Extracellularly Regulated Kinases 1/2 (p44/42 Mitogen-Activated Protein Kinases) Phosphorylate Synapsin I and Regulate Insulin Secretion in
http://www.100md.com
内分泌学杂志 2005年第2期
Institut National de la Santé et de la Recherche Médicale Unité 376 (C.L., S.C., E.H.H., D.B., S.D.), Centre Hospitalier Universitaire Arnaud de Villeneuve, 34295 Montpellier Cedex 5, France; and Unité Mixte de Recherche 5160, Centre National de la Recherche Scientifique (C.B.), Institut de Biologie-Faculté de Médecine, 34060 Montpellier Cedex 1, France
Address all correspondence and requests for reprints to: Dr. Stéphane Dalle, Unité Institut National de la Santé et de la Recherche Médecale U376, Centre Hospitalier Universitaire Arnaud de Villeneuve, 371 Rue du Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France. E-mail: dalle@montp.inserm.fr.
Abstract
The p44/p42 MAPKs (ERK1/2) cascade regulates ?-cell nuclear events, which modulates cell differentiation and gene transcription, whereas its implication in processes occurring in the cytoplasm, such as activation of the exocytotic machinery, is still unclear. Using the MIN6 ?-cell line and isolated rat islets of Langerhans, we investigated whether glucose, by activating the ERK1/2 cascade, induces phosphorylation of cytoplasmic proteins implicated in exocytosis of insulin granules such as synapsin I. We observed that the majority of ERK1/2 activity induced by glucose remains in the cytoplasm and physically interacts with synapsin I, allowing phosphorylation of the substrate. Therefore, we reexamined the potential requirement of ERK1/2 for insulin secretion. Blocking activation of ERK1/2 using MEK1/2, the MAPK kinase inhibitor PD98059 or using small interfering RNA-mediated silencing of ERK1 and ERK2 expressions resulted in partial inhibition of glucose-induced insulin release, indicating that ERK1/2 pathway participates also in the regulation of insulin secretion. Moreover, using the pancreatic islet perifusion model, we found that the ERK1/2 activity participates in the first and second phases of insulin release induced by glucose. Taken together, our results demonstrate new aspects of the glucose-dependent actions of ERK1/2 in ?-cells exerted on cytoplasmic proteins, including synapsin I, and participating in the overall glucose-induced insulin secretion.
Introduction
INCREASING CONCENTRATIONS OF circulating glucose, in the 3–20 mM range, stimulate insulin secretion by generating and amplifying signals in pancreatic ?-cells (1, 2, 3). After transportation into the ?-cell by an insulin-insensitive facilitated transporter, glucose is metabolized, producing ATP (1, 2, 3). The increase in the ATP to ADP ratio induces the closure of ATP-dependent potassium channels, which leads to membrane depolarization, promoting opening of L-type voltage-dependent calcium channels (VDCCs) (1, 2, 3). The resulting calcium entry generates a rapid and sustained rise in cytoplasmic-free calcium concentrations [cytoplasmic (Ca2+)], which trigger the biphasic insulin release (1, 2, 3). The first phase requires a rapid and marked elevation of cytoplasmic [Ca2+] and corresponds to exocytosis of a limited readily releasable pool of insulin secretory granules docked to the plasma membrane (1, 2, 3, 4). The sustained second phase of secretion requires signal(s) additional to the cytoplasmic [Ca2+] rise and corresponds to the recruitment to the plasma membrane of a reserve pool of secretory granules by activation/rearrangement of the components of the exocytotic machinery (1, 2, 3, 4, 5, 6, 7, 8, 9). However, the amplifying signaling pathways, which increase the efficiency of calcium on insulin exocytosis, remain in a large part unidentified (1, 2, 3).
