Calpains Play a Role in Insulin Secretion and Action

Pancreatic islets were isolated from nonfasted 9- to 13-week-old C57BL/6J mice (Jackson, Bar Harbor, ME). Soleus muscle strips and adipocytes were isolated from 8- to 12-week-old male Wistar rats (Harlan Sprague-Dawley, Indianapolis, IN). These studies were approved by the Animal Care and Use Committee of the University of Chicago.

Isolation of mouse pancreatic islets was accomplished as described previously (6). After overnight incubation in RPMI-1640 medium (11.6 mmol/l glucose), islets were exposed to inhibitors in the same medium for 4 h at 37°C. The protease inhibitors used were ALLM (calpain inhibitor II, N-Ac-Leu-Leu-methioninal), ALLN (calpain inhibitor I, N-Ac-Leu-Leu-norleucinal), cathepsin B inhibitor II, (Calbiochem-Novabiochem, San Diego, CA), E-64-d (ethyl (+)-(2S,3S)-3-[(S)-3-Methyl-1-(3-methylbutylcarbamoyl)butyl-carbamoyl]-2-oxiranecarboxylate; Matreya,Pleasant Gap, PA), PD147631 (a gift from Dr. Kevin Wang, Parke-Davis, Ann Arbor, MI), and CA074ME (Peptides International, Louisville, KY). GLP-1 (7–36 amide) was from Peninsula Laboratory (Belmont, CA). Other reagents were from Sigma Chemical (St. Louis, MO). The calpain inhibitors were dissolved in DMSO, which also was added to control cultured islets at the concentration of 0.1%. Islets were preincubated in Krebs-Ringer’s buffer (KRB) containing 2 mmol/l glucose and similar concentrations of inhibitors for 60 min at 37°C. Triplicate groups of five islets then were incubated in borosilicate tubes containing 1 ml of KRB with the same concentration of inhibitor and various insulin secretagogues for 1 h in a moving water bath at 37°C. Insulin concentrations were measured in aliquots of the incubation buffer by radioimmunoassay (7).

Insulin secretion from perifused islets was measured using a modification of a previously described protocol (6,7).

[Ca²⁺]_i and NADPH responses to glucose or other secretagogues and calpain activity were measured in islets plated on glass coverslips and cultured in RPMI-1640 for 2–4 days to allow islets to adhere to the coverslips before the 4-h treatment with calpain inhibitors. We measured [Ca²⁺]_i responses to glucose and other secretagogues using fura-2 as described (6). Intracellular NADPH responses to glucose were monitored by the changes in the autofluorescence at 460 nm (8). For measurements of calpain activity, islets were loaded with the fluorogenic, membrane-permeant calpain substrate t-butoxycarbonyl-Leu-Met-7-amino-4-chloromethylcoumarin (Boc-Leu-Met-CMAC [10 μmol/l]; Molecular Probes, Eugene, OR) (9) in HEPES-buffered (10 mmol/l) KRB with 2 mmol/l glucose. The fluorescence emitted from the proteolytic product in islets was measured with a bandpass combination between 400 and 500 nm after excitation by light at 340 nm. Experiments were performed after a 4-h incubation in the presence of 200 μmol/l ALLM, 200 μmol/l E-64-d, or vehicle.

During these measurements, islets on coverslips were perifused continuously with oxygenated KRB medium containing the inhibitors at 37°C. A computer-driven imaging system (MetaFluor 4.0; Universal Imaging, West Chester, PA) was used to capture and store images after ultraviolet excitation.

Glucose utilization and oxidation rates were measured (8,10) in mouse islets cultured in the presence or absence of calpain inhibitors for 4 h as described above.

