|Home | About | Journals | Submit | Contact Us | Français|
Islet Ca2+-independent phospholipase A2 (iPLA2) is postulated to mediate insulin secretion by releasing arachidonic acid in response to insulin secretagogues. However, the significance of iPLA2 signaling in insulin secretion in vivo remains unexplored. Here we investigated the physiological role of iPLA2 in β-cell lines, isolated islets, and mice. We showed that small interfering RNA-specific silencing of iPLA2 expression in INS-1 cells significantly reduced insulin-secretory responses of INS-1 cells to glucose. Immunohistochemical analysis revealed that mouse islet cells expressed significantly higher levels of iPLA2 than pancreatic exocrine acinar cells. Bromoenol lactone (BEL), a selective inhibitor of iPLA2, inhibited glucose-stimulated insulin secretion from isolated mouse islets; this inhibition was overcome by exogenous arachidonic acid. We also showed that iv BEL administration to mice resulted in sustained hyperglycemia and reduced insulin levels during glucose tolerance tests. Clamp experiments demonstrated that the impaired glucose tolerance was due to insufficient insulin secretion rather than decreased insulin sensitivity. Short-term administration of BEL to mice had no effect on fasting glucose levels and caused no apparent pathological changes of islets in pancreas sections. These results unambiguously demonstrate that iPLA2 signaling plays an important role in glucose-stimulated insulin secretion under physiological conditions.
Phospholipase A2 (PLA2) is a diverse group of enzymes that catalyze the hydrolysis of the sn-2 fatty acyl bond of phospholipids to liberate free fatty acids and lysophospholipids (1). Growing evidence suggests that hydrolysis of arachidonic acid (AA) from the membrane phospholipids of β-cells by activation of intracellular PLA2 is involved in glucose-stimulated insulin secretion (2–5). Certain receptor-mediated insulin secretagogues, such as glucose-dependent insulinotropic polypeptide, cholecystokinin-8, and muscarinic agonist stimulate the release of AA by activation of iPLA2 (6–8). Both Ca2+-dependent and -independent PLA2 (cPLA2 and iPLA2, respectively) are reportedly involved in AA release and insulin secretion induced by insulin secretagogues (9–25). cPLA2, which requires micromolar Ca2+ for its enzymatic activity, was cloned from rat islet cells (26) and has been shown to play an important role in the maintenance of insulin stores (25, 27) and in the control of the rate of exocytosis in β-cells (28).
It is noteworthy that the majority of AA released in response to glucose stimulation is Ca2+ influx independent: AA accumulation occurs in Ca2+-free medium or in the presence of Ca2+-channel blockers (20, 29, 30). These findings suggest that glucose-induced AA release occurs before Ca2+ influx and, therefore, before insulin secretion. Indeed, AA has been found to stimulate a rise of cytosolic Ca2+ by multiple mechanisms (29, 31–33). Interestingly, like glucose-stimulated insulin secretion, glucose-stimulated AA release requires the transport and metabolization of glucose within β-cells, and ATP, thought to be a signal generated by glucose metabolism (34), can stimulate AA release in the absence of Ca2+ (20). An ATP-stimulated, Ca2+-independent PLA2 activity has also been identified in isolated islet and insulinoma cell lines (21). iPLA2 is an attractive candidate for mediating the Ca2+-independent, glucose metabolism-activated hydrolysis of AA from islet membrane phospholipids because bromoenol lactone (BEL), which selectively inhibits iPLA2 (35–37) while having no effect on β-cell metabolism (22), suppresses glucose-stimulated AA release, the rise of intracellular Ca2+, and insulin secretion (21, 22, 24, 38–40).
We have cloned an iPLA2 from rat and human pancreatic islets and confirmed that its catalytic activity is Ca2+ independent, stimulated by ATP, and inhibited by BEL (41, 42). We further examined the role of iPLA2 in mediating glucose-stimulated AA release and insulin secretion by overexpressing cloned iPLA2 cDNA in insulinoma INS-1 cells. Compared with controls, these cells exhibited increased AA release and insulin secretion in response to glucose stimulation; these qualities were further enhanced by elevation of cAMP and inhibited by BEL (24, 43).