Glucose-induced calcium entry has been reported to activate various kinases in the ?-cells, including protein kinase A (PKA) via calcium influx-dependent cAMP production (10), calmodulin kinase II (CAMKII) (11), protein kinase C (12), and p44/42 MAPKs (also called ERK1/2) (10, 13, 14, 15, 16, 17). PKA, protein kinase C, and CAMKII have been shown to be involved in the glucose-induced insulin secretion (18, 19, 20). ERK1/2 are the terminal kinases in a ubiquitous and pleiotropic three-kinase cascade, consisting of Raf isoforms that activate MEK1/2, MAPK kinase which, in turn, activate ERK1 and ERK2. ERK1/2 have been implicated in many cellular events, including proliferation, differentiation, survival, and secretion (21, 22). In the ?-cells, glucose-activated ERK1/2 have been described, so far, to regulate insulin gene transcription (14, 23).
Over its physiological concentration range, glucose activates ERK1/2 in pancreatic ?-cells (10, 13, 14, 15, 16, 17). Signals from extracellular calcium entry, cAMP production, and intracellular calcium stores were suggested to mediate the glucose effects on ERK1/2 (10, 13, 14, 15, 16, 17), besides those observed on insulin release. Thus, increased ERK1/2 activity by glucose may correlate with insulin secretion. Nevertheless, studies have shown a lack of relationship between ERK1/2 activity and insulin release (13, 16, 24). On the other hand, using the insulinoma (INS)-1 ?-cell line, it has been shown that the glucose-activated ERK1/2 partially translocate to purified nuclear fractions, whereas the majority of active ERK1/2 was found in the cytoplasm of glucose-stimulated cells (16). Thus, it is possible that ERK1/2 also regulate glucose-dependent cytoplasmic events involved in the ?-cell biology. To date, no cytoplasmic function for ERK1/2 has been reported in ?-cells.
To address this issue, we used the MIN6 cells and isolated rat pancreatic islets to examine the potential requirement of ERK1/2 activity in phosphorylation and activation of cytoplasmic targets, and we reexamined the potential requirement of ERK1/2 for insulin secretion. We studied synapsin I, a well-known ERK1/2 substrate in neurons, present in the ?-cell cytoplasm and expressed in normal rat islets (8, 9). Synapsin I is associated with the insulin-secretory granules in MIN6 cells (9), and phosphorylation of synapsin I by PKA and CAMKII has been proposed to be implicated in the glucose-induced insulin release (8, 9, 18, 20). We observed that, upon glucose stimulation, ERK1/2 associate with synapsin I and that the glucose-induced ERK1/2 activity leads to phosphorylation of synapsin I. On the other hand, we found that glucose-induced insulin release from MIN6 cells and isolated rat pancreatic islets is partially dependent on ERK1/2 activity. Thus, our data demonstrate that ERK1/2 activity is a significant regulator of glucose-dependent cytoplasmic events, including phosphorylation of synapsin I, and suggest, for the first time, a significant involvement of ERK1/2 in the insulin secretory process.
Materials and Methods
Animals
Male Wistar rats (Iffa-Credo, l’Arbresle, France) had free access to laboratory chow and water. They were housed, handled, and killed according to the rules of the Centre National de la Recherche Scientifique Animal Care and Use Committee.
Materials
Mouse anti-ERK1/2 antibody was obtained from Transduction Laboratories (BD Biosciences Pharmingen, San Diego, CA). Mouse anti-phospho-ERK1/2 (p44/42 MAPK) antibody, which selectively recognizes the doubly phosphorylated, active forms of these kinases, and rabbit anti-synapsin I antibody were purchased from Cell Signaling (New England Biolabs, Beverly, MA). Goat anti-synapsin I antibody, horseradish peroxidase-linked donkey antirabbit, or antimouse antibodies, and protein A/G-plus agarose were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse antiphosphoserine antibody was obtained from Abcam (Cambridge, UK). Fluorescein isothiocyanate-conjugated donkey antimouse antibody was purchased from Vector Laboratories (Burlingame, CA). DMEM and fetal calf serum (FCS) were purchased from Invitrogen Life Technologies (Grand Island, NY). Nitrocellulose transfer membranes (Protran) were obtained from Schleicher & Schuell (Dassel, Germany). PD98059 was purchased from Calbiochem (La Jolla, CA). 45CaCl2 (2 mCi/ml) was from Amersham Pharmacia Biotech (Amersham, UK). All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO).
