Munc13-1 Deficiency Reduces Insulin Secretion and Causes Abnormal Glucose Tolerance

Munc13-1^+/− knockout and Munc13-1^H567K/+ knockin mice were described previously (13,19). The genotypes of the mice were determined by PCR and DNA electrophoresis as described (13,19). Mice were housed on a 12-h light/dark cycle and were allowed free access to standard mouse food and water. All experimental procedures have been approved by the Animal Care Committee of the University of Toronto.

Glucose (1 g/kg body wt) tolerance tests were performed about bimonthly on 12- to 32-week-old Munc13-1^+/+, Munc13-1^+/−, and Munc13-1^H567K/+ mice after a 12- to 16-h fasting period by intraperitoneal injection. The purpose of using older mice of up to 32 weeks of age was to assess a possible recovery from or worsening of glucose intolerance in Munc13-1^+/− and Munc13-1^H567K/+ mice. Within each group, there was no change in glucose tolerance over the period, and the data were therefore pooled. Blood samples were obtained from the tail vein at the times indicated following glucose injection. Blood glucose concentrations were measured with a Lifescan OneTouch basic glucose meter (Lifescan Canada, Burnaby, Canada). During the glucose tolerance test (above), blood samples were also obtained from the legs’ saphenous vein at the times indicated after the glucose load. Serum was separated from plasma by centrifugation and stored at −20°C until serum insulin concentrations were determined using a radioimmunoassay kit (Linco Research, St. Louis, MO). For the insulin tolerance test, human biosynthetic insulin (Novo Nordisk, Toronto, Canada) was injected intraperitoneally at a dose of 0.55 units/kg body wt into mice after a 5-h fasting period. A drop of blood was withdrawn from the tail vein at the indicated times following insulin administration.

Mice were sacrificed by cervical dislocation. The abdominal cavity was opened, the pancreatic duct cannulated under a dissecting microscope, and 2 ml of a collagenase solution (2 mg/ml; Worthington Biochemical, Lakewood, NJ) dissolved in Hanks’ balanced salt solution containing 10 mmol/l HEPES (pH 7.4) was injected, which dilates the pancreas. Pancreas was then excised and incubated in the same solution at 37°C for 26 min with shaking to disrupt the tissue. Pancreatic tissues were then filtered through a nylon gauze and washed extensively with collagenase-free Hanks’ balanced salt solution/HEPES. The islets were then hand picked under an upright stereomicroscope and maintained in culture medium. Dispersed single β-cells were obtained by incubating islets in Ca²⁺/Mg²⁺-free PBS containing 5 mmol/l EDTA and 0.25 mg/ml trypsin for 10 min at 37°C with mild shaking. Isolated cells were cultured in RPMI 1640 medium containing 11 mmol/l glucose supplemented with 10% FCS, 100 units/ml penicillin, 100 μg/ml streptomycin, 10 mmol/l HEPES (pH 7.4), and 1 mmol/l sodium pyruvate at 37°C in a humidified atmosphere containing 5% CO_2. β-Cells were allowed to adhere on coverslips for ∼48 h before electrophysiological experiments. β-Cells were cultured for up to 4 days, but the majority of the electrophysiological studies were performed at 2–4 days.

Isolated islets were solubilized in sample buffer (with 2% SDS), and the indicated amount of protein of each sample was loaded and separated on SDS polyacrylamide gels (Munc13-1, 15 μg of protein/lane, 6% PAGE; nsec1/Munc18-1, syntaxin 1A, 5 μg of protein/lane, 14% PAGE; protein kinase C [PKC]-α and -ε, 18 μg of protein/lane, 7.5% PAGE). Separated proteins were transferred to nitrocellulose membranes and incubated with the following rabbit polyclonal primary antibodies for 1.5–2 h at room temperature: anti-Munc13-1, 1:5,000 (19); anti-syntaxin 1A, 1:1,000 (Calbiochem, San Diego, CA); anti-nsec1/Munc18-1, 1:1,000 (Transduction Laboratories, Lexington, KY); anti-PKC-α, 1:1,000 (Sigma); and anti-PKC-ε, 1:1,000 (Sigma). Primary antibodies were detected with peroxidase-labeled goat anti-rabbit secondary antibodies. Positive reactions were visualized by chemiluminescence (Pierce) and exposure of the membranes to Kodak BMR film (Eastman Kodak, Rochester, NY) for 1 s to 10 min.