Although experimental evidence from INS-1 cells with elevated iPLA2 expression supports the role of iPLA2 in glucose-stimulated insulin secretion, the physiological importance of iPLA2 in regulation of in vivo insulin levels remains unexplored. In this study, we further examined the function of iPLA2 in insulin secretion in INS-1 cells, isolated islets, and mice by using small interfering RNA (siRNA) silencing and BEL administration to block the expression or inhibit the activity of iPLA2. The results presented here demonstrate that inhibition of iPLA2 results in insufficient insulin secretion and hyperglycemia in normal mice.
It has previously been shown that iPLA2-overexpressing INS-1 cells secrete higher levels of insulin than parental INS-1 cells in response to glucose stimulation (24, 43, 44). To further verify the role of iPLA2 in insulin secretion, we applied RNA interference technology (45) to specifically silence iPLA2 expression in INS-1 cells. iPLA2-specific siRNA was constructed and evaluated as described in Materials and Methods. The psiRNA-iPLA2 construct that showed the highest degree of inhibition was used in all subsequent studies. As shown in Fig. 1B, about 50% of cells in the transfected plate, corresponding to the transfection efficiency of psiRNA-iPLA2, lost any detectable expression of iPLA2. In contrast, mock-transfected cells retained uniform iPLA2 expression (Fig. 1A). Western blot analysis confirmed that iPLA2 expression in siRNA-transfected cells was dramatically reduced compared with mock-transfected cells or to cells transfected with glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-specific siRNA (Fig. 1C). The fact that some iPLA2 was still detectable in the Western blot analysis is likely due to the incomplete transfection of the cells with psiRNA-iPLA2 rather than to incomplete silencing of iPLA2 by the siRNA. For each subsequent batch of experiments, we confirmed iPLA2 expression silencing by measuring its activity before measuring the insulin-secretory response of the transfected cells. As shown in Fig. 1D, the glucose response was dramatically inhibited in siRNA-treated cells compared with control and mock-transfected cells. These data, along with our previous results showing a significant increase in insulin secretion when iPLA2 is overexpressed in INS-1 cells (24), demonstrate that iPLA2 signaling is important for glucose-stimulated insulin secretion.
To assess the role of iPLA2 in insulin secretion in vivo, we first used immunohistochemical analysis to examine iPLA2 expression patterns in mouse pancreas, a gland with both endocrine (islet of Langerhans) and exocrine (acinar cells) functions. We removed the pancreas from C57BL/6J mice and stained three consecutive islet-containing sections for insulin, glucagon, and iPLA2, respectively (Fig. 2). Consistent with other reports, insulin, expressed by islet β-cells, was detectable in the core of the islet (Fig. 2A), and glucagon, expressed by islet α-cells, was found on the periphery (Fig. 2B). Interestingly, iPLA2 is expressed predominantly in islet cells costained by insulin but is barely detected in the exocrine acinar cells (Fig. 2C), suggesting that iPLA2 is more important in the endocrine tissue of the pancreas than in the exocrine cells. Consistent with these differences in expression levels, iPLA2 activity was higher in islets than in pancreatic acinar cells (Fig. 2D). Pretreatment of isolated islets with BEL (10 μM) reduced iPLA2 activity to close to the level of acinar cells.
We next examined whether inhibition of this islet-associated iPLA2 activity by BEL also suppresses glucose-stimulated insulin secretion in isolated mouse islets. Freshly isolated, randomly grouped islets from normal mice were used to assess glucose-stimulated insulin secretion by static incubation in the presence or absence of 10 μM BEL. As shown in Fig. 3, the islets secreted a large amount of insulin in response to 20 mM glucose in the absence of BEL. However, this secretory response was dramatically reduced in islets pretreated with BEL, consistent with previous reports on isolated rat islets (46).
BEL has recently been reported to inhibit magnesium-dependent cytosolic phosphatidate phosphohydrolase (PAP-1) in mouse P388D1 macrophages and in human amnionic WISH cells (47, 48). In isolated rat islets, 10 μM BEL inhibits about 23% of total PAP-1 activity (49). To differentiate the effect of BEL on iPLA2 from its effect on PAP-1, we treated the islets with propranolol, a well-known PAP-1-specific inhibitor (50–53). In accordance with other studies (52, 53), 250 μM propranolol completely inhibited PAP-1 activity in the islets. However, in contrast to BEL, propranolol administration caused glucose-stimulated insulin secretion in the islets to increase slightly (Fig. 3A). Taken together with our previous results showing that propranolol does not affect glucose-stimulated AA release (46) and insulin secretion (24), our findings indicate that the suppression of insulin secretion by BEL is not attributable to inhibition of PAP-1.
