Protein Kinase C and Calcium Regulation of Adenylyl Cyclase in Isolated Rat Pancreatic Islets

Adenosine 3′,5′-cyclic phosphoric acid 2′-O-succinyl [¹²⁵I]iodotyrosine methyl ester was from New England Nuclear (Boston, MA). GLP-1 fragment 7-36 amide, cAMP, insulin-free bovine serum albumin (BSA), 3-isobutyl-1-methylxanthine (IBMX), carbamylcholine chloride (CCh), cholecystokinin fragment 26-33 sulfated amide (CCK), phorbol 12,13-dibutyrate (PDBu), phorbol 12,13-didecanoate (PDD), 1H-[1,2,4]oxadiazolo-[4,3-a]quinoxalin-1-1(ODQ), forskolin and phorbol 12-myristate 13-acetate (PMA) were from Sigma Chemical (St. Louis, MO). Fetal bovine serum was from Atlanta Biologicals (Norcross, GA). Collagenase type P was from Roche Molecular Biochemicals (Indianapolis, IN). CMRL-1066 medium was from GIBCO/Life Technologies (Grand Island, NY). Bisindolylmaleimide (GF 109203X) and bisindolylmaleimide XI HCl (Ro-32-0432) were from Calbiochem-Novabiochem (San Diego, CA). 1,2-Bis(o-amino-5-fluorophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetraacetoxymethyl ester (BAPTA/AM) was from Molecular Probes (Eugene, OR). Antibody to cAMP was a gift from Dr. David L. Garbers (Howard Hughes Medical Institute, Dallas, TX). All other chemicals were reagent grade.

Pancreatic islets were isolated using the collagenase method from pancreases excised from decapitated male Harlan Sprague-Dawley rats (225–300 g). All animal procedures were approved by the Institutional Animal Care and Use Committee. Isolated islets were either used immediately as freshly isolated islets or were cultured for 20 h in CMRL-1066 culture medium containing 5.5 mmol/l glucose, 2 mmol/l glutamine, 10% fetal bovine serum, penicillin (100 units/ml), and streptomycin (100 μg/ml), as described previously (26). The islets were cultured in a humidified (95%) atmosphere of 5% CO₂/95% air at 35°C. Levels of cAMP in freshly isolated versus cultured islets were not significantly different under basal or stimulated conditions.

For each experiment, islets (10 islets/sample) were washed three times in a Krebs-Ringer bicarbonate HEPES (KRBH) buffer (pH 7.4) containing 2.8 mmol/l glucose (basal) and 0.01% BSA, as described previously (26). Both fresh and cultured islets were incubated in KRBH buffer for 30–60 min at 37°C in an atmosphere of 95% O₂/5% CO₂ in a shaking water bath. Then, the islets were preincubated at a final volume of 0.1 ml KRBH buffer with various agents: BAPTA/AM, 20 min; GF 109203X and Ro-32-0432, 1 h; and ODQ, 30 min. All islet samples were pretreated with the phosphodiesterase inhibitor IBMX (0.2 mmol/l), 20 min, before time-zero stimulation with agents described in the text. Islet cAMP levels were determined at time-zero and after up to 20 min of incubation. All treatments were performed in duplicate. The reaction was stopped by the addition of ice-cold 0.1 N HCl (final concentration), and samples were immediately frozen. After three freeze/thaw cycles, the samples were sonicated and neutralized with sodium acetate buffer (final concentration 25 mmol/l) before acetylation and radioimmunoassay (RIA), as described previously (7,27). The cAMP levels of islets at time-zero were subtracted from the later time values to determine cAMP production.

Values are means ± SE. Significant differences between treatment groups were determined by one-way analysis of variance (ANOVA) with post hoc analysis using Student-Newman-Keuls multiple-comparison test. Values of P ≤ 0.05 were accepted as significant.