To allow freshly isolated islets to recover after isolation, islets from Munc13-1^+/+, Munc13-1^+/−, and Munc13-1^H567K/+ mice at ages 4–8 months were cultured for 2 h as described above in medium containing 11 mmol/l glucose. The islets were then cultured for 2 h in the same conditions except that the medium contained 2.8 mmol/l glucose. Afterward, the similar-size islets were picked and distributed in batches of ∼30 into 12-well plates, with four replicates for each condition per genotype. They were then washed and preincubated for 40 min at 37°C in 1 ml Krebs-Ringer bicarbonate buffer containing 10 mmol/l HEPES (KRBH; pH 7.4), 0.5% BSA, and 3 mmol/l glucose. The islets were then incubated for 1 h at 37°C in 1 ml of KRBH containing 3 mmol/l glucose, followed by 1 h at 37°C in 1 ml of KRBH containing 16 mmol/l glucose in the absence or presence of 200 nmol/l phorbol 12-myristate 13-acetate ester (PMA) (Sigma). At the end of each 1-h incubation, media were collected for insulin determination. Total insulin content of the islets was measured after acid-ethanol (0.2 mmol/l HCl in 75% ethanol) extraction. Insulin was measured by radioimmunoassay, as described above, and insulin secreted into the supernatant was expressed as percentage of total islet insulin content per hour to ensure accuracy and reduce intra- and interassay variation.

Patch electrodes were pulled from 1.5-mm thin-walled borosilicate glass and polished to a tip resistance of 2–4 MΩ when filled with intracellular solution. β-Cells were identified by the lack of a transient voltage-dependent inward Na⁺ current as described by Gopel et al. (23). For capacitance measurements, the patch electrodes were coated with orthodontic wax (Butler, Guelph, Canada) close to the tips and fire polished. The pipette solution contained (in mmol/l): 125 K-glutamate, 10 KCl, 10 NaCl, 1 MgCl₂, 5 HEPES, 0.05 EGTA, and 4 MgATP, pH to 7.1. The extracellular solutions consisted of (in mmol/l): 138 NaCl, 5.6 KCl, 1.2 MgCl₂, 2.6 CaCl₂, 5 HEPES, and 5 d-glucose, pH to 7.4. For the measurements with PMA stimulations, the β-cells were preincubated with 200 nmol/l PMA for ∼10 min, and this concentration was maintained in the incubation chamber during recordings. Cell capacitance was determined by the Lindau-Neher technique (24), implementing the “Sine + DC” feature of the Lock-in module (40 mV peak-to-peak and a frequency of 500 Hz) in the standard whole-cell mode. Recordings were conducted using an EPC-10 patch clamp amplifier and Pulse software (HEKA Electronics, Lambrecht/Pflaz, Germany). Exocytotic events were elicited by a train of eight 500-ms depolarizing pulses (1-Hz stimulation frequency) from −70 to 0 mV. All capacitance measurements were performed at 30°C.