Inhibition of iPLA2 by BEL and suppression of AA release is thought to lead to the inhibition of glucose-stimulated insulin secretion (21, 22, 24, 38–40). We further examined whether exogenous AA can reverse the inhibitory effect of BEL on glucose-stimulated insulin secretion. Consistent with previous reports (19, 23), addition of exogenous AA significantly increased insulin secretion only in mouse islets incubated at substimulatory, but not stimulatory, glucose concentrations (Fig. 3B). Importantly, exogenous AA also reversed the inhibitory effect of BEL on glucose-stimulated insulin secretion. These data, along with our previous results showing that overexpressing iPLA2 in INS-1 cells significantly increases glucose-stimulated AA release (24), suggest that iPLA2 serves as part of a glucose-sensing pathway that rapidly generates AA to stimulate insulin exocytosis.
Because inhibition of iPLA2 activity by BEL effectively inhibited glucose-stimulated insulin secretion in mouse islets, we were interested in determining the physiological effect of BEL injection, and subsequent iPLA2 inhibition, in mice. Forty minutes before glucose tolerance testing, BEL (3 μg/g body wt) was injected into the tail vein of mice as described in Materials and Methods. As shown in Fig. 4A, fasting glucose levels in the BEL-treated mice were not significantly different from those in the vehicle-treated group. However, blood glucose levels in the BEL-treated mice were significantly higher than in the control group after ip glucose injection, and marked hyperglycemia persisted longer in the BEL-treated mice than in control mice. We did not observe any apparent difference between the sexes, and the data shown here were gathered from both male and female animals. As with the isolated islets, we also treated mice with propranolol to distinguish between iPLA2 and PAP-1 inhibition. Glucose levels in the propranolol-treated group exhibited no apparent difference from those of the control group (Fig. 4A), excluding the possibility that the hyperglycemic state observed in BEL-treated mice was attributable to PAP-1 inhibition.
We also tested insulin levels in blood samples drawn from control and BEL-treated groups 10 min after glucose challenge. As shown in Fig. 4B, the insulin levels of the BEL-treated mice before glucose challenge were similar to those in the control group. After glucose challenge, however, insulin levels were significantly lower. These results demonstrate that BEL administration caused insufficient insulin secretion in the mice, which may contribute to impaired glucose tolerance. To confirm that BEL is capable of inhibiting iPLA2 in vivo, we isolated islets from the mice immediately after administration of BEL or vehicle and measured their iPLA2 activity. As shown in Fig. 4C, iPLA2 activity was significantly lower in islets isolated from the BEL-treated group than from the vehicle-treated group, even after the 4 h required for complete islet isolation. These data indicate that BEL administration to mice can inhibit iPLA2 activity in islets throughout the glucose tolerance testing period.
Next we examined whether BEL administration to mice affects fasting glucose levels or causes pathological changes to pancreatic islets. As shown in Fig. 5A, there was no apparent difference in fasting glucose levels between control and BEL-treated groups after 5 successive days of BEL administration. Additionally, as shown in Fig. 5B, immunohistochemical analysis of pancreatic islets revealed no morphological changes in islets from BEL-treated mice as compared with control mice, and expression of insulin, glucagon, and iPLA2 in the two groups was similar. Insulin content and iPLA2 activity were also not significantly different between control and BEL-treated mice (data not shown). These results indicate the architecture and protein expression profiles of pancreatic islet β-cells were not affected by short-term administration of BEL to mice.
The impaired glucose tolerance we observed in the BEL-treated group could be due either to reduced insulin sensitivity or insufficient insulin secretion. To resolve this question, we measured glucose disposal in the mice using a euglycemic hyperinsuline-mic clamp (54). Age- and weight-matched catheterized mice were iv injected with either vehicle or BEL (3 μg/g body wt), and blood glucose levels were clamped at 100 mg/dl in both groups (Fig. 6A). The insulin levels between groups during the clamp were not significantly different. As shown in Fig. 6B, the glucose infusion rate in the BEL-treated group is similar to that of the controls during clamps. We averaged the glucose infusion rate during the last 60 min to determine the glucose disposal rate as described previously (55). There was no significant difference in glucose disposal rates between the two groups (Fig. 6C). This finding indicates that insulin sensitivity under hyperinsulinemic conditions was not affected by BEL treatment.
