells and somatostatin expressing ␦-cells around the periphery (Fig. 4). Further analysis of pancreatic serial sections showed no obvious change between Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice in islet number per pancreas, islet size, number of ␤-cells, and non-␤-cells per islet or size of ␤-cells and non-␤-cells per islet (data not shown). The weight and size of pancreata from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice were equivalent, and as such it is reasonable to deduce that there was no change in pancreatic ␤-cell mass between Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice. It follows  14). B, a double intraperitoneal arginine stimulation test was performed on 16-h fasted Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice 10 min apart, each a 1 mg/g dose of arginine as indicated by the arrows, and plasma insulin levels measured. A mean Ϯ S.E. of plasma insulin levels are shown, where q represents Rab3A Ϫ/Ϫ mice (n ϭ 16), and f represents Rab3A ϩ/ϩ mice (n ϭ 14). C, blood glucose levels were also measured during the double intraperitoneal arginine stimulation test. A mean Ϯ S.E. of blood glucose levels are shown, where q represents Rab3A Ϫ/Ϫ mice (n ϭ 8), and f represents Rab3A ϩ/ϩ mice (n ϭ 7). that the insulin secretory-deficient phenotype of Rab3A Ϫ/Ϫ mice does not reside at the level of insufficient insulin production.
An increase in ␤-cell glucose metabolism is required to generate secondary signals for glucose-induced insulin secretion (4,21) and proinsulin biosynthesis (17,21). We compared the rate of glucose oxidation in isolated islets from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice and found that there was no significant difference in glucose oxidation at either a basal 2.8 mM or stimulatory 16.7 mM glucose concentration (Fig. 5). As such, the insulin secretory deficiency in Rab3A Ϫ/Ϫ mouse ␤-cells does not lie at the level of defective glucose metabolism.
Downstream of increased glucose metabolism, a rise in cytosolic [Ca 2ϩ ] i in pancreatic ␤-cells is essential to trigger glucoseinduced insulin secretion (4, 5, 7). We examined whether the insulin secretory-deficient phenotype of Rab3A Ϫ/Ϫ mice was due to a defect in generating secondary signals required to trigger insulin exocytosis by measuring glucose-induced increases in ␤-cell [Ca 2ϩ ] i . Changes in cytosolic [Ca 2ϩ ] i was monitored by Fura-2 fluorescence imaging as described previously (19). Increasing the glucose concentration from a basal 2 mM glucose to a stimulatory 14 mM glucose induced a significant rise in cytosolic [Ca 2ϩ ] i , after a short lag period, that was equivalent in isolated islet ␤-cells from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mice (Fig. 6). However, there was a subtle difference in subsequent glucose-induced oscillations in [Ca 2ϩ ] i . Although the amplitude of the [Ca 2ϩ ] i oscillations did not appreciably change the frequency of [Ca 2ϩ ] i , oscillations were slower in Rab3A Ϫ/Ϫ compared with Rab3A ϩ/ϩ islet ␤-cells (Fig. 6, A versus B). Nonetheless, despite this disparity in glucose-induced [Ca 2ϩ ] i oscillations the total increase in [Ca 2ϩ ] i was only 4% lower in Rab3A Ϫ/Ϫ islet ␤-cells. This was reaffirmed in that depolarization of islet ␤-cells with 30 mM KCl at 2 mM glucose FIG. 4. Insulin production is unaffected in Rab3A ؊/؊ mice. A, total insulin content was measured in isolated Rab3A Ϫ/Ϫ (open bar) and Rab3A ϩ/ϩ (closed bar) mouse islets by enzymelinked immunosorbent assay (n Ն 18). B, preproinsulin mRNA and, as a control, glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA levels were measured by nuclease protection assay in isolated islets from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mouse islets incubated for 1 h at basal 2.8 mM or stimulatory 16.7 mM glucose as described under "Experimental Procedures." A representative autoradiograph analysis is shown. C, proinsulin biosynthesis was analyzed by immunoprecipitation of gave an identical rapid increase in cytosolic [Ca 2ϩ ] i in both in Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ islet ␤-cells, which promptly returned to basal cytosolic [Ca 2ϩ ] i levels on removal of the stimulus (Fig.  6). As such, there was not a noticeable defect in inducing a rise in cytosolic [Ca 2ϩ ] i in Rab3A Ϫ/Ϫ mouse islet ␤-cells that was sufficient to explain the insulin secretory deficiency in vivo.