For the measurement of Ca_V currents, pipettes were filled with (in mmol/l): 120 CsCl, 20 tetraethylammonium chloride, 5 EGTA, 5 MgATP, and 5 HEPES (pH 7.2). The external solution was comprised of (in mmol/l): 100 NaCl, 20 BaCl₂, 20 tetraethylammonium, 4 CsCl, 1 MgCl₂, 10 glucose, and 5 HEPES (pH 7.4). Tetraethylammonium was used to block K_V currents. For K_V current measurements, pipettes were filled with (in mmol/l): 125 K-glutamate, 10 KCl, 10 NaCl, 1 MgCl₂, 5 HEPES, 0.05 EGTA, and 4 MgATP, pH to 7.1. The extracellular solutions consisted of (in mmol/l): 138 NaCl, 5.6 KCl, 1.2 MgCl₂, 2.6 CaCl₂, 5 HEPES, and 5 d-glucose, pH to 7.4. To elicit K_V and Ca_V currents, cells were held at −80 mV and depolarized from −80 to +80 mV in 10-mV increments using 500-ms step pulses and from −60 to +80 mV in 10-mV increments using 200-ms step pulses, respectively. Steady-state inactivation properties of Ca_V currents were investigated by depolarizing the cells with 1-s prepulse from −100 to 60 mV. Steady-state inactivation curves were fit with a Boltzmann equation: I/Imax = 1/(1 + exp[V-V1/2]/k), where V1/2 is the voltage at which half of the channels are inactivated, and k is the slope factor. All Ca_V and K_V currents recordings were performed at room temperature.

Where appropriate, results are expressed as means ± SE. Statistical analysis was performed by unpaired Student’s t test or repeated measure ANOVA, where P < 0.05 was considered significant.

Homozygous Munc13-1^−/− and Munc13-1^H567K/H567K mice exhibit severely impaired neurotransmitter release and die shortly after birth (13,19). We have therefore used heterozygous mice (+/− and ^H567K/+) to examine the effects of Munc13-1 deficiency on glucose homeostasis and insulin secretion. The heterozygous mice do not exhibit neurological deficits (13,19). The expression levels of Munc13-1 in Munc13-1^+/+, Munc13–1^+/−, and Munc13-1^H567K/+ mouse brains were determined by Western blotting using a rabbit antibody generated against the NH₂-terminal amino acids 1–305 of Munc13-1 (19,20). Consistent with previous reports (19), Munc13-1 levels were significantly reduced by 20% (n = 2) in the brains of Munc13-1^+/− mice compared with wild-type Munc13-1^+/+ control mice (Fig. 1A). Brain Munc13-1 levels were similar in Munc13-1^H567K/+ mice and wild-type controls, since the H567K mutant protein is of the same size as the native Munc13-1 protein (19). Brain levels of syntaxin 1A and nSec1/Munc18-1 were identical in the three groups of mice. Islet tissue levels of Munc13-1 were reduced by 60% (n = 4) in Munc13-1^+/− mice as compared with wild-type islets, whereas those of Munc13-1^H567K/+ was 97% of those of Munc13-1^+/+ islet levels (Fig. 1B). The levels of syntaxin 1A, nSec1, and PKC-α and -ε in the islets of Munc13-1^+/− and Munc13-1^H567K/+ mice were similar to those of wild-type Munc13-1^+/+ islets. The results indicate that possible defects of insulin secretion in the islet β-cells from the two types of heterozygous mutants would be attributed to the Munc13-1 deficiency per se and not to a deficiency in the expression of Munc13-1–interacting proteins or the other DAG-binding proteins, specifically the putative PKC isoforms reported to influence insulin secretion (25–27). In contrast, complete loss of Munc13-1 (−/−) induces a mild reduction in a number of the t-SNARE syntaxin and several associated proteins in the brain (13,19). A major reduction in the levels of SNARE proteins and nSec1 also accompanies Munc13-1 deficiency in the diabetic GK and Zucker rat islets as we have previously reported (20).

