Glucose-induced Phosphorylation of Myristoylated Alanine-rich C Kinase Substrate (MARCKS) in Isolated Rat Pancreatic Islets*

In order to further evaluate the role of protein kinase C activation in glucose-induced insulin secretion, the extent of phosphorylation of the myristoylated ala-nine-rich C kinase substrate (MARCKS) was examined in freshly isolated rat pancreatic islets prelabeled with [saP]orthophosphate. The islets were incubated with either 2.76 mM glucose done, 2.76 mM glucose + 1 pM phorbol myristate acetate, 20 mM glucose, or 20 mM glucose + 50 nM staurosporine. After stimulation, the homogenized islets were processed by immunoprecipi-tation with a specific polyclonal anti-MARCKS antibody, followed by sodium dodecyl sulfate-polyacryl-amide gel electrophoresis. Densitometric analysis of autoradiograms revealed that phorbol myristate ace-tate caused a 3.78 f 0.97-fold increase in MARCKS phosphorylation over control. In the islets exposed to 20 mM glucose, an increase of 3.43 f 0.46-fold over control was observed. In islets exposed to G20 + 50 nM staurosporine, MARCKS phosphorylation was inhibited by 90 f 4% compared with control islets exposed to 20 mM glucose alone. Islets similarly treated (but incubated without ”P) were examined by immunocytochemistry using an a-PKC-specific monoclonal antibody and visualized by confocal immunofluorescence microscopy.

control was observed. In islets exposed to G20 + 50 nM staurosporine, MARCKS phosphorylation was inhibited by 90 f 4% compared with control islets exposed to 20 mM glucose alone. Islets similarly treated (but incubated without "P) were examined by immunocytochemistry using an a-PKC-specific monoclonal antibody and visualized by confocal immunofluorescence microscopy. The a-PKC redistributed from the cytosol to the plasma membrane in the &cells of islets exposed to 20 mM glucose. In separate experiments, unlabeled but similarly treated islets were shown to respond with a 6-7-fold increase in insulin secretion in static incubation. Thus, when freshly isolated rat pancreatic islets are exposed to stimulatory glucose concentrations, they exhibit both a translocation of a-PKC and a significant increase in the extent of phosphorylation of MARCKS protein. These data suggest that a-PKC is activated during glucose-induced insulin secretion.
Glucose-induced insulin secretion has been shown to depend on the activation of several signal transduction pathways, including increased calcium influx, increased intracellular cAMP (1)(2)(3)(4), and increased phosphoinositide hydrolysis ( 5 ) . More recently, the involvement of yet another regulatory * This work was supported by National Institutes of Health Grants DK-19813 and NIDDK-41230. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. molecule, protein kinase C (PKC),' in glucose-induced insulin secretion has been proposed. For instance, it is well known that phorbol esters (pharmacologic activators of PKC) like phorbol 12-myristate 13-acetate (PMA) can induce insulin secretion (6). When PMA is combined with agents that cause the activation of calcium influx (sulfonylureas) and cAMP production (forskolin), a normal, biphasic pattern of insulin secretion can be induced in isolated pancreatic islets (7). In addition, we have recently reported that a-PKC is present in pancreatic islets and that it translocates to the membrane fraction of islet homogenates in response to glucose stimulation (8). Although translocation of PKC to the membrane fraction has been regarded as indicative of PKC activation in many tissues, it has become evident that there is no simple correlation between the extent of PKC translocation and the magnitude of the cellular response (8). Hence, additional means for assessing PKC activation in intact cells are required. One such means is that of measuring the extent of phosphorylation of specific PKC substrates (9).
