Munc18c Heterozygous Knockout Mice Display Increased Susceptibility for Severe Glucose Intolerance

The rabbit anti-GLUT4 and rabbit anti-Munc18c antibodies were obtained as previously described (32). The syntaxin 4, VAMP2, SNAP-23, and transferrin receptor antibodies were purchased from Chemicon (Temecula, CA), Synaptic Systems (Gottingen, Germany), Affinity Bioreagents (Golden, CO), and Zymed (South San Francisco, CA), respectively. The Akt and phosphoserine (Ser 473)-specific Akt antibodies were purchased from Cell Signaling (Beverly, MA). Goat anti-rabbit horseradish peroxidase (HRP) and anti-mouse HRP secondary antibodies were purchased from Bio-Rad (Hercules, CA). The rat insulin radioimmunoassay kit was acquired from Linco Research (St. Charles, MO).

A panel of PCR oligonucleotide primers was designed spanning the entire murine Munc18c cDNA. PCR amplification of mouse liver DNA using the O/F primer pair (O primer: 5′-GGAAAGAAGGCGAATGGAAG; F primer: 5′-TCACTCCAGGGTTTTCATCC) yielded a 506-bp product. DNA sequencing revealed that this matched the cDNA sequence, indicating an absence of intronic DNA. The O and F primers were sent to Genome Systems (St. Louis, MO) for a PCR-based screen of the murine 129/Sv/J library and yielded two 120-kb bacterial artificial chromosome (BAC) clones containing the Munc18c gene. Digestion with BamHI yielded a 9-kb fragment of the Munc18c gene. This 9-kb fragment was isolated from each of the two BAC clones, and each was subcloned into pBluescript (Stratagene, La Jolla, CA) and sequenced in both directions complemented by Southern analysis using the O/F DNA; two exons spanning ∼1.4 kb of the genomic sequence were contained in this 9-kb fragment. The O/F DNA was subjected to fluorescence in situ hybridization (Incyte Genomics, St. Louis, MO), which showed the location of this gene at a position within 15% of the distance from the heterochromatic-euchromatic boundary to the telomere of chromosome 19, an area that corresponds to band 19B. A total of 80 metaphase cells were analyzed, with 73 exhibiting specific labeling.

The Munc18c targeting vector, pMunc18c-KO, was constructed using the positive-negative selection vector, pOSDUPDEL (Gene Targeting, University of Iowa, Iowa City, IA). The 5′ homologous region in the targeting vector was a 4-kb Not I−Xho I fragment containing all intronic DNA upstream of the exon that includes the start codon, taken from the pBluescript 9-kb clone using a Not I site from the vector just 20 bp upstream of the 5′ BamHI site of the 9-kb fragment. This fragment was inserted into the Not I−Sal I cloning site upstream of the neomycin gene (neomycin cassette is in reverse orientation) in the pOSDUPDEL vector. The 3′ homologous region in the targeting vector was generated by PCR amplification of the 568-bp intron separating the two exons in the 9-kb segment of the sequenced gene, engineered with Xho I−Xba I ends and ligated into the Xho I−Nhe I sites of pOSDUPDEL just downstream of the neomycin cassette. The entire DNA region spanning the 5′ homologous region, the neomycin cassette, and the 3′ homologous region was sequenced for verification. The vector also contained a thymidine kinase cassette distal to the 3′ homologous region. The targeted gene therefore lacked a 1.3-kb region that included the 100-bp exon containing the start codon and the 1 kb of intronic DNA immediately proximal to the exon.