Characterization of regulated insulin secretion from isolated islets indicated a defect downstream of generating secondary signals in Rab3A Ϫ/Ϫ islet ␤-cells, at the level of insulin exocytosis. In static in vitro incubation experiments, basal insulin secretion from isolated Rab3A Ϫ/Ϫ mouse islets was equivalent to control Rab3A ϩ/ϩ mouse islets (Fig. 7, A and B). However, secretagogue-stimulated insulin release was compromised in Rab3A Ϫ/Ϫ islets. Qualitatively, the glucose dose-response pattern for stimulated insulin secretion from Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ islets was similar, with a typical threshold glucose concentration between 5 and 6 mM glucose reaching a maximum insulin secretory rate at Ͼ15 mM glucose (Fig. 7A). However, Rab3A Ϫ/Ϫ islets secreted only 30 -40% of the amount of insulin secreted by control Rab3A ϩ/ϩ islets (Fig. 7A). Likewise, the potentiation of glucose-induced insulin secretion by GLP-1, the sulfonylurea, glyburide, or the fatty acid, oleate, was decreased by ϳ60% in Rab3A Ϫ/Ϫ islets compared with control Rab3A ϩ/ϩ islets (Fig. 7B). Examination of the biphasic pattern of insulin secretion in perifused islets from Rab3A Ϫ/Ϫ mice indicated the kinetics for glucose-stimulated insulin secretion to be comparable with Rab3A ϩ/ϩ islets; however, a major abnormality was found in a blunted first phase of insulin release (between 3 and 13 min) that consequentially diminished the second phase in Rab3A Ϫ/Ϫ islets (Fig. 7C). As a result, the accumulated glucose-induced insulin secretion over 40 min in perifused Rab3A Ϫ/Ϫ islets was significantly reduced by ϳ70% compared with control Rab3A ϩ/ϩ islets (Fig. 7D).

DISCUSSION
Rab3A Ϫ/Ϫ mice are glucose-intolerant and exhibit loss of first-phase insulin secretory response to glucose, but the insulin secretory response to a single dose of arginine was similar in Rab3A Ϫ/Ϫ versus Rab3A ϩ/ϩ control mice, which are typical characteristics of insulin secretory dysfunction in type-2 diabetes (1,3). This insulin secretory dysfunction in the Rab3A Ϫ/Ϫ mice did not appear to be driven by peripheral insulin resistance, but rather by a primary defect in the pancreatic ␤-cell. This was reaffirmed in that stimulated insulin secretion from Rab3A Ϫ/Ϫ mouse isolated islets in vitro was markedly decreased compared with control Rab3A ϩ/ϩ islets. However, since Rab3A Ϫ/Ϫ mice may have impaired secretory function in other neuroendocrine cells (11), it is possible that the ␤-cell secretory defect could be secondary, particularly when considering hypothalamic or adrenergic cells. However, when isolated islets from Rab3A Ϫ/Ϫ mice are cultured for 24 or 48 h, that would give time to recover from any in vivo influence of circulating factors, the insulin secretory response to glucose Ϯ glyburide, GLP-1, or oleate remained 60 -70% inhibited compared with similarly cultured Rab3A ϩ/ϩ control islets (data not shown), as found in freshly isolated islets (Fig. 7). Moreover, if the blunted insulin secretory response in ␤-cells were secondary to increased ␣-adrenergic activity, one would also expect to see a blunted first phase insulin secretory response to arginine in vivo (22,23), a decrease in glucose-and K ϩ -induced rise in cytosolic [Ca 2ϩ ] i (24,25), as well as glucose-induced proinsulin biosynthesis (26). However, there were no differences found in these parameters between Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ control mouse islets (Figs. 3A, 4, and 6). Thus, the most likely scenario is that the insulin secretory dysfunction in the Rab3A Ϫ/Ϫ mice is due to a primary defect in the ␤-cell itself.