Myristoylated alanine-rich C kinase substrate (MARCKS) is a specific PKC substrate initially described in brain tissue, and later found to be present in a variety of tissues, including the rat pancreas (10, 11). Measurement of the extent of MARCKS phosphorylation has been used to specifically detect PKC activation in several tissues (12, 13). Using specific anti-MARCKS antibodies to immunoprecipitate 32P-labeled MARCKS, others have reported the absence of increased phosphorylation of MARCKS in glucose-stimulated cultured rat pancreatic islets (9). From these observations, these workers concluded that PKC activation is probably not an important feature in the glucose-induced signaling events in @-cells. This result differs from our observation showing glucoseinduced a-PKC translocation in freshly isolated rat pancreatic islets (8). The contradictory results of these two studies may imply that glucose induces PKC translocation but not MARCKS phosphorylation or that differences in the experimental techniques account for the apparent discrepancy. In order to clarify this issue, we examined the effect of glucose on the extent of MARCKS phosphorylation using freshly isolated rat pancreatic islets prelabeled with [32P]orthophosphate. Using this approach, we show that glucose-stimulated freshly isolated islets exhibit a significant increase in the phosphorylation of the PKC-specific substrate MARCKS and that translocation of a-PKC and insulin secretion also occur under the same experimental conditions. collagenase were purchased from Boehringer Mannheim. The monoclonal a-PKC antibody was obtained from Seikugaku America. All other reagents were purchased from Sigma unless otherwise indicated. The anti-MARCKS antibody was a generous gift from Drs. Paul Greengard and Angus Nairn from The Rockefeller University. The antibody was raised in rabbits against purified rat brain MARCKS and shown to recognize MARCKS in Western blot and immunocytochemical studies (14). Statistical analysis was performed using the two-tailed Student's t test for comparison of paired data and the Bonferroni test for comparison of multiple means.
Isolation of Rat Pancreatic Islets-The pancreas was dissected from anesthetized male Sprague-Dawley rats fed ad libitum and weighing between 260 and 300 g. The pancreas was collagenase-digested, and islets were picked manually and placed in a vial with 2 ml of buffer (bicarbonate-buffered media containing 115 mM NaCl, 5  islets were then stimulated with G2.75 + staurosporine or G20 + staurosporine. At the end of the incubation, the incubation buffer was removed and the islets placed on 100 pl of homogenization buffer (20 mM Tris, pH 7.4,5 mM EGTA, 1% Triton X-100,50 p~ leupeptin, 1 mM phenylmethylsulfonyl fluoride, 25 pg/ml aprotinin, 10 p M pepstatin A, and 0.1% p-mercaptoethanol). The islets were homogenized by bath sonication in ice-cold water. Parallel experiments were performed using the same protocol but without [32P]orthophosphate in the incubation buffer (21 islets/variable) and without Triton X-100 or @-mercaptoethanol in the homogenization buffer. These samples were assayed for protein content using the Bio-Rad Dc protein assay method. MARCKS Immunoprecipitation-The homogenized samples were loaded unto BM-Quick-Spin G50 Sephadex Columns (Boehringer Mannheim) and centrifuged at 1,100 X g for 4 min to dispose of unincorporated label. The eluate was recovered and processed for immunoprecipitation. Each sample was first precleared by incubation with 20 p1 of rabbit preimmune serum in 300 p1 of PBS, 1% Triton X-100 for 3 h at room temperature and with gentle shaking, followed by incubation with CL-4B-protein A beads (1 mg/20 pl) for 3 h. The samples were then centrifuged for 2 min at 2,000 rpm in an VS-15 Shelton centrifuge. The supernatant was saved for further processing. The supernatant from each sample was incubated overnight at 4 "C with 10 pl of anti-MARCKS antibody in 300 pl of PBS, 1% Triton X-100 adjusted to 100 mM NaCI. After overnight incubation, 20 pl of CL-4B-protein A was added to each sample and incubated for 3 h at room temperature. Again, the samples were centrifuged at 2,000 rpm for 2 rnin; this time the CL-4B-protein A pellet with the bound MARCKS was saved and the supernatant discarded. After washing the pellet twice with 200 pl of PBS, each sample was resuspended in 40 pl of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (0.25 M Tris pH 6.8, 2 mM EGTA, 4% SDS, 20% glycerol, 0.001% bromophenol blue, 10% p-mercaptoethanol) and boiled for 5 min. The samples were processed by SDS-PAGE using 8% gels (16). The gels were then fixed by incubation with 50% methanol, 20% acetic acid for 15 min, washed with deionized water, and dried in a Hoeffer SE 540 slab gel dryer. The labeled bands were visualized by autoradiography. Quantification of the extent of MARCKS phosphorylation was accomplished by densitometric analysis of the 86.5-kDa band corresponding to MARCKS, using a BioImage Visage 2000 densitometer, and expressed as optical density (O.D.). MARCKS phosphorylation was expressed as the fold-increase of each sample's O.D. over control O.D. (the control is a sample of islets that were only exposed to G2.75).