The targeting construct was linearized with Not I and introduced into 2 × 10⁷ pluripotent embryonic stem cells (R1 clone) by electroporation (Gene Pulser; Bio-Rad). Embryonic stem clones that were G418- and ganciclovir-resistant were isolated, amplified, and screened for targeting fidelity by Southern blot analysis. Two targeted clones were obtained from the 253 that were analyzed, and both were subsequently reconfirmed by Southern blot analysis by hybridizing the HindIII-digested embryonic stem DNA with the O/F probe. Cells from the two targeted clones (AB46 and AB49) were microinjected into donor C57BL/6J blastocysts and implanted into pseudopregnant recipients. Chimeric animals resulting from the microinjections were bred to C57BL/6J mice, and agouti pups were screened from germline transmission of the mutant allele. The genotypes from these matings and all subsequent matings were determined by PCR on DNA from tail-biopsy specimens using primers that annealed to the 5′ and 3′ flanking DNA of the neomycin substitution site (AA: 5′-CTTGTGTTCGTGGGTGTTCC; AH: 5′-GCTCATGTATGAGAGCAAAG), whereby the PCR product of the WT allele was 1,300 bp and that of the targeted allele was 1,500 bp. An additional two PCR screens were performed to verify the orientation of the insertion site of the neomycin cassette. Founder animals were bred to C57BL/6J stock and offspring carrying the targeted allele were outbred for an additional five generations to enrich the C57BL/6J background at the University of Iowa College of Medicine animal care unit, and carried through an additional two generations upon arrival at the Indiana University School of Medicine laboratory animal resources center, all according to animal care guidelines.

Tissues were homogenized in a 1% Igepal detergent buffer (25 mmol/l Tris [pH 7.4], 1% Igepal, 10% glycerol, 50 mmol/l sodium fluoride, 10 mmol/l sodium pyrophosphate, 137 mmol/l sodium chloride, 1 mmol/l sodium vanadate, 1 mmol/l phenylmethylsulfonyl fluoride [PMSF], 10 μg/ml aprotinin, 1 μg/ml pepstatin, and 5 μg/ml leupeptin) for 15 s then centrifuged at 2,000g for 5 min. The resulting supernatants were microcentrifuged at 13,500g for 20 min at 4°C. Proteins were separated on 10 or 12% SDS-PAGE and then transferred to polyvinylidine fluoride or nitrocellulose membrane for immunoblotting.

Mice (age 3 months) were individually housed for the 10-week high-fat (HF) diet study. The study was conducted in three separate modules with two or more mice per feeding group (WT-HF, Munc18c^−/+-HF, WT−normal diet [ND], or Munc18c^−/+-ND). Each module was conducted at a different time over the span of 9 months in an effort to obtain randomized results from independent litters of mice. Body weight and food intake measurements were taken daily along with the replenishment of fresh diet. The proximate profile of the HF diet (cat. #F3282; Bio-Serv, Frenchtown, NJ) was 20% protein, 35.5% fat, 32.7% carbohydrate, 3.7% ash, 0.1% fiber, and <10% moisture. The diet’s fatty acid composition was 161 g/kg oleic, 89 g/kg palmitic, 44 g/kg stearic, 35 g/kg linoleic, 11 g/kg cis-9-hexadecanoic, 5 g/kg myristic, and 4.7 g/kg cis-11-eicosenoic, with all fatty acids at 1.5 g/kg or less. The caloric profile of the HF diet was 16% protein, 59% fat, and 25% carbohydrate with 5.286 kcal/g energy. The caloric profile of the ND (Teklab cat. # 7,017; Harlan, Indianapolis, IN) was 22% protein, 11% fat, and 60% carbohydrate with 3.41 kcal/g metabolizable energy.

Male Munc18c KO and WT mice (age 4–6 months) were fasted for 18 h before the intraperitoneal glucose tolerance test (IPGTT). After a sample of fasted blood was collected, animals were given glucose (2 g/kg body wt) by intraperitoneal injection; blood glucose readings were then taken at 30-min intervals over 120 min for the IPGTT. For the insulin tolerance test (ITT), male mice (age 4–6 months) were fasted for 6 h. After a sample of fasted blood was collected, animals were injected intraperitoneally with Humulin R (0.75 units/kg body wt); blood glucose readings were then taken after 15, 30, 60, and 90 min. Blood was collected from the tail vein and diluted in saline for measuring blood glucose using the Hemocue glucometer (Mission Viejo, CA). Undiluted blood glucose readings from mice fasted for 18 h ranged from 80 to 100 mg/dl.