Intraperitoneal glucose tolerance tests (IPGTTs) indicated that Munc13-1^+/− mice displayed a glucose-intolerant phenotype (Fig. 2). Male Munc13-1^+/− mice exhibited significantly (P ≤ 0.05) higher glucose levels in response to a glucose load at t = 15 and 30 min. Fasting glucose levels were, however, not significantly different between Munc13-1^+/− (6.8 ± 0.3 mmol/l) and wild-type Munc13-1^+/+ control (6.6 ± 0.2 mmol/l) mice. Munc13-1^H567K/+ mice also showed normal fasting glycemia (6.8 ± 0.2 mmol/l). Surprisingly, the Munc13-1^H567K/+ mice exhibited overall normal glucose tolerance, indicating that expression of the Munc13-1 mutant variant may not lead to sufficient insulin secretory deficiency to cause an abnormal glucose homeostasis, although glucose levels at t = 20–30 min were slightly but not significantly elevated. We next investigated whether this abnormal glucose homeostasis is due to either a defect in insulin secretion or a defect in insulin sensitivity of the peripheral tissues. Following insulin injection, glucose levels were similarly lowered in the three groups, indicating no difference in insulin sensitivity (Fig. 3).

Munc13-1^+/− mice showed similar fasting serum insulin levels (0.41 ± 0.04 ng/ml) as those of Munc13-1^+/+ mice (0.47 ± 0.05 ng/ml). However, in response to a glucose load during IPGTT, Munc13-1^+/− mice consistently showed lower serum insulin levels with respect to Munc13-1^+/+ mice (Fig. 4A). The t = 10 min level, which corresponds to the first phase of insulin release, was lower in Munc13-1^+/− (0.65 ± 0.07 ng/ml) compared with wild-type (0.95 ± 0.08 ng/ml) mice. Serum insulin levels at t = 30 min, which corresponds to the second phase of insulin release, was also lower in the Munc13-1^+/− mice (0.52 ± 0.03 ng/ml) as compared with wild-type (0.73 ± 0.05 ng/ml) mice. A similar observation was made at t = 120 min. Insulin secretion as indirectly assessed as the area under the curve representing the time course of release over the entire 2-h period, encompassing first- and second-phase release, showed an ∼50% reduction in Munc13-1^+/− mice (Fig. 4B). As for the Munc13-1^H567K/+ mice, serum insulin levels were lower at t = 10 min (0.59 ± 0.15 ng/ml), but during the second phase of insulin release, insulin levels were similar at 30 and 120 min. The area under the curve in Munc13-1^H567K/+ mice showed no difference with respect to wild-type mice.

We then examined insulin secretion from freshly isolated pancreatic islets of the different Munc13-1 mutant mice (Fig. 5). There was no difference in basal (3 mmol/l glucose) insulin secretion between the groups. However, GSIS (16 mmol/l glucose-stimulated secretion above basal) was greatly reduced in Munc13-1^+/− mice islets with respect to Munc13-1^+/+ mice islets. Relative to basal insulin release, Munc13-1^+/− islet GSIS was 1.68-fold basal release, which is a reduction of 52% with respect to Munc13-1^+/+ mice islets (3.51-fold basal release) and a reduction of 43% with respect to Munc13-1^H567K/+ mice (Fig. 5A). The results corroborate the above in vivo results (Fig. 4B). Diacylglycerol (DAG) and phorbol esters are known to potentiate insulin secretion, mainly by increasing the size of insulin granules pools, and by accelerating the mobilization of insulin granules for subsequent priming and release (2). We therefore assessed the potentiation of GSIS by PMA, a phorbol ester that binds Munc13-1 at the same site as DAG (18–20). Indeed, 200 nmol/l PMA stimulation potentiated 16 mmol/l GSIS in Munc13-1^+/− islets (Fig. 5B), indicating that the residual Munc13-1 proteins in the islets of heterozygous Munc13-1^+/− mice are functional. Nevertheless, PMA potentiation of GSIS relative to 16 mmol/l GSIS was greatly suppressed by 49.2% in Munc13-1^+/− islets (2.42-fold) with respect to Munc13-1^+/+ islets (4.76-fold) (Fig. 5B). These results indicate that the binding of PMA to the residual Munc13-1 proteins in the Munc13-1^+/− islets, but not PKC, which is not deficient in these transgenic mice (Fig. 1B), primarily accounts for enhancement of insulin secretion by PMA.