Isolated islets were dispersed into single β-cells by gentle trituration in Ca²⁺ and Mg²⁺-free phosphate-buffered saline containing 0.025% trypsin-0.1 mmol/l EDTA. Cells were plated on glass coverslips and maintained in culture in RPMI-1640 containing 11.6 mmol/l glucose for 48–96 h before being treated with inhibitors. Calcium current measurements were obtained in the whole-cell patch-clamp configuration. Calcium currents were activated by step depolarizations to either +10 or +20 mV for either 20 or 100 ms, from a holding potential (HP) of −80 mV. All current records were corrected for leak and capacitance. The data were filtered at 2 kHz and then sampled every 100 μs. Pipette resistances were 1.5–2.5 MΩ. Series resistance was partially compensated (∼80%) using the compensation circuit of the Axopatch-1C amplifier.

Cells were incubated for 4 h in RPMI-1640 at 37°C in either 0.1% DMSO (control) or 100 μmol/l ALLM and then transferred for an additional 1–2 h to KRB containing similar concentrations of DMSO or ALLM. For recording, cells were bathed in a solution containing (in mmol/l) 145 NaCl, 2 KCl, 1 MgCl₂, 2 glucose, 10 HEPES, 10 CaCl₂ (pH 7.3; adjusted with NaOH), and either DMSO or ALLM. After the whole-cell configuration was established, the bath solution was exchanged for a tetraethylammonium (TEA)-based recording medium that contained (in mmol/l) 155 TEA-Cl, 2 glucose, 10 HEPES, 10 CaCl₂ and 100 nmol/l TTX (pH 7.3; adjusted with TEA-OH), and either DMSO or ALLM. The intracellular pipette solution consisted of (in mmol/l) 110 CsCl, 4 MgCl₂, 20 HEPES, 10 EGTA, 0.35 guanosine triphosphate, 4 ATP, and 14 creatine phosphate (pH 7.3; adjusted with CsOH).

Membrane capacitance changes in dispersed pancreatic β-cells were used to assess insulin exocytosis and were carried out in the perforated whole-cell configuration. For these experiments, the phase-tracking technique (11) was used, whereby a 60-mV peak-to-peak sine wave was superimposed on a holding potential of −80 mV (12), as described previously (13). β-cells were stimulated with a train of 10-step depolarizations to +20 mV (HP = −80 mV). Each step depolarization lasted 150 ms and was separated by a 400-ms interpulse duration. The data were collected at a 500-μs sampling rate and filtered at 5 kHz. Recordings with series resistance of >20 MΩ were discarded. Series resistance compensation was applied in all recordings.

The extracellular solution for the capacitance measurements contained (in mmol/l) 130 NaCl, 2 glucose, 10 Na-HEPES, 1 MgCl₂, 2 KCl, and 5 CaCl₂ (pH 7.3) with NaOH. The pipette solution contained (in mmol/l) 135 Cs-glutamate, 10 Na-HEPES, 9.5 NaCl, 0.5 TEA-Cl, and 0.5 CaCl₂ (pH 7.3) with CsOH. The pipettes were backfilled with an identical solution to which amphotericin B (final concentration of 0.5 mg/ml) was added and then sonicated. The amphotericin B stock solution (125 mg/ml) was kept frozen at −20°C and used for 1 week. ALLM pretreatment was as described above. All electrophysiological recordings were carried out at room temperature (22–24°C).

Measurement of glycogen synthesis rates was performed using a modification of a previously described protocol (14). Soleus muscle strips were isolated from nonfasted male rats and incubated in KRB–5 mmol/l glucose–10 mmol/l HEPES, 0.2% BSA in the presence and absence of 100 μmol/l ALLM, 200 μmol/l E-64-d, or vehicle.

2-deoxyglucose (2-DOG) uptake by isolated soleus muscle strips from male rats was measured using a modification of a previously described protocol (14). After a 30-min preincubation in KRB containing no glucose, 2 mmol/l pyruvate, 10 mmol/l HEPES, 0.2% BSA, and 100 μmol/l ALLM or 200 μmol/l E-64-d, the muscle strips were transferred to identical medium containing 0.1 mmol/l 2-deoxy-[2,6-³H]glucose (0.5 μCi/ml) and 0.1 mmol/l [¹⁴C]sucrose (0.2 μCi/ml) and incubated for another 30 min at 37°C. The uptake of 2-DOG was calculated as described previously (14).