Measurements of Insulin Secretion-For insulin secretion in static incubation, 20-40 islets were laid on a nylon filter and placed in a plastic vial with 200 pl of incubation buffer with G2.75, sealed, and incubated for 90 min at 37 "C. After 90 min the buffer was carefully replaced with fresh buffer containing G2.75 and incubated for an additional 15 min (basal secretion). This buffer was collected and replaced with buffer containing G20. After 15 min of stimulation the buffer was collected (stimulated secretion) and the islets fixed with ice cold 4% paraformaldehyde in phosphate-buffered saline (140 mM NaCl, 10 mM Na2POI, pH 7.4). The aliquots of incubation buffer were analyzed for insulin content by radioimmunoassay (17), and the islets were processed for immunocytochemistry to assess the localization of a-PKC, as described below.
The capacity of the islet preparation to respond with a biphasic insulin secretion to a glucose stimulus was also assessed. Filters with 20-40 islets were first placed in vials with 200 pl of incubation/ perifusion buffer, sealed and incubated for 90 min at 37 "C as described above to simulate the 32P-labeling conditions. The filters containing the islets were then transferred to a perifusion chamber and perifused for 30 min with G2.75 to establish a stable basal insulin secretory rate. The islets were then perifused with G20 for maximal stimulation. Samples of the perifusate were collected at timed intervals for insulin analysis by radioimmunoassay (17).
Immunocytochemistry/a-PKC Translocation-At the end of the static incubation, the filter with islets was removed from the incubation chamber and immediately immersed in ice-cold 4% paraformaldehyde and fixed for 30 min on ice. The islets were washed three times with PBS. The filter was then placed in a plastic Petri dish filled with PBS and the fixed islets manually picked and loaded individually onto a drop of PBS that had been placed on a gelatinsubbed slide. Six to ten islets were placed on each slide, which was then air-dried and heat-fixed. The slides were stored at 4 "C. When ready for immunocytochemistry, the fixed islets were rehydrated with PBS for 10-15 min and permeabilized by incubating with 0.3% Triton X-100 in PBS for 15 min. This was followed by incubation with 50 mM ammonium chloride for 15 min and three consecutive washes with PBS-0.5% Triton X-100. After blocking nonspecific binding sites by incubating with 1% BSA in PBS for 15 min, the islets were exposed to a monoclonal anti-a-PKC antibody (MC-3a Seikugaku diluted 1:20 in PBS, 0.5% BSA, 0.15% Triton X-100) for 90 min and then washed three times with PBS, 0.5% Triton X-100. This was followed by incubation with anti-mouse fluorescein isothiocyanate (diluted 1:lOO) for 60 min. After three final washes with PBS, 0.5% Triton X-100 the sample was covered with Antifade (1 mg/ml of pphenylenediamine in 75% glycerol and 25% 0.1 M carbonate buffer, pH 9.6). PKC immunoreactivity was visualized by immunofluorescence confocal microscopy using a Bio-Rad MRC 600 confocal microscope.