For the Munc13-1^H567K/+ islets, GSIS was mildly but not significantly reduced when compared with Munc13-1^+/+ mice islets (Fig. 5A), corroborating the above in vivo insulin secretion data (Fig. 4B). The potentiation (secretion above that evoked by 16 mmol/l glucose alone) of insulin secretion by 200 nmol/l PMA was nonetheless modestly reduced but less so than in Munc13-1^+/− islets (Fig. 5B). The reason for the better GSIS response of the Munc13-1^H567K/+ islets compared with the Munc13-1^+/− islets is that the Munc13-1 H567K DAG-binding site mutant protein, although defective in its ability to translocate to the plasma membrane, contains an intact priming domain that is distinct from the DAG-binding site (28,29) and therefore is able to retain some priming function (19,22). Nonetheless, these results show that an intact C1 PMA/DAG-binding domain of Munc13-1 is required for a fully efficacious potentiation of GSIS by PMA.

The differences in GSIS between the groups suggest the possibility of compensatory mechanisms such as increased islet mass (30) or increased (pro)-insulin biosynthesis. Total insulin contents were comparable between isolated islets of the three groups (Fig. 5C), which argues against a difference in insulin biosynthesis as a compensatory mechanism explaining the differences in insulin secretory responses in these mice.

To examine the direct contribution of islet β-cell insulin secretory defect in the Munc13-1^+/− mice, we employed patch-clamp membrane capacitance (C_m) measurements of single-islet β-cells. We used a protocol of eight serial depolarizing pulses where the first two pulses would release the initially primed RRP of insulin granules, and subsequent pulses would cause the mobilization of granules from the reserve pool(s) to refill the releasable pool (31,32). Compared with Munc13-1^+/+ β-cells, C_m increases in Munc13-1^+/− β-cells were significantly reduced at every depolarizing pulse, with a 62% reduction of RRP size (6.45 ± 0.69 vs. 2.42 ± 0.36 fF/pF at 2nd pulse, P < 0.05) and a 36% reduction of granules refilling/mobilization (ΔCm_3rd-8thpulse) (7.65 ± 0.73 vs. 4.86 ± 0.33 fF/pF, P < 0.05) (Fig. 6A and C). This result is consistent with the result of the whole-islet secretion assays (Fig. 5). PMA pretreatment enhanced insulin exocytosis in Munc13-1^+/+ β-cells by threefold compared with no PMA pretreatment (39.5 ± 3.18 vs. 13.23 ± 0.72 fF/pF at 8th pulse, P < 0.05), where the RRP size was 15.54 ± 2.03 fF/pF at 2nd pulse, and ΔCm_3rd-8thpulse was 24.56 ± 2.80 fF/pF (Fig. 6B and D). In contrast, PMA potentiation of insulin exocytosis in Munc13-1^+/− β-cells was reduced to only twofold (17.33 ± 1.28 vs. 7.46 ± 0.50 fF/pF at 8th pulse), where the RRP size is 7.18 ± 0.73 fF/pF at 2nd pulse, and ΔCm_3rd-8thpulse was 11.98 ± 0.80 fF/pF (Fig. 6B and D). In absolute terms, the cumulative increase in C_m was reduced even more (by 56%) in Munc13-1^+/− β-cells compared with Munc13-1^+/+ β-cells (17.33 ± 1.28 vs. 39.5 ± 3.18 fF/pF), including a 54% reduction (P < 0.05) in the RRP (ΔCm_{1st–2nd pulse}) and 51% reduction (P < 0.05) in the rate of granules pool mobilization (ΔCm_{3rd–8th pulse}). These results suggest that a deficiency in Munc13-1 expression greatly impairs Ca²⁺-evoked and PMA-potentiated exocytosis, specifically by reducing the size of the primed RRP and impairing the rate of granules refilling/mobilization.