There was no difference in pancreatic ␤-cell mass in between Rab3A Ϫ/Ϫ versus Rab3A ϩ/ϩ mice and neither was insulin production and intracellular stores of insulin affected in Rab3A Ϫ/Ϫ mouse islets. As such, the insulin secretory deficiency in Rab3A Ϫ/Ϫ mice was not due to an insufficient store or supply of insulin. Glucose metabolism is a prerequisite to generate secondary signals for glucose-induced insulin secretion and proinsulin biosynthesis, but rates of glucose oxidation were similar in Rab3A Ϫ/Ϫ and Rab3A ϩ/ϩ mouse islets (Fig. 5). This indicated that insulin secretory dysfunction in Rab3A Ϫ/Ϫ mice was not due to defective ␤-cell glucose sensing or metabolism. Indeed, the observation that glucose-induced translational control of proinsulin biosynthesis was unaffected in Rab3A Ϫ/Ϫ indicated that stimulus-response coupling mechanisms were intact in Rab3A Ϫ/Ϫ ␤-cells. This was reaffirmed in that glucose-and K ϩ -induced depolarization of Rab3A Ϫ/Ϫ islet ␤-cells caused an increase in cytosolic [Ca 2ϩ ] i similar to that in Rab3A ϩ/ϩ islet ␤-cells. As such, the defect in insulin secretion in Rab3A Ϫ/Ϫ mouse ␤-cells most likely lies downstream of generating secondary signals, at the level of insulin exocytosis. A defective exocytosis mechanism in Rab3A Ϫ/Ϫ islet ␤-cells would also be consistent with the observation that the insulin secretory response to all secretagogues tested in vitro was reduced by Ն50% in Rab3A Ϫ/Ϫ islets, despite these secretagogues stimulating insulin secretion via distinct signaling mechanisms (4).
In pancreatic ␤-cells, Rab3A is mostly located on the cytosolic face of ␤-granule membranes and has been implicated to play a role in control of insulin exocytosis (8,9), as it does for Ca 2ϩdependent exocytosis in other neuroendocrine cell types (10). Here, we find that the absence of Rab3A in ␤-cells decreases secretagogue-induced insulin secretion in vivo and in vitro, reaffirming an important role of Rab3A in control of insulin exocytosis. Rab GTP-binding proteins specifically direct vesicular transport by an interaction with a particular "effector protein," characteristic of an individual Rab protein and neuroendocrine cell type (10). Rab3A has been shown to interact with several candidate effector proteins (10), but the Rab3Acalmodulin interaction appears most pertinent in control of Ca 2ϩ -regulated insulin exocytosis (27). Recently, it has been found that the Rab3A-calmodulin interaction on ␤-granules provides a platform for local activation of the Ca 2ϩ /calmodulindependent phosphoprotein phosphatase, calcineurin, in response to increased [Ca 2ϩ ] i (15,28). Activation of calcineurin on ␤-granules then leads to dephosphorylation activation of the ATP-dependent motor, kinesin, and subsequent transport of ␤-granules to a "readily releasable pool" docked at the ␤-cell plasma membrane committed to undergo exocytosis (15). Therefore, in the absence of Rab3A, calmodulin cannot be readily sequestered to the ␤-granule membrane for local activation of calcineurin and kinesin in response to elevated [Ca 2ϩ ] i , so that ␤-granule transport becomes much less efficient, the number of ␤-granules recruited to a readily releasable pool is reduced, and consequently insulin exocytosis is impaired. A defect in replenishing the readily releasable pool of ␤-granules in ␤-cells of Rab3A Ϫ/Ϫ mice would be consistent with the observations of a blunted first-phase glucose-induced insulin release in vitro and in vivo, as well as a severely diminished insulin secretory response to a second arginine stimulus where the prior arginine stimulus would have emptied the readily releasable pool of ␤-granules. Moreover, such a mechanism is also consistent with observations of a decay in synaptic transmissions in hippocampal neurons of Rab3A Ϫ/Ϫ mice in response to multiple repeat Ca 2ϩ -induced stimulations, which are normally sustained as found in Rab3A ϩ/ϩ mice (11).