RESULTS
Phosphorylation of MARCKS in Situ-Immunoprecipitation of samples of 32P-labeled pancreatic islets using an anti-MARCKS antibody revealed a band with a molecular mass of 86.5 kDa. This value is within the range of the molecular mass values described for MARCKS (Fig. 1B). A basal degree of MARCKS phosphorylation was found in samples of unstimulated pancreatic islets. Stimulation of the islets with 1 MM PMA caused a marked increase in the extent of phosphorylation. When the islets were stimulated with G20, a marked increase in MARCKS phosphorylation was observed again (Fig. 1B). When the autoradiograms were analyzed by densitometry we found that the increase in phosphorylation in the PMA-treated islets was 3.98 k 0.97-fold over control (n = 5; p < 0.01) and the increase in phosphorylation in the glucose-stimulated islets was 3.43 f 0.46-fold over control ( n = 9; p < 0.01) (Fig. lA). The results are expressed as mean f S.E.. Preliminary data examining the extent of MARCKS phosphorylation after 2 and 5 min of stimulation seems to indicate that phosphorylation is first detectable after 5 min.
If the islets were preincubated for 30 min with the PKC inhibitor staurosporine, the glucose-induced increase in phosphorylation was only 10% of a positive control of islets stimulated with G20 (90%&4% inhibition) ( n = 4; p 0.002), as one would expect if this phosphorylation was indeed dependent on PKC activation (Fig. 2). Incubation of the islets with 50 nM staurosporine in the presence of low glucose (G2.75) did not induce any statistically significant change in the extent of basal phosphorylation of MARCKS (1.21 f 0.28 expressed as -fold increase over basal; p = 0.6).
Protein assays of the unlabeled islets that were incubated and homogenized under conditions that permit accurate pro- Freshly isolated rat pancreatic islets were prelabeled with [3'P] orthophosphate. This was followed by stimulation for 15 min with G2.75, 1 PM PMA, or G20. After stimulation, the islets were homogenized and MARCKS protein immunoprecipitated with a specific anti-MARCKS antibody, as described under "Experimental Procedures." The phosphorylated MARCKS was visualized by autoradiography. A, the extent of MARCKS phosphorylation was quantified by densitometric analysis of the autoradiograms. The absorbance changes were expressed as -fold over control. Islets exposed to PMA (n = 5) had an increase in MARCKS phosphorylation 3.98 f 0.97fold over control ( p < 0.01). Islets exposed to G20 (n = 9) had an increase in MARCKS phosphorylation of 3.43 f 0.46-fold over control ( p < 0.01). The difference between PMA-treated and G2O-treated islets was not statistically significant. B, a typical autoradiogram is shown. tein measurements (no Triton X-100 or 8-mercaptoethanol) revealed that there was no significant difference between the amount of protein in stimulated (G20 or PMA) uersus unstimulated (G2.75) islets (not shown).
Insulin Secretion-As previously described, the islets were incubated for 90 min at 37 "C in the presence of G2.75 only, to mimic the 32P-labeling incubation. Subsequently the islets were either stimulated with G20 and insulin secretion assayed in a static incubation or transferred to a perifusion chamber and perifused with G20. Basal secretion in the presence of G2.75 was used as a control. When G20 was added to islets in static incubation, a 5.6 f 1.5-fold increase in the rate of insulin secretion was observed (Fig. 3A). When the islets were perifused, the insulin secretory response to glucose displayed 20-40 islets were laid on a nylon filter and incubated for 90 min at 37 "C, to mimic the [32P]orthophosphate labeling conditions. The number of islets was exactly matched for all variables within an experiment. The islets were then exposed to either G2.75 or G20 for 15 min. Samples for insulin secretion were obtained as described under "Experimental Procedures." The results are expressed as -fold increase over basal. The islets exposed to G20 exhibited a secretory response which was 5.6 f 1.5-fold greater than basal. B, in other experiments, groups of 20-40 islets were incubated for 90 min at 37 "C and then transferred to a perifusion chamber. The islets were perifused with buffer containing G2.75 for 30 min followed by G20 for 15 min. Timed samples were collected for insulin content determination. The results are expressed in pg/islet/min. The rate of insulin secretion 15 min after the addition of G20 was 13-fold over basal. a normal biphasic pattern (Fig. 3B). The rate of insulin secretion in the perifused islets 15 min after the addition of G20 was 13-fold over basal.