As for Munc13-1^H567K/+ β-cells, there is also a reduction in the initial size of the primed RRP compared with Munc13-1^+/+ β-cells (3.83 ± 0.94 v.s. 6.45 ± 0.69 fF/pF at 2nd pulse, P < 0.05) (Fig. 6A and C). However, the subsequent pulses evoked C_m increases (ΔCm_{3rd–8th pulse}) in Munc13-1^H567K/+ β-cells comparable to those in Munc13-1^+/+ β-cells, indicating that the rate of refilling of the RRP was not impaired (8.49 ± 1.21 vs. 7.65 ± 0.73 fF/pF, P = 0.56) (Fig. 6A and C). These results are consistent with the in vivo results showing an impaired first-phase but not second-phase secretion in Munc13-1^H567K/+ mice (Fig. 4). This deficiency of Munc13-1 binding to phorbol ester/DAG caused only a mild impairment of priming of granules in RRP as to cause a reduction of the initial size of the RRP but seemed to not affect the refilling of RRP under these stimulatory conditions. However, upon increased demand by PMA pretreatment, the priming deficiency became accentuated by the deficiency of Munc13-1 binding to phorbol ester/DAG, revealing not only a reduction in the initial size of the RRP (by 27%) compared with Munc13-1^+/+ β-cells (11.42 ± 1.07 vs. 15.54 ± 2.03 fF/pF at 2nd pulse, P < 0.05) but, remarkably, also a similar reduction (by 36%) in the refilling (ΔCm_{3rd–8th pulse}) of the RRP compared with Munc13-1^+/+ β-cells (15.76 ± 0.88 vs. 24.56 ± 2.80 fF/pF, P < 0.05) (Fig. 6B and D).

The impaired insulin exocytosis phenotype described above could be caused by perturbations in essentially any step of the insulin vesicle cycle and in particular the Ca²⁺-triggering step (33). Ca²⁺ plays a role not only as a trigger for exocytosis but also in the priming of insulin vesicles (2). It is thus possible that Munc13-1 deficiency could alter Ca²⁺ release pathways directly or indirectly via its actions on syntaxin 1A, the latter having direct effects on Ca²⁺ channels in pancreatic β-cells (34,35). We therefore examined possible alterations in the voltage-gated Ca²⁺ channel activity by the Munc13-1 deficiency. In both Munc13-1^+/− and Munc13-1^H567K/+ β-cells, we observed no significant changes in either Ca_V current amplitudes (Fig. 7A and B) or channel gating (steady-state inactivation; data not shown). The waveform of action potentials is shaped by the activity of voltage-gated K⁺ channels in the pancreatic β-cells, the majority of which are the delayed rectifier, Kv2.1, whose activity is modulated by syntaxin 1A (36,37). However, we observed no significant changes in K_V current amplitudes in the Munc13-1 mutant mice islet β-cells (Fig. 7C and D).

Munc13-1 is an essential secretory vesicle–priming protein (10,13,14) that facilitates the formation of a stable SNARE complex by directly interacting with (38) and switching syntaxin 1A from a “closed” to an “open” conformation (8,16) and thus allowing syntaxin 1A to bind the cognate SNARE proteins (39). As in neurons, we previously provided evidence that Munc13-1 in pancreatic β-cells is the major phorbol ester/DAG receptor involved in phorbol ester/DAG-induced potentiation of insulin exocytosis (20). In this report, we directly show that Munc13-1 reduction alone is sufficient to cause abnormal GSIS and glucose homeostasis mimicking diabetes. The dramatic phenotype of impaired synaptic transmitter release in Munc13-1 knockout mice is a consequence of a reduced primed and RRP size (13,18,19). For neurotransmitter release, total Munc13-1 deletion is required to abrogate priming (13,19). In the present study, a deletion of one Munc13-1 allele resulting in the depletion in islet tissues of Munc13-1 to levels mimicking those seen in diabetic islets (20,21) was sufficient to induce an insulin secretory defect. This occurred with no detectable effects on the levels of Munc13-1–interacting exocytic proteins, which was observed with the complete Munc13-1 deletion (13). These observations in brain and islet tissues may be explained by a functional redundancy of Munc13 isoforms in neurons to compensate for the Munc13-1 deficiency (19) but which may not be present or abundant in pancreatic islets.