Hindlimb skeletal muscle was subfractionated into plasma membrane and intracellular membrane components, as previously described (37). Munc18c^−/+ and WT mice (age 4–6 months) were fasted overnight for 16 h, injected intraperitoneally with 21 units/kg body wt Humulin or vehicle, and killed after 40 min for removal of the soleus and hindquarter muscles into homogenization buffer (20 mmol/l HEPES [pH 7.4], 250 mmol/l sucrose, 1 mmol/l EDTA, 5 mmol/l benzamidine, 10 μg/ml aprotinin, 5 μg/ml leupeptin, 1 μg/ml pepstatin, and 1 mmol/l PMSF) for Polytron homogenization. Homogenates were centrifuged at 2,000g for 5 min at 4°C, and the supernatant was then centrifuged at 9,000g for 20 min at 4°C. The supernatant was centrifuged at 180,000g for 90 min. Pellets were resuspended and combined with sample buffer for separation by 10% SDS-PAGE and subsequent immunoblotting for GLUT4.

Pancreatic mouse islets were isolated as previously described (34). Briefly, pancreases from male mice (age 8–12 weeks) were digested with collagenase and purified using a Ficoll density gradient. After being isolated, islets were cultured overnight in CMRL-1066 medium. Fresh islets were hand-picked into groups of 10, preincubated in Krebs-Ringer bicarbonate buffer (10 mmol/l HEPES [pH 7.4], 134 mmol/l NaCl, 5 mmol/l NaHCO₃, 4.8 mmol/l KCl, 1 mmol/l CaCl₂, 1.2 mmol/l MgSO₄, and 1.2 mmol/l KH₂PO₄) containing 2.8 mmol/l glucose and 0.1% BSA for 2 h, then stimulated with 20 mmol/l glucose for 2 h. Medium was collected to measure insulin secretion and islets were harvested in NP-40 lysis buffer to determine cellular insulin content by radioimmunoassay.

To determine the importance of Munc18c protein on glucose homeostasis in vivo, we cloned a 9-kb piece of the murine Munc18c gene that contained the start AUG codon and constructed a targeting vector that replaced 1.1 kb of the Munc18c genomic sequence. The ATG-coding exon was removed and replaced with a neomycin resistance cassette oriented in the opposite orientation of the endogenous Munc18c gene (Fig. 1A). Of the 253 neomycin-resistant embryonic stem cell clones analyzed, two demonstrated the correct recombination as determined by PCR screening and Southern blot analysis (Fig. 1B and C). Both clones (AB46 and AB49) were used to generate chimeric founder mice. Targeted embryonic stem cell clones and DNA from F1 generation heterozygous mice tail snips exhibited the appearance of the 1,500-bp band in addition to the 1,300-bp endogenous Munc18c band, indicating the addition of the neomycin cassette insertion into the targeted allele (Fig. 1B). Southern blotting using an exon probe confirmed the presence of the 3.3-kb band of the targeted allele in addition to the endogenous 2-kb band of the Munc18c gene (Fig. 1C). Real-time PCR was used to further confirm the reduction in Munc18c mRNA in the Munc18c heterozygous KO mice (data not shown). Heterozygotic offspring of the F1 generation were crossed in an effort to obtain homozygotic offspring; of >150 offspring, none were homozygous, suggesting that homozygotic disruption of the Munc18c gene resulted in embryonic lethality. Subsequent analysis of embryonic development revealed that the Munc18c^−/− embryos died before embryonic day 7.5, and that by embryonic day 9.5, ∼30% of the embryos were WT and 70% were heterozygotic. Subsequent studies were performed using the AB49 line of Munc18c heterozygous KO mice, enriched for the C57BL/6J strain background by more than seven generations of outbreeding into the C57BL/6J strain.