Pancreatic islets preparation and MIN6 cells culture
Pancreatic islets were isolated from fed male Wistar rats weighing 280–320 g on the day the rats were killed. Islets were isolated by collagenase digestion followed by Ficoll gradient separation, with a method adapted from (25). The MIN6 cell line (passages 18–24) was maintained as described previously (26, 27, 28). Cultures were never allowed to become completely confluent. MIN6 cells used in the experiments were tested for the dose-dependent insulin response to glucose, which always was in the physiological range (26).
Insulin secretion from static incubation of isolated pancreatic islets
Isolated islets were stabilized for 2 h at 37 C in HEPES-balanced Krebs-Ringer buffer (KRB) [119 mM NaCl; 4 mM KCl; 1.2 mM KH2PO4; 1.2 mM MgSO4; 2.5 mM CaCl2; 20 mM HEPES, pH 7.2)] containing 0.1% BSA (KRB) and 2.8 mM glucose and then stimulated (3 islets/tube) by 16.7 mM glucose in static incubation. When used, nifedipine (2 μM) was added during the 2.8- and 16.7-mM glucose stimulation periods, whereas PD98059 (20 μM) was added from the beginning of the stabilization period and maintained until the end of the stimulation. To prevent an effect related to the toxicity of the solvent-carrier dimethyl sulfoxide, control islets were always treated with dimethyl sulfoxide in the same conditions of concentration and time as for the PD98059-treated islets. At the end of stimulation, islets were pelleted by centrifugation and lysed in acid-ethanol for assessment of insulin content. Insulin secretion was quantified by RIA using [125I]porcine insulin, rat insulin (Novo Nordisk, Copenhagen, Denmark) as standard, and the guinea pig antiporcine insulin antibody 41 (29). Results are presented as insulin secreted (nanograms per milliliter) normalized to islet insulin content extracted with acid-ethanol overnight at –20 C.
Perifusion of isolated pancreatic islets
For perifusion experiments, after a 2-h static incubation in KRB containing 2.8 mM glucose in the presence or the absence of PD98059, batches of 30 islets were placed in perifusion chambers (two chambers per bath) at 37 C and perifused at a flow rate of 1.5 ml/min with KRB containing various concentrations of glucose as indicated in the presence or the absence of PD98059. Islets were equilibrated for 20 min in KRB supplemented with 2.8 mM glucose and then stimulated for 30 min with 16.7 mM glucose. The 16.7-mM glucose stimulation period was followed by a 20-min perifusion in 2.8 mM glucose.
Insulin secretion from MIN6 cells
MIN6 cells were plated in 24-well plates at a density of 1 x 106 cells/well for 3–5 d. Insulin secretion studies were performed using a static incubation method in 5% CO2 at 37 C with cells (70–80% confluence) as described previously (26). For the experiments with PD98059, MIN6 cells were preincubated for 2 h in KRB in the absence or presence of PD98059 before stimulation of insulin secretion with 10 mM glucose, and PD98059 was present throughout the incubation period unless otherwise indicated. For determination of insulin content, MIN6 cells were lysed for 30 min at 4 C in PBS + 1% Triton X-100.
Measurement of 45Ca2+ influx
MIN6 cells were plated in 24-well plates at a density of 1 x 106 cells/well for 3–5 d. Twenty-four hours before the experiments, the culture medium was changed. On the day of the experiment, the cells were preincubated for 2 h in 500 μl KRB in 5% CO2 at 37 C. The preincubation solution was then replaced by 250 μl KRB containing 8 μCi/ml 45CaCl2 and the test agents. The reaction, developed at 37 C, was stopped by aspiration of the medium. The cells were rapidly washed three times with 1 ml ice-cold buffer (135 mM NaCl, 5 mM KCl, 2.5 mM CaCl2, 1 mM LaCl3, 10 mM HEPES). The cells were then solubilized in 1 ml KRB containing 0.1% Triton X-100 for 1 h at room temperature. An aliquot of the solution (100 μl) was then assayed for 45Ca2+ content in a ?-counter after addition of a liquid scintillation medium (complete phase combining system, Amersham Pharmacia Biotech).