In diabetes, the first phase of insulin secretion is abrogated, and the subsequent sustained second-phase response is blunted. As in neurons, the primed RRP granules pool accounting for the first phase of insulin release requires the formation of the SNARE ternary complex (2). The reduced levels of Munc13-1 in type 2 diabetes rat models (20) and diabetic human islets (21) could therefore be a major contributing factor to the reduced first phase of insulin secretion as observed with the Munc13-1^+/− mice. The second phase of release is accounted for by additional insulin granules being recruited to refill the RPP (4). We show that the second phase of insulin release was also severely reduced in the Munc13-1^+/− mice, indicating the important role of Munc13-1 in both phases of insulin secretion. Indeed, our C_m studies support this thinking, demonstrating that both the initial size of primed RRP and subsequent mobilization of granules from the reserve pool to the RRP are reduced in the Munc13-1^+/− β-cells (Fig. 6A and C).

A surprising result is that the Munc13-1^H567K/+ mice exhibited only a reduced first phase of insulin release (Fig. 4A) but a normal second phase, which was not sufficient to induce an abnormal glucose tolerance. Consistently, the C_m study showed only a reduced sized of the primed RRP but normal mobilization of the reserve pool (Fig. 6A and C). The normal glucose tolerance was not due to a compensatory increase in islet mass or insulin biosynthesis since total insulin content per islet (Fig. 5C) remained unaltered. It could be attributed to possible compensatory mechanisms such as the actions of the other islet secretory cells, which could also be regulated by Munc13-1 priming actions, including a reduction of paracrine inhibition by somatostatin released by islet δ cells or a reduction of counterregulatory neuronal and endocrine modulation. Nonetheless, when additional demand by PMA potentiation was imposed on the Munc13-1^H567K/+ β-cells, a deficiency in mobilization of the reserve pool to the RRP was also observed, albeit less severe than in Munc13-1^+/− β-cells (Fig. 6B and D). This is corroborated by reduced PMA potentiation of GSIS in Munc13-1^H567K/+ islets (Fig. 5B). Although our C_m protocol would be considered to be supraphysiologic stimulation, high and sustained glycemic demand on the islets certainly occur in diabetes and would have a similar imposition on the deficient islet Munc13-1 levels (20,21). Taken together, these results suggest that Munc13-1 is a major phorbol ester and DAG substrate in the islet β-cell to prime insulin granules within the RRP and which determines the initial size of the RRP and the efficiency of mobilization of granules into the RRP.

These results also complement our previous studies (20,40) showing that an intact C1 domain is essential for soluble Munc13-1 translocation to the plasma membrane in response to phorbol ester and DAG activation. Translocation of soluble Munc13-1 to the plasma membrane is thought to create additional priming sites, particularly when the membrane-bound Munc13-1 in the presynaptic active zone is functionally saturated (19). The Munc13-1^H567K/+ mice islet β-cells secreted more insulin than Munc13-1^+/− mice, which we attribute to the fact that the Munc13-1^H567K mutant variant, although defective in its C1 domain, still has some capability of transducing its priming action since the priming domain was recently localized to a α-helical domain in the COOH-terminus (28,29,41). Consistently, overexpression of Munc13-1^H567K–EGFP (enhanced green fluorescent protein) in insulinoma cells (20) and adrenal chromaffin cells (40) potentiated exocytosis over vector-transfected controls but less than overexpression of the wild-type Munc13-1 protein.