To determine AC activity in isolated rat pancreatic islets, cAMP levels were quantitated in the presence of a phosphodiesterase inhibitor, IBMX (0.2 mmol/l). Stimulation of muscarinic receptors with CCh increased islet cAMP levels in a concentration-dependent manner (Fig. 1). Maximal stimulation of AC occurred between 200 and 500 μmol/l CCh with a 4.5-fold increase in cAMP, although significant stimulation was observed at 10 μmol/l CCh with a 2.7-fold increase in cAMP (Fig. 1). CCh stimulation increased cAMP levels maximally within 5 min and levels remained significantly higher than control values after 20 min (Fig. 2). The presence of the muscarinic receptor antagonist atropine (50 μmol/l) did not significantly change basal cAMP levels compared with control values (data not shown) but inhibited the CCh (200 μmol/l)-induced increase in cAMP by 81 ± 17% (P < 0.001), such that there was no significant change compared with control values.

The specificity of the cAMP response was tested in islets in two ways. In the absence of IBMX, CCh failed to significantly increase cAMP levels (3.3 ± 0.3 fmol cAMP/10 islets) compared with basal values (1.3 ± 1.3 fmol cAMP/10 islets) after a 20-min stimulation. Moreover, in the presence of a specific guanylyl cyclase inhibitor, ODQ (10 μmol/l), islet cAMP generation in response to CCh (727 ± 124 fmol cAMP/10 islets) was not significantly different (P > 0.05) from levels with CCh treatment alone (629 ± 103 fmol cAMP/10 islets). Thus, guanosine 3′,5′-cyclic monophosphate did not contribute to the perceived effects of CCh on cAMP generation.

CCK is an intestinal hormone that stimulates receptors on β-cells and regulates insulin secretion (28). Similar to CCh, CCK (400 nmol/l) increased cAMP levels 4.1-fold compared with control values (Fig. 3).

The effects of the PKC-activating phorbol ester, PDBu, on islet cAMP levels were determined. PDBu increased cAMP levels 6.2-fold compared with control (Fig. 3), whereas PDD, a biologically inactive phorbol ester, did not significantly affect cAMP generation.

The combination of CCh and PDBu did not potentiate cAMP accumulation compared with PDBu alone, although a significant increase in cAMP generation was observed compared with CCh (Fig. 3). A combined stimulus with PDBu and CCK increased cAMP levels as much as 6.3-fold, but this was not different from the PDBu response alone (Fig. 3). Thus, neither CCh nor CCK effects were additive with PDBu.

The relative efficacy of PDBu was compared with the direct-acting AC activator forskolin and the GLP that activates receptor-mediated GTP-binding protein G_αs and AC. Forskolin (10 μmol/l) increased islet cAMP levels up to 58-fold (Table 1), and GLP (0.1 μmol/l) increased cAMP levels almost 10-fold (2,694 ± 827 fmol cAMP/10 islets) (P < 0.001) after a 20-min incubation. Thus, PDBu did not maximally activate islet AC compared with these stimuli.

Glucose is the primary stimulus for insulin secretion in islets and has been previously reported to increase cAMP levels in islets (2). In the present study, glucose (20 mmol/l) stimulation increased cAMP levels 2.7-fold. When glucose (20 mmol/l) was combined with PDBu (1 μmol/l), there was a 6.7-fold increase in cAMP (Fig. 4). However, the glucose and PDBu responses were not additive and were not significantly higher than cAMP levels with PDBu stimulation alone.

A role for PKC stimulation in islet cAMP generation was confirmed with the specific PKC inhibitors GF 109203X and Ro-32-0432. The presence of GF 109203X and Ro-32-0432 resulted in an ∼60% decrease in PDBu-induced cAMP levels (Fig. 5A). Similarly, both GF 109203X and Ro-32-0432 decreased CCh-stimulated cAMP levels by ∼50% (Fig. 5B). GF 109203X also inhibited CCK-stimulated cAMP generation by ∼30%, but Ro-32-0432 completely inhibited the CCK response (Fig. 5B). Neither GF 109203X nor Ro-32-0432 affected basal cAMP levels or forskolin-stimulated cAMP levels (Table 1).

To determine whether Ca²⁺ mediated the activation of AC, islets were treated with an intracellular Ca²⁺ chelator, BAPTA/AM (29). In BAPTA/AM-treated islets, the cAMP responses to CCh and CCK were inhibited by 50 and 45%, respectively, although there was no significant effect on basal cAMP levels (Fig. 6). BAPTA/AM also significantly reduced forskolin-stimulated cAMP production by ∼25% (Table 1). In contrast, PDBu responses were not significantly affected by the presence of BAPTA/AM (Fig. 6).