Because BEL treatment does not significantly affect the glucose disposal rate in mice, we used hyperglycemic clamp experiments (56) to evaluate β-cell insulin secretion. Blood glucose levels were clamped at 3 times the normal range in both control and BEL-treated mice (Fig. 7A), and the glucose infusion rate was adjusted according to ongoing measurements of blood glucose levels (Fig. 7B). Although we began the experiment with the same infusion rate for both groups, glucose levels rose faster in the BEL-treated group than in the controls, and the mice in the control group required larger infusions of glucose within the first 30 min to elevate blood glucose to designated levels (Fig. 7B). As the experiment continued, the BEL-treated mice required less glucose to maintain the hyperglycemic state than did the controls. We then assessed insulin levels in blood samples from animals from both groups that had reached a steady-hyperglycemic state and found that average insulin levels in the BEL-treated group were significantly lower than in controls (Fig. 7C). In contrast, no differences were observed under basal conditions, consistent with the results of the glucose tolerance test (Fig. 4B). When these data are taken together, we can conclude that the hyperglycemia observed in the BEL-treated mice (Fig. 4A) results from insufficient insulin secretion rather than from decreased insulin sensitivity.
It has long been proposed that glucose-stimulated hydrolysis of AA from islet membrane phospholipids by activation of PLA2 plays a signaling role in glucose-stimulated insulin secretion (4, 19, 57–59). However, a lack of in vivo studies has made it difficult to determine whether this hypothesis accurately represents the physiological process of insulin secretion. In the present study, we investigated the physiological role of iPLA2 in glucose-stimulated insulin secretion in β-cell lines, isolated islets, and C57BL/6J mice. The following results provide the first experimental evidence that iPLA2 signaling is important for sufficient insulin secretion in response to high glucose in vivo: 1) siRNA that specifically silences iPLA2 expression in cultured INS-1 cells significantly suppresses glucose-stimulated insulin secretion in those cells; 2) pancreatic islet β-cells express substantially more iPLA2 than do pancreatic exocrine acinar cells; 3) inhibition of iPLA2 catalytic activity by BEL in isolated mouse islets dramatically suppresses glucose-stimulated insulin secretion, and this inhibitory effect is reversed by exogenous AA; 4) inhibition of iPLA2 by administering BEL to mice causes sustained hyperglycemia and reduces insulin levels; and 5) BEL treatment of mice causes insufficient insulin secretion rather than alteration of insulin sensitivity.
If iPLA2 signaling is indeed physiologically important for insulin secretion, as we demonstrated in mice, the following questions should be considered: 1) Do islet cells contain a large pool of AA in their membrane phospholipids? Previous studies showed that AA was the most abundant fatty acid measured in the phospholipids of rodent and human islets and constituted more than 30% of the total esterified fatty acyl mass of islet glycerolipids (22, 38). It is also apparent that abundant AA in islet phospholipids is essential for insulin secretion by β-cells: reduced insulin secretion by late-passage INS-1 cells is associated with the loss of phospholipid-associated AA, and AA supplementation partially restores responses to insulin secretagogues in these cells (60). 2) Is the amount of unesterified AA released after glucose stimulation sufficient to stimulate insulin secretion under physiological conditions? It has been shown that intracellular concentrations of free AA can achieve levels of 100 μM or more during glucose stimulation (29, 30), well above the levels required to stimulate insulin secretion in in vitro studies. In these studies, 50 μM or greater of exogenous AA stimulates significant insulin secretion from intact or permeabilized islets even at substimulatory concentrations of glucose (19, 27, 57–59). The pattern of AA-induced insulin stimulation is also similar to the biphasic response to glucose (58). 3) Are islet β-cells capable of releasing AA quickly enough to stimulate insulin secretion? Exposure of islets to glucose causes a rapid release of AA (4), suggesting β-cells have an immediate response system. Because adjusting the amount of key enzyme through synthesis, degradation, and protein transport is too time consuming, most cells instead turn to ways of regulating the activity of enzymes already on hand when rapid responses to changing physiological conditions are necessary. We have shown here that islet cells express a substantial amount of iPLA2 (Fig. 2). Activation of this existing large pool upon glucose stimulation may contribute to the rapid release of AA. Consistent with this hypothesis, glucose-stimulated AA release in isolated islets was fully inhibited by BEL but not propranolol (46), iPLA2-overexpressing INS-1 cells release more AA than do parental INS-1 cells in response to stimulatory glucose concentrations, and this response can be inhibited by BEL (24).