The underexpression of Munc18c protein in the heterozygous^−/+ KO mice was directly compared with endogenous levels in WT littermate mice in heart, skeletal muscle (hindlimb), liver, fat (epididymal), and pancreas by immunoblot analysis (Fig. 2A). Other tissues such as kidney, lung, brain, and spleen also showed reduced expression of Munc18c protein (data not shown). Quantitation by optical density scanning analyses of immunoblots showed Munc18c to be reduced in tissues by 50 ± 9% in the Munc18c heterozygous^−/+ KO mice. In contrast, no differences in abundance of the other SNARE proteins—syntaxin 4, SNAP-23, or VAMP2—were detected in any of the tissue homogenates. No changes in the abundance of the GLUT4 protein were detected in insulin-responsive tissues that displayed characteristic tissue-specific expression such as the heart, skeletal muscle, and fat, but not the liver (Fig. 2B). Similarly, there were no alterations in the abundance of the GLUT4 vesicle cargo protein insulin-responsive aminopeptidase (data not shown).

To investigate if the reduction in Munc18c protein, particularly in skeletal muscle and adipose tissues, would reduce the insulin sensitivity of the Munc18c^−/+ mice, we subjected Munc18c^−/+ and WT mice to an ITT. Mice were fasted for 6 h and fasting glucose levels were determined before insulin for the ITT was injected (Fig. 3). Munc18c^−/+ mice had elevated levels of blood glucose at all time points after the insulin injection compared with WT mice, and calculation of the area under the curve of the ITT data revealed the reduction in insulin responsiveness of the Munc18c^−/+ mice to be statistically different from that of the WT mice (P < 0.05). These data demonstrate that the Munc18c^−/+ mice had significantly impaired insulin sensitivity.

The mechanism for glucose uptake into peripheral tissues involves initiation of the insulin signaling cascade, which leads to the translocation of GLUT4-containing vesicles to the cell surface membranes. We have previously shown that underexpression of syntaxin 4 and Munc18c together in skeletal muscle of syntaxin 4^−/+ KO mice results in inhibition of insulin-stimulated GLUT4 translocation (38), as assessed by skeletal muscle homogenization coupled with sucrose velocity sedimentation to separate sarcolemma/transverse tubule enriched fractions (P1 and P2) from the fractions enriched for intracellular membranes (37). Munc18c^−/+ mice were evaluated using this method in the present study. Under basal conditions, WT and Munc18c heterozygous KO mice showed less GLUT4 protein in the plasma membrane fractions, with the majority being observed in intracellular membrane fractions, as determined by sucrose velocity sedimentation (Fig. 4A). However, under insulin-stimulated conditions, the amount of GLUT4 in the P2 fraction from Munc18c^−/+ mice failed to increase >1.3-fold after intraperitoneal injection with insulin in three independent assays (Fig. 4A). To verify the partitioning of proteins in the skeletal muscle fractions, the enrichment of the plasma membrane protein syntaxin 4 and the endosome-localized transferrin receptor (TnFR) was examined (Fig. 4B). Syntaxin 4 was found to be enriched in the P1 and P2 fractions, with little to no detection in the sucrose fractions, as has been previously described (33,39). TnFR was enriched in early sucrose fractions, distinct from the pattern of the GLUT4-containing vesicles, a finding that is consistent with the original publication of this fractionation method (37). Optical density quantitation of GLUT4 immunoblots confirmed that the skeletal muscle of Munc18c^−/+ KO mice translocated significantly less GLUT4 to the P2 fraction in response to insulin compared with skeletal muscle of WT mice (Fig. 4C). In contrast, proximal insulin signaling was unaffected, as Munc18c^−/+ and WT mice injected with insulin showed equivalently increased levels of Ser 473 phosphorylation and expression of protein kinase B/Akt in skeletal muscle tissue homogenates (Fig. 4D). Taken together, these data demonstrate that the Munc18c^−/+ mice had significantly impaired skeletal muscle insulin-stimulated GLUT4 translocation, independent of effects on proximal insulin-signaling events.