Adipocytes were isolated from epididymal fat pads (15). For the measurement of basal and insulin-stimulated transport of glucose, aliquots of 200 μl of adipocytes (2 × 10⁵ cells/ml) were incubated for 120 min at 37°C in KRB with different concentrations of insulin either with or without 100 μmol/l ALLM or E-64-d. Then, another 50 μl of KRB containing 5 mmol/l 2-DOG (final concentration 1 mmol/l), 0.5 μCi of 2-deoxy-[2,6-³H]glucose was added and cells were incubated for an additional 5 min at 25°C. The transport was stopped by addition of cytochalasin B (final concentration 50 μmol/l), and cells were spun through 250 μl of dinonyl phthalate oil (Fisher Scientific, Pittsburgh, PA). Cells then were transferred to scintillation vials for counting.

Epitrochlearis muscles were isolated from overnight fasted rats and were incubated for 1 h at 35°C in 3 ml of oxygenated Krebs-Henseleit buffer (KHB) supplemented with 5 mmol/l glucose, 15 mmol/l mannitol, 0.1% BSA, with or without 100 μmol/l ALLM. Muscle strips then were incubated in identical media with or without 12 nmol of insulin or were transferred to KHB medium gassed with 95% N₂–5% CO₂ for 80 min. Muscle glucose transport activity was measured using minor modifications of previously described methods (16). After the initial incubations, all muscles were washed for 15 min at 30°C in 3 ml of KHB containing 20 mmol/l mannitol and 0.1% BSA to remove glucose from the extracellular space. Muscles then were transferred to 1.5 ml of KHB containing 8 mmol/l 3-O-[³H]methyl-d-glucose (3-MG, 2 μCi/ml), 12 mmol/l [U-¹⁴C] mannitol (0.2 μCi/ml), and 0.1% BSA at 30°C for 10 min. When present during the previous incubations, insulin and ALLM also were added to the media for the wash step and transport assay. Extracellular space and intracellular 3-MG concentrations (μmol · ml intracellular water^–1) were determined as described previously (16).

Data are expressed as mean ± SE. The statistical significance of differences was assessed at the 5% level. Differences between two groups were tested using the nonpaired Student’s t test when the data were normally distributed or the Wilcoxon’s rank-sum test when the data were not normally distributed. A paired t test was used when differences between responses in two aliquots of islets from the same mouse incubated under different conditions were being tested. Analysis of variance (ANOVA) was used to test differences among more than two groups.

Exposure to calpain inhibitor II (ALLM, 250 μmol/l) and E-64-d (200 μmol/l) enhanced the insulin secretory responses to 20 mmol/l glucose in islets 1.97 ± 0.3-fold (n = 5, P < 0.01, paired t test) and 1.77 ± 0.1-fold (n = 6, P < 0.001), compared with islets exposed to 20 mmol/l glucose without the inhibitors. No effect was observed on basal insulin secretion (at 2 mmol/l glucose). Significant effects of ALLM and E-64-d on the insulin secretory response to 20 mmol/l glucose were seen at inhibitor concentrations >100 μmol/l (Figs. 1A and B, Table 1). Pretreatment of islets with calpain inhibitor I (ALLN, 100 μmol/l) resulted in a 2.7-fold increase and with the nonpeptide α-mercaptoacrylic acid derivative PD 147631 (a relatively specific nonpeptidic inhibitor of calpain with an IC₅₀ value for calpain that is >100 times less than that for cathepsin B and other cysteine proteases; K. Wang, personal communication) resulted in a 1.8-fold increase in the response to 20 mmol/l glucose compared with control experiments performed in the absence of inhibitors (Table 1). Enhancement of the insulin secretory response to glucose by ALLM was observed only at glucose concentrations >8 mmol/l (Fig. 1C) and also was observed in a dynamic islet perifusion system (Fig. 1D). Islet insulin content was unchanged after a 4-h exposure to 100 μmol/l ALLM in the presence of either 2 or 20 mmol/l (2 mmol/l glucose: 30.9 ± 1.6 vs. 31.7 ± 3.0 ng/islet in ALLM and control, respectively; 20 mmol/l glucose: 28.3 ± 4.7 vs. 32.4 ± 2.0 ng/islet in ALLM- and control-treated islets, respectively).