Structural features of iPLA2 also support a physiologically important role in mediating glucose-stimulated AA release. iPLA2 contains an ATP-binding domain that is highly homologous to that of protein kinases (5, 61). The catalytic activity of islet iPLA2 is stimulated by ATP in vitro (41, 42), and iPLA2 protein binds to an ATP-sepharose column (62). The fact that glucose stimulates more AA release in iPLA2-overexpressing INS-1 cells than in parental INS-1 cells (5, 24) indicates that glucose can activate iPLA2, probably through its metabolite ATP. In addition, glucose stimulates iPLA2 translocation from the cytosol to the perinucleus, where it can access its membrane-phospholipid substrates. This translocation is further enhanced by elevation of cellular cAMP levels by 3-isobutyl-1-methylxanthine or forskolin, which is also accompanied by increased insulin secretion by INS-1 cells (24, 44). Taken together, we conclude that iPLA2 plays a key role in glucose-stimulated AA release and that the iPLA2/AA signaling pathway represents a distinct, physiologically relevant mechanism to mediate glucose-stimulated insulin secretion in vivo.
Tissue culture medium RPMI 1640 and fetal bovine serum (FBS) were purchased from Mediatech, Inc. (Herndon, VA). CMRL-1066, Hanks’ balanced salt solution, and penicillin/streptomycin were from Invitrogen (San Diego, CA). BSA, propranolol, Ficoll, and glucose solution were obtained from Sigma-Aldrich Co. (St. Louis, MO). BEL was from Cayman Chemical Co. (Ann Arbor, MI). Ultrasensitive insulin ELISA kit was from Crystal Chem, Inc. (Chicago, IL). Liberase RI was from Roche Molecular Biochemicals (Indianapolis, IN). FuGENE 6 transfection reagent was obtained from Roche. Radioimmunoprecipitation assay (RIPA) lysis buffer was from United States Biological (Swampscott, MA), and enhanced chemiluminescence Western blotting detection reagents were from Amersham Pharmacia Biotech (Piscataway, NJ). Human insulin was from Eli Lilly & Co. (Indianapolis, IN).
pSilencer 2.0-U6 and pSilencer 3.0-H1 Vectors (Ambion, Inc., Austin, TX) were used to express iPLA2-specific siRNAs in INS-1 cells. pSilencer 2.0-U6 contains the human U6 promoter that has been used extensively to express siRNAs and ribozymes in mammalian cells, and pSilencer 3.0-H1 contains the H1 RNA promoter (H1 RNA is a component of ribonuclease P). The siRNA expression vectors were constructed according to the manufacturer’s instructions. Four pairs of siRNAs against rat iPLA2 mRNA were synthesized, annealed, and subcloned into both pSilencer 2.0-U6 and pSilencer 3.0-H1 Vectors. After verification and purification, all constructs were transfected into INS-1 cells to evaluate the inhibition of iPLA2 expression.
INS-1 cells were cultured in RPMI 1640 medium containing 10% FBS, 10 mM HEPES buffer, 2 mM glutamine, 1 mM sodium pyruvate, 50 mM β-mercaptoethanol, and 0.1% (wt/vol) of penicillin in a 5% CO2 in air atmosphere at 37 C. When the cells were approximately 70% confluent, they were transfected with FuGENE 6 transfection reagent according to the instruction manual. The cells were harvested and lysed, 3 d after transfection, in RIPA lysis buffer containing protease inhibitors. The supernatant from each transfectant was analyzed by the Western blot to determine whether iPLA2 expression was inhibited by the siRNA. The pSilencer 3.0-H1 construct corresponding to the sequence 5′-CTTGGCATTCGGGAGTGTT-3′ in the encoding region of iPLA2 mRNA showed the most inhibition in INS-1 cells. This vector was designated as psiRNA-iPLA2 and used to express siRNA in INS-1 cells in all of the experiments. GAPDH-specific siRNA was used as a negative control.
Cells were collected and lysed in RIPA buffer containing protease inhibitors. Aliquots containing 60 μg of protein were resolved on 8% SDS-PAGE. After electrophoresis, proteins were blotted onto nitrocellulose membranes, and iPLA2 was probed with goat antihuman iPLA2 polyclonal IgG (T-14, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), followed by incubation with antigoat IgG conjugated with horseradish peroxidase. iPLA2 proteins were detected by enhanced chemiluminescence Western blotting detection reagents.