To investigate the effects of the reduced expression of Munc18c on insulin secretion, glucose-stimulated insulin secretion was quantitated from islets isolated from Munc18c^−/+ and WT mice. Glucose stimulation (20 mmol/l) resulted in a 25-fold increase in insulin release compared with unstimulated islets of WT mice, whereas islets isolated from Munc18c^−/+ mice showed <40% of this induction (Fig. 5A). No significant alterations of total insulin content of WT or Munc18c^−/+ islets incubated with or without glucose were detected (Fig. 5B). These data indicate that the islets of the Munc18c^−/+ mice had significantly impaired responsiveness to glucose.

To determine if the reduction in Munc18c protein resulted in aberrant whole-body glucose tolerance, WT and Munc18c^−/+ mice were subjected to an IPGTT. Glucose tolerance was monitored in animals that were fasted overnight and subjected to an intraperitoneal injection of glucose (2 g/kg); blood glucose levels were monitored over a 2-h period (Fig. 6). We also tested glucose tolerance in male Munc18c^−/+ and WT littermate male mice (age 3 months) that had been fed the HF diet to exacerbate potential aberrations in glucose tolerance, as high fat intake negatively impacts mechanisms involved in insulin action and secretion. Three separate feeding trials were conducted over a 9-month period for randomization, and glucose tolerance testing was performed after 5 or 10 weeks on the diet. After 5 weeks on the diet, fasting glucose levels were similar in all four groups of mice, as were peak blood glucose levels 30 min after glucose injection (Fig. 6A). In mice fed the normal diet (ND), peak blood glucose was reached 30 min after injection in the WT mice and was cleared over the remaining 90 min. Similarly, the Munc18c^−/+ mice fed the ND demonstrated the ability to clear glucose to the same level as WT mice throughout the experiment. Both WT and Munc18c^−/+ mice had increased serum levels of insulin within 30 min after glucose injection, indicating responsiveness to the glucose injection (data not shown). However, 2 h after the glucose injection, the glucose levels of the Munc18c^−/+ mice fed the HF diet remained significantly higher than those of WT mice on the same diet, showing essentially no clearance of glucose (Fig. 6A). This glucose intolerance worsened over time as these Munc18c heterozygous mice displayed an even greater level of glucose intolerance after an additional 5 weeks on the HF diet, with significantly elevated blood glucose throughout the clearance period compared with the WT mice on the same diet. Serum insulin levels were not determined for the HF-fed mice due to insufficient quantities of serum that could be collected from the tails of these mice during the IPGTT. These data indicate that the Munc18c^−/+ mice could clear glucose normally when fed a balanced diet, but were unable to compensate as well as the WT mice when challenged with the HF diet.

Daily food intake monitoring showed that mice fed the HF diet consumed slightly less food than those fed the ND, although there were no differences in food consumption between WT and Munc18c^−/+ mice fed the HF diet (Fig. 6B). Furthermore, no significant differences in overall body weight were detected between WT and Munc18c heterozygous mice, although those HF-fed mice weighed significantly more than mice fed the ND after 8 weeks of feeding (Fig. 6C). Tissue weights did not differ between Munc18c heterozygous KO and WT mice fed the ND (Table 1), with liver accounting for ∼5%, fat for 2%, kidney and hindlimb skeletal muscle for 1.5%, and heart, lung, and spleen for <1% of total body weight. Munc18c heterozygous KO mice fed the HF showed parallel changes in tissue weight compared with HF-fed WT mice, with epididymal fat weight significantly increased to 7% of total body weight and other significant increases in the weight of kidneys and skeletal muscle due to infiltration of fat in the tissue of the HF-fed mice. These data indicate that the Munc18c^−/+ mice had an increased susceptibility to HF diet−induced glucose intolerance, whereas food consumption, body weight gain, and tissue weight redistribution did not differ from that of WT mice.