Our study of the Rab3A Ϫ/Ϫ insulin secretory phenotype indicates that the root of ␤-cell secretory dysfunction in type-2 diabetes does not necessarily lie at the level of glucose-sensing and ␤-cell metabolism from which secondary signals emanate (e.g. a rise in [Ca 2ϩ ] i ), but rather at the level of a defective insulin exocytosis mechanism. This would be symptomatic of insulin secretory dysfunction in type-2 diabetes arising from an "overworked" ␤-cell trying hard, but not quite succeeding, to compensate for peripheral insulin resistance (3). In the face of persistent hyperglycemia in type-2 diabetes, the ␤-cell is chronically stimulated to produce and secrete much higher amounts of insulin than usual, and as such there is a higher rate of ␤-granule turnover. This, in turn, compromises ␤-granule transport that depletes the ready releasable pool of ␤-granules, so that insulin secretory insufficiency develops leading to ␤-cell secretory dysfunction (29). Such is the case with Rab3A Ϫ/Ϫ mice, best illustrated by a severe reduction in the in vivo insulin secretory response to a second consecutive arginine FIG. 7. Secetagogue-induced insulin secretion is impaired from isolated islets of Rab3A ؊/؊ versus Rab3A ؉/؉ control islets. A, insulin secretion from isolated islets incubated for 1 h at 37°C was assessed over a range of glucose concentrations. Results are shown as a mean Ϯ S.E., where q represents Rab3A Ϫ/Ϫ mice (n ϭ 8), and f represents Rab3A ϩ/ϩ mice (n ϭ 8). B, insulin secretion from isolated islets incubated for 1 h at 37°C was assessed at either a basal 2.8 mM glucose or stimulatory 16.7 mM glucose in the additional presence of 1 nM GLP-1, 5 M glyburide, or 125 M oleate complexed to 1% (w/v) BSA as indicated. Results are shown as a mean Ϯ S.E., where open bars represents Rab3A Ϫ/Ϫ mice (n ϭ 6), and closed bars represents Rab3A ϩ/ϩ mice (n ϭ 6). C, insulin secretion from isolated islets was assessed in perifusion incubation experiments at either a basal 2.8 mM glucose or stimulatory 16.7 mM glucose as indicated. Results are shown as a mean Ϯ S.E., where q represents Rab3A Ϫ/Ϫ mice (n ϭ 10), and f represents Rab3A ϩ/ϩ mice (n ϭ 10). D, total insulin secretion from 16.7 mM glucose-induced insulin secretion from perifused isolated Rab3A Ϫ/Ϫ (open bar) versus Rab3A ϩ/ϩ (closed bar) mouse islets as indicated under the curve (AUC) derived from C. Results are shown as a mean Ϯ S.E., where Ⅺ represents Rab3A Ϫ/Ϫ mice (n ϭ 10), and f represents Rab3A ϩ/ϩ mice (n ϭ 10). stimulus that is contrastingly augmented in normal Rab3A ϩ/ϩ animals. However, there are no reports in the literature of an insulin secretory response to such a double consecutive arginine stimulus test in human type-2 diabetics, but if pursued could well reveal a similar insulin secretory insufficiency found in Rab3A Ϫ/Ϫ mice that would be supportive of insulin secretory dysfunction in human type-2 diabetes arising from deficiencies in the insulin exocytotic mechanism due to ␤-cell exhaustion (3,29). In this regard, it should be noted that if the ␤-cell is induced to rest its secretory activity in type-2 diabetics, normal insulin secretory function can be recovered (30). Notwithstanding, the glucose-intolerant/insulin secretory deficient phenotype of Rab3A Ϫ/Ϫ mice, in the absence of insulin resistance or changes in ␤-cell mass, emphasizes the important contribution that insulin secretory dysfunction makes in the pathogenesis of type-2 diabetes and further suggests that protection of ␤-cell function is a worthy consideration for the treatment of type-2 diabetes (2).