In Munc13-1^H567K/+ mice, the normal second phase of release by GSIS (Fig. 4A) and normal mobilization of insulin granules into the RRP (Fig. 6A and C) indicate that additional cell signals or molecules in islet β-cells may come into play during the second phase, which are capable of compensating for the milder Munc13-1 dysfunction in the Munc13-1^H567K/+ islets, but become less efficient when Munc13-1 levels are reduced as in the Munc13-1^+/− islets (and in diabetes [20,21]). The major DAG substrates include many PKC isoforms, particularly cPKC (PKC-α) and nPKC (PKC-ε), which were unchanged in the islets of the three mice (Fig. 1B). Nonetheless, this does not preclude the possibility that normal levels of these (25,27,42) as well as other PKCs, such as cPKC-βII and nPKC-δ (26,43,44), may also serve as DAG receptors capable of being activated, undergo cellular translocation, and exert additional and independent priming actions on insulin exocytosis. More studies performed on these mice islets will be required to see if these PKCs act either sequentially to Munc13-1 or in parallel, perhaps in a manner capable of bypassing the Munc13-1 deficiency.

l-Type voltage-gated calcium channels in pancreatic β-cells provide the primary source of Ca²⁺ entry into the pancreatic β-cells to trigger insulin exocytosis (3). However, the density of the channels in pancreatic β-cells is considerably lower than that in synapses of nerve cells and in adrenal chromaffin cells (3,45). It was suggested that a subset of insulin granules in the RRP is situated in the immediate vicinity of the l-type Ca²⁺ channels to allow efficient triggering of exocytosis by Ca²⁺ entry (3). This was proposed to be caused by tethering of the Ca²⁺ channels to the subset of granules within the RRP (45). Ca²⁺ channels as well as the major membrane-repolarizing K⁺ channel in islet β-cells, Kv2.1, are also profoundly influenced by distinct conformations of syntaxin 1A (36,37,46). It is therefore possible that a Munc13-1 deficiency could indirectly influence Ca²⁺ or Kv2.1 channel gating by actions on syntaxin 1A. However, we observed here that Munc13-1 deficiency did not alter voltage-gated Ca²⁺ or Kv channels activity. Moreover, calcium ionophores cannot bypass the Munc13-1 mutant phenotype in hippocampal neurons, indicating that Munc13-1 deficiency does not interfere with Ca²⁺ triggering of transmitter release (13).

In summary, our study shows that a 60% reduction of Munc13-1 expression or function is sufficient to cause defects in both first and second phases of insulin release, more so than in the neuronal synapse, and is in fact sufficient to cause abnormalities in glucose homeostasis mimicking type 2 diabetes. Since the RRP in the islet β-cells is much smaller than in neurons and chromaffin cells (47) and because islet β-cells express only low levels of Munc13-1 proteins, a smaller perturbation of Munc13-1 expression or function would be expected to cause defects in insulin exocytosis. In fact, pancreatic islet Munc13-1 levels in type 2 diabetic rodents (20) and humans (21) are reduced to levels similar to those observed in the Munc13-1–deficient mice studied. Thus, the results provide direct evidence that Munc13-1 is an important player in the transduction mechanism of the β-cell in response to glucose. It can be hypothesized that glucose-induced increases in DAG following inhibition of fat oxidation by malonyl CoA (48) or enhanced lipolysis (49) promote insulin exocytosis via the DAG receptor Munc13-1. As such, Munc13-1 may be the late-signaling effector of fuel-induced insulin secretion. The possibility should be considered that upregulation of Munc13 levels or priming function may be therapeutically effective in restoring the defective insulin secretion in diabetes.

This work was supported by grants to H.Y.G. from the Canadian Institutes of Health Research (CIHR; MOP64465) and equipment grants from CIHR (MMA 66939) and the University of Toronto Banting and Best Diabetes Center (BBDC). E.P.K. is supported by a Canada Graduate Scholarships Doctoral Award from CIHR (CGD 76317) and the University of Toronto-BBDC (Tamarack Studentship Award). N.B. is supported by the German Research Foundation (SFB406/A1).