Two of the parameters used to clinically identify the metabolic syndrome are elevated triglyceride levels and impaired fasting glucose. After 10 weeks on the HF diet, the Munc18c^−/+ mice had significantly increased fasting blood glucose compared with mice on the ND (Table 2) as well as increased fed serum insulin levels (P = 0.06). However, fasting levels of serum triglycerides did not differ between WT and Munc18c^−/+ mice under standard conditions, nor did levels of serum cholesterol and nonesterified fatty acids (NEFAs) (Table 3). After 10 weeks on the HF diet, both WT and Munc18c^−/+ mice showed elevated fasting serum cholesterol (174 ± 12) and triglyceride (128 ± 7) levels, with no change in NEFAs. Female mice showed slight increases in NEFA levels, but no differences were seen between WT and Munc18c^−/+ mice (1.5 ± 0.2 for both). Thus the Munc18c^−/+ mice had increased susceptibility for developing hyperglycemia without significant aberrations in serum triglycerides as compared with WT mice.

In this study, we documented the generation and characterization of Munc18c^−/+ mice. Our data clearly show that Munc18c^−/+ mice had an ∼50% loss of Munc18c protein and increased susceptibility toward developing severe glucose intolerance. Consistent with this finding, the Munc18c^−/+ mice showed significantly impaired insulin sensitivity and inherent deficiencies in insulin-stimulated GLUT4 translocation in skeletal muscle, in addition to reduced glucose-stimulated insulin secretion in pancreatic islets. Although capable of seemingly normal glucose disposal when fed the ideal standard diet, the Munc18c^−/+ mice became severely glucose intolerant and incapable of glucose clearance after just 5 weeks on the HF diet, which may suggest that compensatory mechanisms used to maintain glucose homeostasis were no longer sufficient with the added stress of the HF diet. These data indicate that Munc18c is an important positive regulator in vivo in multiple cell types critical for insulin secretion and insulin action.

The loss of insulin-stimulated GLUT4 translocation from hindlimb skeletal muscle of Munc18c^−/+ mice was consistent with the detection of impaired insulin sensitivity as determined by the ITT, although it is unclear as to why the Munc18c^−/+ mice failed to exhibit a detectable defect in glucose tolerance as measured by the IPGTT under standard conditions. Only with the added perturbation of high-fat feeding did the underlying defects in insulin action and insulin secretion become apparent in the IPGTT assay. Possible explanations for the inconsistency might be 1) Munc18c^−/+ mice retain expression of 50% of Munc18c protein, which may be sufficient for function in skeletal muscle depots with different cell/fiber types therein (33); 2) the mice experience compensation from GLUT1-mediated glucose uptake, which is insensitive to Munc18c (32); 3) there was a contribution by adipose tissue glucose uptake, as adipocytes in vivo appear to be less sensitive to alterations in protein levels of Munc18c and syntaxin 4 complexes than skeletal muscle in vivo or 3T3L1 adipocytes in culture (34,38,39); or 4) there was enhanced glucose uptake by the liver, as has been demonstrated in the muscle-specific GLUT4 KO mice (40). Taken together with the fact that mice lacking GLUT4 altogether still show insulin-stimulated glucose uptake into soleus muscle (41), our findings strongly suggest that there may be variations in GLUT4-dependent and even GLUT4-independent activation of muscle glucose uptake that might explain the glucose tolerance of the Munc18c^−/+ mice. Furthermore, alterations in glucose homeostasis clearly detected by hyperinsulinemic-euglycemic clamp analysis have been otherwise missed by the relatively insensitive IPGTT assay (39); future studies of these mice should include this more sensitive analysis as well as ex vivo tissue analyses.