ALLM produced a 1.55 ± 0.2–fold (n = 6, P < 0.05, paired t test) increase in the insulin secretory response to 50 nmol/l GLP-1, an agent that promotes insulin secretion through its effects on adenyl cyclase and resulting increase in cAMP (17). ALLM did not, however, significantly increase the insulin secretory responses to 30 mmol/l KCl, an agent that directly depolarizes the β-cell or 100 μmol/l carbachol that mobilizes Ca²⁺ from intracellular stores (18) (data not shown).

We also examined the effects of other protease inhibitors on insulin secretion. Insulin secretory responses to 20 mmol/l glucose were not altered in the presence of pepstatin A (100 μmol/l), an aspartyl protease inhibitor, or cathepsin B inhibitor II (100 μmol/l), a lysosomal cysteine protease inhibitor, indicating that the stimulatory effects of ALLM and E-64-d on insulin secretion are specific to these protease inhibitors (Table 1).

Exposure of islets to calpain inhibitors for 4 h did not have any significant effects on glucose utilization and oxidation rates and islet NADPH measurements. Rates of glucose utilization at basal (2 mmol/l glucose) and stimulatory glucose concentrations (20 mmol/l) in the presence of 100 μmol/l ALLM (14.5 ± 3.6 and 89.5 ± 3.0 pmol · islet^–1 · h^–1, respectively) or 200 μmol/l E-64-d (15.5 ± 4 and 79.5 ± 9.5 pmol · islet^–1 · h^–1, respectively) were not significantly different from those in islets incubated in their absence (14.5 ± 2.1 and 76.5 ± 6.5 pmol · islet^–1 · h^–1; n = 3 in each case). Similarly, there was no significant difference in the glucose oxidation rates at basal or stimulatory glucose concentrations in the presence of ALLM (6.0 ± 0.7 and 39.5 ± 4.1 pmol · islet^–1 · h^–1) and E-64-d (5.2 ± 0.4 and 40.0 ± 2.4 pmol · islet^–1 · h^–1) compared with those measured in the absence of inhibitor (4.4 ± 0.8 and 32.5 ± 6.5 pmol · islet^–1 · h^–1; n = 3 in each case). Consistent with a lack of effect of ALLM and E-64-d on β-cell glucose metabolism, the NADPH response to an increase in the glucose concentration from 2 to 14 mmol/l in the presence of 100 μmol/l ALLM (2.7 ± 0.4-fold increase; n = 4) and E-64-d (2.8 ± 0.3-fold increase; n = 2) was not significantly different from controls (2.6 ± 0.2-fold increase; n = 4).

The stimulatory effects of ALLM and E-64-d on insulin secretion in response to high glucose were not associated with increases in [Ca²⁺]_i (Figs. 2A and B). Calcium currents were similar in control and ALLM-treated cells (Fig. 2C); no differences in amplitude or kinetics were apparent. In addition, no shifts in voltage dependence were observed (data not shown), and the average peak calcium current density obtained in control and ALLM-pretreated cells was comparable (Fig. 2D).

Membrane capacitance measurements confirmed the large enhancement of insulin secretion observed after ALLM pretreatment (Figs. 2E and F). Representative capacitance changes, from control (top) and ALLM-pretreated cells (bottom) are shown in Fig. 2E. Stimulation induced much larger average changes in peak membrane capacitance in ALLM-pretreated cells in comparison with control cells (Fig. 2F).