Additional experiments were performed in the absence of extracellular Ca²⁺ and the presence of EGTA (100 μmol/l) to chelate residual Ca²⁺. Under Ca²⁺-free conditions, cAMP generation in response to PDBu, CCh, and CCK was significantly inhibited by ∼50% for each agent (Fig. 7). Forskolin-stimulated cAMP production, however, was not significantly affected by Ca²⁺-free conditions (Table 1).

AC activation in the absence of extracellular Ca²⁺ and the presence of EGTA and BAPTA/AM was determined. Even with this stringent Ca²⁺ depletion, CCh significantly increased cAMP levels by 11-fold compared with the effects of CCh in the absence of BAPTA/AM (Fig. 8A). Likewise, with Ca²⁺ depletion, CCK significantly increased cAMP levels by 5.4-fold compared with effects in the absence of BAPTA/AM (Fig. 8B). To determine whether PKC played a role in AC activation under stringent Ca²⁺-depleted conditions, PKC inhibitors were included in the islet treatment. In the presence of CCh plus GF 109203X or Ro-32-0342, cAMP levels were decreased by 75 and 51%, respectively, compared with Ca²⁺-depleted islets in the absence of inhibitors (Fig. 8A). The PKC inhibitors also decreased CCK-induced cAMP production to a similar extent in Ca²⁺-depleted islets (Fig. 8B).

To determine whether PKC downregulation modulated the cAMP response to CCh or CCK, islets were cultured in the presence of the phorbol ester PMA (500 nmol/l) for 20 h to induce PKC downregulation (30). In PMA-pretreated islets, PMA, CCh, and CCK each failed to increase cAMP production above basal levels, whereas each of the stimuli significantly increased cAMP production in untreated islets (Fig. 9A). Moreover, islets that were pretreated with PMA showed a complete inhibition in AC activity in response to PMA, CCh, and CCK compared with similarly stimulated paired control islets (Fig. 9A, shaded versus open bars). Under stringent Ca²⁺ depletion in the absence of extracellular Ca²⁺ and in the presence of EGTA and BAPTA/AM, PMA-pretreated islet cAMP production in response to PMA, CCh, and CCK was significantly inhibited by 41, 59, and 72%, respectively, compared with similarly stimulated control islets that were not downregulated (Fig. 9B). In comparison with basal values, PMA but not CCh or CCK evoked a small stimulation of cAMP production in downregulated islets (Fig. 9B). In contrast, PMA, CCh, and CCK significantly stimulated cAMP production by 400–500% above basal levels under Ca²⁺-depleted conditions in control islets (Fig. 9B).

AC production of cAMP in β-cells has been attributed to glucose-evoking changes in ATP levels and Ca²⁺ mobilization/calmodulin activation (31,32,33,34). Indeed, Ca²⁺-stimulated AC activity has been reported for plasma membrane preparations of rat insulinoma cells (35). Stimulation of muscarinic receptors by CCh activates PLC with the generation of inositol 1,4,5-trisphosphate that mobilizes Ca²⁺ in rat islets (36,37) and diacylglycerol that stimulates PKC activation, while voltage-dependent Ca²⁺ channels (VDCC) (32,38,39) and store-operated channels (40) contribute to Ca²⁺ influx. Yet, there is controversy as to whether muscarinic receptor stimulation regulates islet AC activity. An early report described rat islet AC in homogenates as being sensitive to acetylcholine (0.1 mmol/l) (41), but later studies reported that muscarinic receptor stimulation by acetylcholine or CCh in intact rat islets did not affect cAMP production (32,42,43). However, these latter studies were performed in the absence of a phosphodiesterase inhibitor that is required to detect changes in islet cAMP levels. A lack of effect of acetylcholine on ob/ob mouse islet cAMP levels has also been reported (44); however, ob/ob mouse islets may be deficient in AC isoforms (45).