Isolated mouse pancreases were fixed overnight in 10% formalin at 4 C. Sections (4-μm) were cut using ThermoShandon Finesse Microtome (Thermo Shandon, Inc., Pittsburgh, PA). Slides were incubated in 4% H2O2 to block the endogenous peroxidase, and then Powerblock universal blocking reagent (10X, BioGenex Laboratories, Inc., San Ramon CA) was used to eliminate the background. Insulin, glucagon, and iPLA2 were stained with the rabbit antiinsulin antibodies (Santa Cruz Biotechnology), rabbit antiglucagon antibodies (DAKO Corp., Carpinteria, CA), or goat anti-iPLA2 antibodies (Santa Cruz Biotechnology), respectively. Biotinylated antirabbit IgG (BioGenex) or biotinylated rabbit antigoat IgG (Vector Laboratories, Inc., Burlingame, CA) were used as second antibody as appropriate, followed by a peroxidase-conjugated streptavidin (BioGenex) and DAKO liquid 3,3′-diaminobenzidine substrate-chromogen reagent (DAKO). Slides were then counterstained with hematoxylin to localize nuclei of cells. Finally, the slides were mounted with Cytoseal xylene-based mounting medium (Stephens Scientific, Kalamazoo, MI) for visualization. For immunocytochemical analysis, INS-1 cells were spun on a Silane-prep slide (Sigma Diagnostics, St. Louis, MO) using ThermoShandon Cytospin-4 (Thermo Shandon, Inc., Pittsburgh, PA) and stained with anti-iPLA2 antibodies. Immunoreactivity was detected as described above.
The activity of iPLA2 was determined using a modified kit originally designed for cPLA2 (cPLA2 assay kit, Cayman Chemical Co.) as described previously (63, 64). Briefly, after specific treatments, tissues from mice or cultured cells are collected and homogenized in buffer (50 mM HEPES, pH 7.4; 1 mM EDTA) followed by centrifugation at 14,000 × g for 20 min at 4 C. The supernatant was removed, and the concentration of proteins was determined. iPLA2 activity was assayed by incubating the samples with arachidonoyl thio-PC (phosphorylcholine) for 1 h at 25 C in a Ca2+-free buffer (4 mM EGTA; 160 mM HEPES, pH 7.4; 300 mM NaCl; 8 mM Triton X-100; 60% glycerol; 2 mg/ml BSA). The reaction was stopped by addition of 5,5′-dithio-bis (nitrobenzoic acid) for 5 min, and the absorbance was determined at 414 nm using a μQuant microplate reader (Bio-Tek Instruments, Inc., Winooski, VT). The specific activity of iPLA2 was calculated and expressed in nanomoles/min/mg of total proteins. The background basal lipase activity, which was determined by inhibiting all specific iPLA2 activity in control samples with BEL, was subtracted from all readings as described elsewhere (63, 64).
For determination of insulin secretion by islets, isolated mouse islets were cultured overnight in RPMI-1640 medium containing 11 mM glucose, 2 mM L-glutamine, 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. The following day, islets were transferred to CMRL-1066 medium containing 5.5 mM glucose and preincubated for 120 min at 37 C. The islets were then distributed in batches of five islets per well in 0.5 ml of Krebs-Ringer bicarbonate buffer (KRB) with or without 10 μM BEL for 30 min, and subsequently transferred under dissecting microscope to medium containing 3 mM glucose or 20 mM glucose or 100 μM AA incubating for 60 min. The medium was then removed from each well, and insulin levels were measured with an ultrasensitive insulin ELISA kit. Insulin secretion was also measured in INS-1 cells with and without transfected psiRNA-iPLA2. Cells in six-well plates were washed in KRB buffer and incubated in the KRB buffer containing 3 mM glucose for 60 min. The medium was then removed and replaced with KRB buffer containing either 3 mM or 20 mM glucose at 37 C for 60 min. The KRB buffer in each well was collected for insulin measurement.
Islets were isolated from C57BL/6J mice by liberase digestion, discontinuous Ficoll gradient separation, and hand pickup, a modification of the original method of Lacy and Kostianovsky (65). Complete islet isolation from mice takes about 4–5 h. Islets were cultured in an atmosphere of 95% air, 5% CO2 in CMRL-1066 medium containing 5.5 mM glucose, 2 mM L-glutamine, 10% (vol/vol) heat-inactivated FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin.