To document that ALLM and E-64-d were inhibiting calpains rather than other cysteine proteases, we measured calpain activity in isolated islets using the fluorogenic calpain-specific substrate Boc-Leu-Met-CMAC (Fig. 3) (9). Although this compound does not distinguish between different calpain isozymes, it does seem to be specific for calpains and not for other lysosomal proteases under physiological conditions (9). In islets incubated in the presence of 200 μmol/l ALLM or 200 μmol/l E-64-d, the rate of generation of the fluorescent signal was lower than in islets incubated in the absence of the calpain inhibitors (Fig. 3A). The area under the curve (AUC), which is a measure of the rate of generation of the fluorescent product, was reduced to 35 ± 4% (n = 3, P < 0.01) and 45 ± 5% (n = 4, P < 0.03, using ANOVA and post hoc Duncan test) of control values in the presence of ALLM (200 μmol/l) and of E-64-d (200 μmol/l), respectively (Fig. 3B).

We measured insulin-stimulated 2-DOG uptake in adipocytes and muscle strips in the presence of 100 μmol/l ALLM and 200 μmol/l E-64-d. In adipocytes, insulin (12 nmol/l) increased uptake of 2-DOG from 457 ± 59 pmol · 2 × 10⁵ cells^–1 · 5 min^–1 to 1,384 ± 178 pmol · 2 × 10⁵ cells^–1 · 5 min^–1 (P < 0.01, n = 4, paired t test; Fig. 4A). In the presence of 100 μmol/l ALLM, insulin increased 2-DOG uptake from 598 ± 102 pmol · 2 × 10⁵ cells^–1 · 5 min^–1 to 751 ± 71 pmol · 2 × 10⁵ cells^–1 · 5 min^–1. This change represented an 84% reduction in the increase in 2-DOG uptake in response to insulin (P < 0.01) compared with experiments performed in the absence of calpain inhibitors. In the presence of E-64-d, 2-DOG uptake was 361 ± 29 and 749 ± 129 pmol · 2 × 10⁵ cells^–1 · 5 min^–1 in the absence and presence of insulin, respectively (n = 4), a 58% reduction in insulin-stimulated 2-DOG uptake (P < 0.03).

Insulin-mediated 2-DOG transport into strips of soleus muscle also was reduced by 100 μmol/l ALLM or 200 μmol/l E-64-d (Fig. 4B). Insulin (12 nmol/l) increased 2-DOG uptake into soleus muscle strips from 0.26 ± 0.01 to 0.47 ± 0.03 μmol · ml H₂O^–1 · 30 min^–1 (P < 0.001, n = 5, paired t test). In the presence of ALLM, 2-DOG uptake was 0.28 ± 0.04 and 0.34 ± 0.05 μmol · ml H₂O^–1 · 30 min^–1 in the absence and presence of insulin, respectively (n = 5), whereas in the presence of E-64-d, 2-DOG uptake was 0.31 ± 0.02 and 0.36 ± 0.02 μmol · ml H₂O^–1 · 30 min^–1 in the absence and presence of insulin, respectively (n = 5). The magnitude of the increase in 2-DOG uptake in response to insulin was reduced by 71 and 75% in the presence of ALLM and E-64-d, respectively (P < 0.001 in each case using ANOVA and post hoc Duncan test). Similar results were observed with the measurement of 3-MG uptake in epitrochlearis muscles (Fig. 4C). Thus, after a 1-h treatment with 100 μmol/l ALLM, the rate of 3-MG uptake in epitrochlearis stimulated by 12 nmol/l insulin was reduced by 61% (0.44 ± 0.06 vs. 1.13 ± 0.07 μmol · ml^–1 · 10 min^–1, for n = 5 muscles/group; P < 0.01).