Translocation of a-Protein Kinase C-We have previously shown by Western blot that a-PKC translocates to the membrane fraction of pancreatic islets upon stimulation with G20 and that this translocation correlates with insulin secretion (8). We have now found that MARCKS phosphorylation also occurs upon stimulation with glucose. To show that a-PKC translocation occurred under the same experimental conditions, we examined the localization of a-PKC by immunocytochemistry in islets incubated for 90 min with G2.75 followed by stimulation with either 1 p~ PMA or G20. Visualization by immunofluorescence confocal microscopy showed a diffuse cytosolic staining pattern in the control islets (Fig. 4A). Control islets that were not exposed to the primary antibody did not show detectable staining (not shown). The islets treated with either PMA or high glucose showed a redistribution of the a-PKC with a significant increase in the staining in the periphery of the /3-cell (Fig. 4B).

DISCUSSION
The possible role of protein kinase C in glucose-induced insulin secretion has been a matter of debate (18,191. Several groups have reported negative results in experiments designed to detect PKC activation during glucose-induced insulin secretion by detecting PKC translocation (9,20,21). These studies used either measurements of PKC activity or measurements of [3H]4-phorbol 12,13-dibutyrate binding in the cytosolic and membrane fractions of homogenized islets to detect translocation. Using isoenzyme-specific anti-PKC antibodies, we recently showed by Western blot that a-PKC is present in pancreatic islets. We also demonstrated that a-PKC translocates from the cytosolic fraction to the membrane fraction of isolated rat pancreatic islets stimulated with high FIG. 4. Immunocytochemical localization of a-PKC in rat pancreatic islets. Groups of 20-40 islets were exposed to G2.75,

I'MA, or G20 in a static incuhation as descrihed under "Experimental
Procedures." The islets were then fixed and probed hy immunocytochemistry with a monoclonal anti-n-PKC antihody, using fluorescein isothiocyanate anti-rahhit as a secondary antihody. The &cells were visualized hy immunofluorescence confocal microscopy. A , in the islets exposed t o G2.75 alone the pattern of immunostaining was found to he diffuse and cytosolic. I?, in the islets exposed to either I'MA or C20 the imrnunostaining WAS ohserved to redistrihute to the periphery of the &cell. A photomicrograph of the G20 stimulated islets is shown. Control islets that were not exposed to the primary mtihody had no detectahle staining (not shown).
glucose as well as other fuel agonists. This translocation correlates temporally with insulin secretion and is blocked by inhibition of glucose metabolism with mannoheptulose (8). Most recently we have confirmed and extended these observations by showing a glucose-induced translocation of a-PKC to the plasma membrane of the P-cell with the use of immunocytochemistry.2 In addition, we have shown that the PKC inhibitor staurosporine causes a marked inhibition of glucoseinduced insulin secretion under conditions where staurosporine does not inhibit the rate of glucose metabolism (22).
In many systems, translocation of PKC alone has been accepted as evidence of its activation during a given cellular event (23,24). However, this assumption has been challenged by some authors. For example, Trilivas and co-workers (25) found that, in 1321N1 astrocytoma cells, stimulation with carbachol induces a translocation of PKC much earlier than a rise in diacylglycerol. Since PKC is thought to require both Ca" and diacylglycerol for its activation, these data would argue that in the case of the 1321N1 astrocytoma cells PKC translocates to the membrane hut without activation of its kinase activity. In another example, Salari and co-workers (26) reported that platelet-activating factor, thrombin and prostacyclin induce an increase in the PKC activity (measured by histone 1 phosphorylation in vitro) found in the cytosolic and particulate fractions of rabbit platelets, without appreciable change in the relative distribution of the total PKC activity (i.e. without evidence of translocation).