Mice (8- to 12-wk old) were fasted overnight (16–18 h) and randomly assigned to receive BEL, vehicle (ethanol), or propranolol via tail vein injection. BEL (10 mM) was freshly prepared in 100% ethanol just before the injection. Glucose tolerance tests (ip) were carried out 40 min after injection. Blood was drawn from the tail vein before, and 5, 10, 15, 30, 45, 60, and 120 min after glucose administration (3 mg/g body wt) to the peritoneal cavity. Blood glucose levels were measured by a Glucometer Elite (Bayer Corp., Elkhart, IN) and Glucose Analyzer (Analox GM7 Analyser, Anolox Instruments USA, Inc. Lunenberg, MA). We determined the optimal amount of BEL by comparing the effect of 1, 3, and 5 μg/g body wt. Although glucose tolerance was affected by every level of BEL, 3 and 5 μg/g gave more consistent results. We used 3 μg BEL/g body wt in all experiments: approximately the same concentration (10 μM = 3.17 μg/ml) that inhibits iPLA2 activity and insulin secretion in islets (21, 22, 38–40). The amount of propranolol (5 μg/g body wt) was optimized based on the daily dosage (160 mg/83 kg body wt) used in humans (66).
At 4–5 months of age, male age-matched C57BL/6J mice (25–32 g) were anesthetized with a Ketamine/xylazine mixture, and an indwelling catheter was inserted into the right jugular vein as previously described (54). The venous catheter was used for the multiple infusions. Experiments were performed in awake, unrestrained, chronically catheterized mice. The catheterized mice were grouped into BEL-treated and vehicle-treated groups.
Insulin sensitivity was assessed using the euglycemic hyperinsulinemic clamp technique (67) with modification (54). Briefly, food was removed for 5–6 h before the catheterized mice were iv injected with either BEL (3 μg/g body wt) or vehicle. A prime-continuous insulin infusion (21 mU/kg/min) was administered during the 90-min experiment. Blood glucose was clamped at 100 mg/dl ± 10 mg/dl by adjusting the rate of 10% glucose infusion according to blood glucose measurements performed every 10 min using a Glucometer Elite (Bayer Corp.). The glucose infusion rate (mg/kg/min) was calculated as the mean value for each 10-min interval during the last 60 min of the clamp. Under the euglycemic hyperinsulinemic state, and the amount of glucose infused is considered equivalent to whole body glucose disposal rate. Glucose disposal rate during the last 60 min was expressed as mean ± SEM mg/kg/min.
A 120-min hyperglycemic clamp was used to evaluate insulin secretion in β-cells. A continuous infusion of 10% glucose was used to maintain plasma glucose concentrations at 3 times the normal range (~300 mg/dl) as described previously (56). BEL (10 mM solution in 100% ethanol; infusion rate at 1 μl/min) was continuously administered during the experiment to inhibit any newly synthesized iPLA2. During the first 30 min, the glucose infusion rate was adjusted at 5-min intervals to elevate glucose concentrations to approximately 300 mg/dl. After this time, glucose levels were maintained around this level by adjusting the infusion rate at 10-min intervals. Insulin concentrations were measured in plasma samples obtained before and 30, 60, 90, and 120 min after the experiment had begun. Insulin levels were measured using an ultrasensitive insulin ELISA kit (Crystal Chem, Inc.). We averaged the insulin levels from the last four time points to determine the hyperglycemic challenge state as described (55).
C57/BL6J mice (The Jackson Laboratory, Bar Harbor, ME) were housed under a standard 12-h light, 12-h dark cycle with standard mouse chow and water ad libitum. Age-matched mice were used to perform each experiment. All procedures involving animals were approved by the Institutional Animal Care and Usage Committee (IACUC) of the Mount Sinai School of Medicine.
Results were expressed as mean ± SEM. The statistical significance of differences was analyzed with a two-tailed unpaired t test, where P < 0.05 was considered significant.
We thank Drs. Charles Mobbs, H. Henry Dong, and X. Sherry Chi for critical discussion of the manuscript.
This work was supported by grants from the National Institutes of Health/National Institutes of Diabetes and Digestive and Kidney Diseases NIDDK (RO1DK063076) and the Juvenile Diabetes Research Foundation, International JDRF (1-2002-646).