To determine whether the observed effects were specific for calpain inhibitors, we used the cathepsin inhibitor compound CA074ME (a cell permeable peptide inhibitor of cathepsin B [19]) in similar experiments. Maximal insulin-stimulated glucose uptake (12 nmol/l insulin) in epitrochlearis muscle strips was 1.1 ± 0.14 μmol · ml H₂O^–1 · 10 min^–1 (n = 8) in the absence and 1.1 ± 0.14 μmol · ml H₂O^–1 · 10 min^–1 (n = 6) in the presence of 1 μmol/l CA074ME.

ALLM also inhibited hypoxia-induced 3-MG uptake in skeletal muscle (Fig. 4D). An 80-min incubation of epitrochlearis muscle in hypoxic conditions caused an increase in 3-MG uptake in the absence of ALLM (0.26 ± 0.05 vs. 1.21 ± 0.2 μmol · ml^–1 · 10 min^–1), and a 1-h preexposure to 100 μmol/l ALLM suppressed the hypoxia-induced 3-MG uptake by 58% (0.50 ± 0.09 vs. 1.21 ± 0.2 μmol · ml^–1 · 10 min^–1 for n = 4 muscles/group; P < 0.05). The basal 3-MG uptake rate also was significantly inhibited by ALLM (0.26 ± 0.05 vs. 0.07 ± 0.02 μmol · ml^–1 · 10 min^–1 for n = 4 muscles/group; P < 0.05).

Insulin (6 nmol/l) increased the rate of muscle glycogen synthesis in soleus muscle strips from 0.58 ± 0.08 to 1.55 ± 0.20 nmol glucose · mg^–1 · h^–1 (n = 6, P < 0.005, using paired t test; Fig. 5). In the presence of 100 μmol/l ALLM, the glycogen synthesis rates were 0.27 ± 0.03 and 0.40 ± 0.05 nmol glucose · mg^–1 · h^–1 in the absence and presence of insulin, respectively (n = 6), whereas in the presence of 200 μmol/l E-64-d, the glycogen synthesis rates were 0.49 ± 0.08 and 0.80 ± 0.14 nmol glucose · mg^–1 · h^–1 in the absence and presence of insulin, respectively (n = 6). The magnitude of the increase in glycogen synthesis in response to insulin was 87 and 68% lower in the presence of ALLM and E-64-d, respectively (P < 0.01 in each case using ANOVA and post hoc Duncan test).

Genetic studies have shown that variation in the gene encoding calpain-10 is associated with type 2 diabetes (2). Low levels of calpain-10 mRNA in skeletal muscle also are associated with insulin resistance, suggesting that calpain-10 may have a role in determining the normal response of skeletal muscle to the effects of insulin (3). The presence of calpain-10 mRNA in pancreatic islets, muscle, and liver, the three most important tissues that control blood glucose levels, suggests that calpain-10 may regulate pathways that affect insulin secretion, insulin action, and hepatic glucose production, each of which is altered in patients with type 2 diabetes. As a first step in identifying the biochemical pathways that may be regulated by calpain-10 and/or other calpains, we studied the effects of calpain inhibitors on insulin secretory responses to glucose in islets and insulin action in muscle and adipocytes.

The present studies provide some insight into the mechanism by which ALLM and E-64-d alter the insulin secretory responses to glucose. After a 4-h exposure to calpain inhibitors, the increase in glucose-induced insulin secretion was not associated with increased [Ca²⁺]_i, rates of glucose oxidation and utilization, or NADPH generation. Thus, pathways in the β-cell that are responsible for the uptake and metabolism of glucose were unaffected. We believe that the primary site(s) of calpain action in the regulation of insulin secretion may be in the pathways that regulate the movement or fusion of insulin secretory granules with the plasma membrane. This hypothesis is supported by the observation that membrane capacitance is increased in islets after short-term exposure to ALLM, in the absence of effects on glucose metabolism and mitochondrial function.