Several groups have also studied the phosphorylation pattern of phosphoproteins during glucose-induced insulin secretion using PAGE of crude extracts of "'P-labeled pancreatic islets (27) or two-dimensional gel electrophoresis and immunoprecipitation of selected phosphoproteins with specific antibodies (9). The phosphorylation of several phosphoproteins of different molecular weights were induced by phorbol esters, suggesting that they may he PKC substrates (27) MARCKS, a specific substrate for PKC. The extent of MARCKS phosphorylation in islets was markedly stimulated with either PMA or glucose. Preliminary data seems to indicate that this is detectable as early as 5 min after the initiation of the glucose stimulus, which correlates with our previous data in perifused islets showing (r-PKC translocation after 5 min in response to G20 (8). These results strongly suggest that glucose stimulates PKC-dependent phosphorylation of proteins in pancreatic islets and argue for the activation of PKC during glucose-induced insulin secretion.
The glucose-induced increase in MARCKS phosphorylation was blocked by inhibiting the action of PKC with staurosporine. Staurosporine is a very effective inhibitor of PKC activity as well as a potent inhibitor of insulin secretion (22. 28). However, staurosporine is not a completely specific inhibitor of PKC since it can also inhibit protein kinase A (PKA), with a higher K,, hut still within the same order of magnitude (29). The phosphorylation domain of MAHCKS contains, in addition to the four PKC phosphorylation sites, a 3-amino acid sequence that has heen identified as a phosphorylation site in PKA suhstrates. Although there is evidence for weak phosphorylation of all of the above mentioned sites by PKA in uitro, PKA activation in intact cells has been shown to have no effect on MARCKS phosphorylation (30. 31). In a previous study, we showed that concentrations of staurosporine between 20 and 100 nM inhihit the second phase of glucose-induced insulin secretion by 70-80T. However, staurosporine did not affect the glucose-induced increases in phosphoinositide hydrolysis and glucose usage and actually enhanced the first phase of insulin secretion (22). These results make it unlikely that staurosporine's effect on insulin secretion or MARCKS phosphorylation is due to nonspecific toxic effects on the (j-cell. Thus, the inhibition of glucoseinduced MARCKS phosphorylation by staurosporine in intact pancreatic islets argues that this is a PKC-mediated event. The observation that staurosporine does not affect the hasal phosphorylation of MARCKS at the concentration and time of exposure used for our experiments may indicate that at least one of MARCKS phosphorylation sites has a slow turnover. Easom and co-workers have also studied the effect of glucose on MARCKS phosphorylation in rat pancreatic islets (9). They found that stimulation of cultured islets with 500 p M carbachol induces a statistically significant increase in MARCKS phosphorylation after 5 min of stimulation and that this is potentiated by concomitant stimulation with 500 PM carbachol and 10 mM glucose. However, they did not observe a significant increase in MARCKS phosphorylation when islets were exposed to high glucose alone. They could find no evidence of glucose-induced PKC translocation either (9). The exact reasons for the difference between their results and the present results are not clear. The main difference in experimental protocols is that Easom's group used islets cultured overnight in medium containing 5.5 mM glucose, whereas our group used freshly isolated islets. I t has been shown that overnight culture of pancreatic islets in medium containing low glucose suppresses glucokinase activity and that these islets are less sensitive to glucose stimulation (32, 33). The latter observation may translate into an impairment in signal transduction in islets cultured overnight, as suggested by the absence of glucose-induced PKC translocation and MARCKS phosphorylation in cultured islets.