In the present study, CCh evoked a concentration-dependent stimulation of cAMP production in islets in the presence of a low substimulatory concentration of glucose (2.8 mmol/l). The specificity of the muscarinic receptor response was demonstrated by the inhibitory effect of atropine. Concentrations of acetylcholine (10⁻⁶ − 10⁻³ mol/l) similar to those for CCh in the present study also increased intracellular Ca²⁺ in rat islet β-cells at a low concentration of glucose (4.4 mmol/l) (46), suggesting that intracellular Ca²⁺ levels modulated CCh-induced AC activity. Moreover, CCK, a gut hormone that stimulates PLC activation (47) and elevates β-cell Ca²⁺ levels (46), also stimulated AC activity in rat islets. Glucose promotes insulin secretion by increasing Ca²⁺ influx (33) and Ca²⁺ mobilization (2,37). The present study confirms previous reports that glucose stimulates AC activity (1,27). Thus, CCh, CCK, and glucose increase cellular Ca²⁺ levels that together with calmodulin could modulate activity of AC type 3 (Fig. 10).

Several PKC isoforms are expressed in rat islets (48,49,50), including the conventional Ca²⁺-dependent/diacylglycerol-activated isoforms α and β, the novel Ca²⁺-independent/diacylglycerol-activated isoforms δ and ε (48), and the atypical Ca²⁺-independent/diacylglycerol-insensitive isoforms ζ and ι (50). PKC isoform sensitivity to phorbol esters parallels their diacylglycerol sensitivity (51). Since glucose (52,53,54), CCh (52,55), and CCK (56) increase islet and/or β-cell PKC activation, the effects of a PKC-activating phorbol ester were investigated for effects on AC activity. PDBu and PMA stimulated islet cAMP generation, whereas a biologically inactive phorbol ester, PDD, did not. However, no potentiation or additivity was evident among CCh, CCK, glucose, and PDBu in terms of AC activation. The results suggest that PDBu evoked a maximal PKC activation response—a mechanism shared by CCh, CCK, and glucose. However, the PKC-mediated AC activation was not maximal in terms of the islet cell capacity to generate cAMP. Both forskolin and GLP evoked larger increases in cAMP production than did PDBu, CCh, or CCK, suggesting that not all AC isoforms were maximally activated by the latter agents alone or in combination (Fig. 10). Forskolin is characterized as directly activating all AC isoforms, whereas GLP is receptor mediated and activates all isoforms through G_s regulation (13). It is probable that PKC stimulated AC types 2, 5, and 7 in islets (Fig. 10).

PKC mediation of the effects of PDBu on AC was confirmed by two specific PKC inhibitors; GF 109203X inhibits PKC-α, -βI, -βII, -γ, -δ, and -ε (57), whereas Ro-32-0432 inhibits PKC-α, -β1, and -ε (58). At the concentrations used in this study, the inhibitors are specific for PKC (59,60). GF 109203X may inhibit PKA (K_i = 2 μmol/l) and has been reported to affect intracellular Ca²⁺ stores at a concentration 10 times higher than that used in the present study (61). In islets, both PKC inhibitors reduced PDBu-, CCh-, and CCK-stimulated cAMP production. CCh-stimulated cAMP levels were decreased by half in the presence of either inhibitor, suggesting that PKC activation was partly responsible for mediating CCh-stimulated cAMP production. Similarly, GF 109203X partially inhibited CCK-stimulated cAMP production. However, Ro-32-0342 completely prevented CCK-stimulated increases in cAMP. The reason for the latter effect is not clear. Ro-32-0342 may be more potent at the possibly lower stimulation caused by nanomoles of CCK versus micromoles of CCh. However, the high concentration of CCh may induce other effects in the β-cell, such as changes in ion flux (38) that potentially affect AC activity and are not mediated by PKC. Neither GF 109203X nor Ro-32-0342 affected AC activity under basal conditions.

Further evidence that PKC mediates the activity of AC in islets came from experiments in which PMA pretreatment was used to downregulate PKC. In PMA-pretreated cells, neither PMA, CCh, nor CCK evoked a change in AC activity, as compared with the robust increase in cAMP observed with these agents in the absence of PKC downregulation. Thus, these data support the hypothesis that PKC activation mediates cholinergic and CCK stimulation of AC activation and cAMP generation in rat islets.