Glucose transport is the major rate-limiting step in the metabolism of glucose in muscle and fat tissue, and impaired glucose uptake by theses tissues is an important feature of insulin resistance (20). Treatment of muscle strips with ALLM and E-64-d resulted in a significant reduction in the rate of insulin-stimulated glucose uptake, as measured by two different assays (2-DOG and 3-MG uptake) and the incorporation of glucose into glycogen. The two calpain inhibitors also inhibited 2-DOG uptake, into adipocytes. Although the exact site where calpains act in the signaling pathways of insulin action needs further study, the results of the effects of calpain inhibitors on hypoxia-induced 3-MG uptake may shed light on this issue. The hypoxia-associated increase in glucose uptake is not mediated by the insulin receptor-IRS-phosphatidylinositol 3-kinase system (21,22). The inhibition of hypoxia-stimulated glucose uptake by ALLM suggests that the calpain-regulated step is distal to this system.

The specific calpain isozyme(s) or cysteine protease(s) implicated in the control of insulin secretion and insulin action in the studies described above is unknown. Because isozyme-specific calpain inhibitors are not available, it is not possible to alter selectively the activity of calpain-10 using inhibitors. ALLM and E-64-d are not specific inhibitors of calpains; they also inhibit cathepsins, and for that reason we cannot be certain that the effects that we are attributing to calpain inhibition are due to inhibition of these proteases. However, a number of factors point strongly to calpain inhibition as the mechanism underlying the effects observed. First, the inhibition of hydrolysis of the calpain-specific substrate Boc-Leu-Met-CMAC by ALLM and E-64-d in pancreatic islets is consistent with the hypothesis that ALLM and E-64-d affect insulin secretion by inhibiting calpain activity rather than lysosomal cysteine proteases such as the cathepsins. Second, different calpain inhibitors that work by different mechanism, such as ALLN and the specific calpain inhibitor PD 147631, have similar effects on glucose-induced insulin secretion as ALLM and E-64-d. Third, cathepsin inhibitors and the aspartyl protease inhibitor pepstatin A showed no effect on insulin secretion. The identification of the specific calpain(s) involved must await the development of specific inhibitors or animal models of either increased or decreased expression of various forms of calpains in different tissues, including the pancreatic β-cell and muscle and adipocyte tissue, to determine which calpains are involved in the regulation of insulin action and insulin secretion.

The possibility that the enhancement in insulin secretion, observed in islets exposed to calpain inhibitors, results from nonspecific toxic effects of the calpain inhibitors on the islets was considered. We do not believe that this is the mechanism for the following reasons. First, the effects are seen with calpain inhibitors of different structures that act by different mechanisms. Second, toxic effects on islets would be expected to result in nonspecific release of insulin from islets, but both basal insulin secretion and islet insulin content were unaffected by exposure to calpain inhibitors.

The present studies demonstrate enhancement of glucose-induced insulin secretion in pancreatic islets exposed to calpain inhibitors for 4 h. The mechanism(s) responsible for this change seems to include effects on the insulin secretory machinery responsible for movement of insulin secretory granules to and fusion with the β-cell membrane. Calpain inhibition also results in reduced insulin-stimulated glucose uptake into adipocytes and muscle and reduced glycogen synthesis rates in muscle. These latter results are consistent with physiological studies in nondiabetic subjects in which low levels of calpain-10 mRNA were associated with insulin resistance (3). Taken together, the in vitro studies presented here and in vivo studies in human subjects (3) support the hypothesis that calpains are involved in the regulation of insulin secretion and insulin action.

This work was supported by grants from the U.S. Public Health Service (DK-31842, DK-20595, DK-44840, DK-47486, and DK-55889), a gift from the Blum-Kovler Foundation, and by the Howard Hughes Medical Institute.

S.K.S., Y.-P.Z., and K.O. contributed equally to the conduct of this study.

We thank K. Fox for expert technical assistance.