In order to correlate MARCKS phosphorylation with the insulin secretory response we measured insulin secretion u nder conditions similar to those used for the phosphorylation experiments. Measurement of insulin secretory rates of idets in static incuhation after stimulation with G20 showed a 5-7-fold increase over basal. Because of the wide differences in the incubation protocols (different glucose concentration, time of incubation, volume of incubation, etc.) it is difficult t o make direct comparisons of the secretory capacity of the islets employed by different investigators. However, the secretory rates we obtained seem to be comparable to those reported by others using islets stimulated in static incubation (21,34). However, this rate of secretion is less than half of the usual secretory rates obtained when freshly isolated islets are perifused (22). Because of this difference, the issue of islet responsiveness was analyzed further by taking islets which were preincubated under conditions similar to those used for the phosphorylation experiments, and then perifused. The perifused islets responded to glucose stimulation with a normal, biphasic secretory pattern and with a second phase of insulin secretion as high as 13-fold above basal. The lower secretory response of the islets under static incubation conditions may be related to a lower oxygen tension in the incubation buffer than that achieved in perifusion, where the buffer is gassed constantly (35). It could also be due to a negative feedback of insulin on its own secretion, a phenomenon that has been described by others (36). Nonetheless, we have shown that insulin secretion occurs under the experimental conditions under which glucose-induced MARCKS phosphorylation is observed.
The cellular function of MARCKS has not been yet established. It is known that MARCKS is a calmodulin-binding protein and that its affinity for calmodulin is dependent on its phosphorylation state (37,38). PKC phosphorylation of MARCKS results in the rapid release of calmodulin making it available for activation of calmodulin-dependent enzymes (38). This kind of interaction could allow for mechanisms for cross-talk between the PKC and the calmodulin-dependent signal transduction pathways. There is also evidence that MARCKS is localized to cytoskeletal structures in its unphosphorylated state and that upon PKC phosphorylation it detaches and redistributes to the cytosol (39). This observation raises the possibility that the phosphorylation-dependent regulation of MARCKS' membrane binding might locally modify the interaction between the cytoskeleton and the membrane and in this manner influence membrane associated cellular events, like the exocytotic secretion of peptide hormones such as insulin.
Our previous published work with freshly isolated perifused islets shows translocation of a-PKC from the cytosolic to the membrane fraction of islets stimulated with glucose as demonstrated by Western blot analysis of the fractions (8). In a set of experiments to be published separately a-PKC's localization within the P-cell was studied by immunocytochemistry using an a-PKC specific monoclonal antibody and visualized by confocal immunofluorescence microscopy.* These experiments showed that a-PKC co-localizes with insulin to the pcells of pancreatic islets. In the basal state the a-PKC has a diffuse cytosolic distribution; after perifusion in the presence of high glucose the a-PKC translocates to the periphery of the B-cell, suggesting that a-PKC is translocating to the plasma membrane of the &cells or to a closely associated domain. Similar results were observed in fixed sections of pancreata of rats infused intravenously with glucose and sacrificed after 15 min. Our present study confirms that this pattern of change in a-PKC localization can also be observed in islets stimulated with glucose in a static incubation, within the same time frame that MARCKS phosphorylation is detected.
In summary, we have established that high glucose by itself is capable of inducing the phosphorylation of MARCKS in isolated rat pancreatic islets, and that this change in MARCKS phosphorylation correlates with a-PKC translocation from the cytosol to the plasma membrane of the p cell or to a compartment closely associated with the plasma membrane, as assessed by immunocytochemistry. These two separate lines of evidence strongly support the hypothesis that PKC is activated in islets exposed to glucose. The fact that insulin secretion is also observed within the same time frame suggests that a-PKC may be actively involved in mediating glucose-induced insulin secretion. A better understanding of the physiological role of MARCKS may eventually allow us to more firmly establish a relationship of cause and effect between PKC activation and insulin secretion.