The role of extracellular Ca²⁺ on AC activation was studied by incubating islets in a Ca²⁺-free KRBH buffer with EGTA to chelate Ca²⁺. PDBu-, CCh-, and CCK-stimulated cAMP generation was significantly inhibited under Ca²⁺-free conditions, suggesting that Ca²⁺ influx plays an important role in AC activation. CCh promotes Ca²⁺ influx through VDCC (39) and by capacitative Ca²⁺ entry through store-operated channels (40,62). CCK-induced cytoplasmic Ca²⁺ mobilization in rat islet cells (63) may mediate effects on AC activation. However, the present results also suggest that Ca²⁺ influx—perhaps through capacitative Ca²⁺ entry—is a component of islet CCK effects.

The dependence of AC activity on intracellular Ca²⁺ was studied by inclusion of an intracellular Ca²⁺ chelator, BAPTA/AM. The presence of BAPTA/AM decreased CCh- and CCK-stimulated cAMP levels by ∼50%, suggesting that intracellular Ca²⁺ was partially responsible for the receptor agonist stimulatory effects. However, BAPTA/AM did not affect the response to PDBu, suggesting that phorbol ester and PKC effects on AC activation were not dependent on increases in intracellular Ca²⁺ levels. Interestingly, glucose augments insulin release when PKA and PKC are activated simultaneously in the absence of changes in intracellular Ca²⁺ with BAPTA (29).

The question of whether AC activity would be further affected by depletion of both extracellular and intracellular Ca²⁺ was investigated. Unexpectedly, stringent Ca²⁺ depletion resulted in increased CCh- and CCK-stimulated cAMP production compared with responses in Ca²⁺-free buffer alone. PKC may play a role in this stimulatory response, since the increase in cAMP levels was reversed by the inclusion of PKC inhibitors. Similarly, PKC downregulation also markedly reduced AC responses to CCh and CCK under Ca²⁺-depleted conditions. There was a small AC stimulation above basal by PMA in PKC-downregulated Ca²⁺-depleted islets that may be due to residual PKC activity. The relationship between PKC activity and Ca²⁺ depletion under these conditions is not entirely understood. It is likely that some islet PKC isoforms are active under stringent Ca²⁺-depleted conditions (48,50), such as novel PKC-δ and PKC-ε. Ca²⁺-independent novel PKCs have been implicated in upregulation of AC activity in macrophages (22), and AC activation in erythroid progenitor cells (64). Alternatively, PKC can phosphorylate βγ subunits and increase activation of type 2 AC (65). In islets, PKC isoform effects on AC may become apparent in the absence of Ca²⁺-supported enzyme activity.

CCh or CCK stimulation under stringent Ca²⁺ depletion also suggests that a Ca²⁺-related AC inhibition was relieved. Whether or not there is a Ca²⁺-mediated inhibitory component (66,67) to islet AC regulation is not known. However, if AC activity were relieved from an endogenous inhibitory effect of Ca²⁺, it is possible that PKC effects might become more pronounced, as in the present study.

In conclusion, this is the first study to characterize a role for PKC in AC activation in islets. A model for AC regulation in the β-cell (Fig. 10) shows CCh, CCK, and glucose stimulation activating PLC, with a resulting increase in inositol 1,4,5-trisphosphate–mediated Ca²⁺ mobilization and Ca²⁺ influx through either VDCC or store-operated channels. Ca²⁺/calmodulin activates AC type 3. Both intracellular Ca²⁺ mobilization and extracellular Ca²⁺ influx are important components of the islet AC response. The alternative diacylglycerol-mediated pathway, like phorbol esters, activates PKC, which then potentially activates AC types 2, 5, and 7. Forskolin and G-protein coupled receptor stimuli can stimulate each of the cyclase isozymes. A previous report characterized glucose stimulation of AC in β-cells as dependent on glucagon secreted from α-cells (68). In the present study, effects of CCh, CCK, or PDBu on glucagon secretion and stimulation of G-protein–coupled receptors (Fig. 10) cannot be ruled out as a mechanism contributing to the changes in cAMP production in islets. Moreover, the cAMP levels observed could be generated in any or all of the heterogeneous cell types that compose the pancreatic islet. However, the results imply that cholinergic innervation and gastrointestinal CCK may play an important role in the regulation of cAMP production, as well as insulin secretion, in islets of Langerhans.

This work was completed in partial fulfillment of the Doctor of Philosophy degree (for Y.T.), and was supported by a grant (DK-25705) from the National Institutes of Health (awarded to S.G.L.).