The incongruity of the data showing the defect in glucose-stimulated insulin secretion of the Munc18c^−/+ islets with the otherwise normal levels of circulating insulin detected during the IPGTT may also be an indication of altered hepatic function in the Munc18c^−/+ mice. Because circulating insulin reflects not only insulin secretion but also insulin clearance, these data suggest that the Munc18c^−/+ mice may have altered hepatic insulin clearance. Future studies are required to fully comprehend the contributions of the liver to the overall metabolic profile of the Munc18c^−/+ mice.

It was interesting to observe that the administration of the HF diet uncovered the latent susceptibility of the Munc18c^−/+ mice to severe glucose intolerance. This may have occurred due to inhibition of glucose transport by free fatty acids and/or hyperglycemia (42). It has also been shown that chronic hyperglycemia induces alterations of the insulin-regulated trafficking of Munc18c in 3T3L1 adipocytes (43). In addition, we recently reported that increased glucosamine leads to O-linked glycosylation of Munc18c under basal conditions in 3T3L1 adipocytes (44), and O-linked glycosylation of proteins can induce insulin resistance in skeletal muscle independent of attenuated phosphorylation of Akt and glycogen synthase kinase-3α and -β (45). Thus, a high-fat/high-calorie diet may alter the glycosylation state or cellular locale of existing Munc18c protein and result in inhibition of GLUT4 translocation by disruption of the VAMP2−syntaxin 4 interaction.

The induction of severe glucose intolerance may also be linked to the inherent insufficiencies in glucose-stimulated insulin secretion in the islets of the Munc18c^−/+ mice that was further exaggerated by consumption of the HF diet. In support of this concept, intake of a diet high in fat results in reduced β-cell function, which has been shown to contribute significantly to whole-body glucose intolerance in dogs (46), mice (47), and humans (rev. in 48). Consistent with this, glucose-stimulated insulin secretion from islets isolated from WT and Munc18c^−/+ mice fed the HF diet was reduced as compared with that of mice fed the standard diet (data not shown). The mechanism by which the HF diet exerted effects in the current study may be explained by the inhibition of glucose-stimulated insulin granule exocytosis by lipotoxicity and β-cell apoptosis (49,50). However, our WT and Munc18c^−/+ mice fed the HF diet had insulin contents similar to those of the ND-fed mice. Alternatively, consumption of the HF diet may have induced loss of first-phase insulin secretion, as has been previously described (51). Further experiments will be required to establish the role of Munc18c protein in biphasic insulin secretion.

In sum, these data support a model whereby hyperglycemic and hypertriglyceridemic conditions exacerbate the defect in the fusion machinery of the Munc18c^−/+ mice required to facilitate GLUT4 translocation and insulin granule exocytosis, further weakening the glucose homeostatic mechanisms. This is the first report of the importance of Munc18c protein in insulin secretion, a finding that will now warrant further investigation.

This study was supported by a career development award from the American Diabetes Association (to D.C.T.) and National Institutes of Health Grants DK-067912 (to D.C.T.) and DK-33823 (to J.E.P.).

We are very grateful to Drs. Baoli Yang and Roger Williamson for generating the knockout mice at the University of Iowa, and to Diana Boeglin for expert technical assistance at the University of Iowa. We also thank Angela Nevins for technical assistance with islet isolation. The Vanderbilt Mouse Phenotyping Core Facility was invaluable for their assistance with metabolic measurements of serum samples.

E.O. contributed data for all tables and figures and managed the project. B.A.S. contributed to the collection of data for Figs. 3, 4, 5, and 6A and Tables 2 and 3. D.C.T. isolated the murine Munc18c gene, constructed the targeting vector and contributed to the generation of the Munc18c heterozygous knockout mice at the University of Iowa in the laboratory of J.E.P. Data in all tables and figures except Fig. 1 were collected at Indiana University School of Medicine in the laboratory of D.C.T.