Small interfering (si)RNA-mediated silencing of endogenous expression of ERK1 and ERK2 in MIN6 cells
Expression of either ERK1 or ERK2 was specifically silenced in MIN6 cells using mouse 20–25 nucleotide prevalidated siRNA duplexes purchased from Santa Cruz Biotechnology. On the day before transfection, MIN6 cells were resuspended in 12-well plates in appropriate growth medium with 15% FCS but without antibiotics. Cells were then grown overnight to reach 40–50% confluency. The day of the experiment, siRNA complexes were prepared and transfection was performed according to the manufacturer’s instructions. Concentration of ERK1 and ERK2 siRNAs for MIN6 cell transfection required optimization (data not shown). Briefly, 3.6 μl of 10 μM of ERK1 or ERK2 siRNA (final concentration 50 nM) was mixed gently with 60 μl siRNA transfection medium (solution A), and 3.6 μl siRNA transfection reagent was mixed with 14.5 μl siRNA transfection medium (solution B). After a 5-min incubation at room temperature, solution A and solution B were combined and incubated at room temperature for 20 min to form transfectant-siRNA complex. Medium from MIN6 cells was then removed and replaced by 600 μl fresh growth medium with 15% FCS but without antibiotics. Transfectant-siRNA complexes were added dropwise while gently rocking the 12-well plates. MIN6 cells were transfected with ERK1 or ERK2 siRNA for at least 5 h at 37 C before switching to fresh growth medium with 15% FCS including antibiotics. Then 24, 48, 72, and 96 h after transfection, cells were lysed and expressions of either ERK1 or ERK2 were assayed by Western blot. The amounts of siRNA, siRNA transfection reagent, and siRNA transfection medium were proportionally scaled down to the surface area of cell culture (24-well plates format) to perform insulin secretion.
Western blotting and immunoprecipitation
Pancreatic islets were stabilized for 2 h in KRB containing 2.8 mM glucose in the absence or the presence of PD98059 as described for insulin release experiments, washed, and further incubated in groups of 100 islets for 10 min at 37 C in KRB supplemented with effectors as described in figure legends. After the 10-min incubation, islets were rapidly centrifuged and 1 μl/islet of cold lysis buffer [50 mM HEPES; 1% Nonidet-P40; 2 mM Na3VO4; 100 mM NaF; 10 mM PyrPO4; 4 mM EDTA; 1 mM phenylmethylsulfonyl fluoride (PMSF); 1 μg/ml leupeptin; 1 μg/ml aprotinin] was added. Islets were sonicated (10 sec), centrifuged at 14,000 rpm for 30 min at 4 C to remove insoluble materials, and supernatant were stored at –20 C until use for subsequent protein determination by Bradford assay and Western blotting.
For experiments with MIN6 cells, after a 2-h preincubation in KRB without glucose, 6-well plates MIN6 cells (70–80% confluence) were incubated in KRB containing various glucose concentrations as indicated in figure legends. Cells were then washed with ice-cold PBS, and lysis buffer (50 mM HEPES, 1 mM EDTA, 1% Triton X-100, 0.1% sodium dodecyl sulfate, 1 mM Na3VO4, 1 mM PMSF, 30 mM PyroPO4, 10 mM NaF, and 1 mg/ml bacitracin) was added. Cell lysates were centrifuged at 14,000 rpm for 30 min at 4 C to remove insoluble materials, and protein content of the supernatant was determined by Bradford assay. For immunoprecipitation, the supernatants (400–800 μg total protein) were incubated overnight at 4 C with the primary antibody. Immunocomplexes were precipitated from the supernatant with protein A/G-plus agarose for 4 h at 4 C and washed three times with ice-cold cell lysis buffer. For Western blot analysis, MIN6 cells, islet lysates (25–50 μg protein/lane), and immunoprecipitates were denatured by boiling (3 min) in Laemmli sample buffer containing 100 mM dithiothreitol and resolved by SDS-PAGE as previously described (30). Immunoprecipitation and Western blot analysis were carried out with an equal amount of total proteins. Visualization and quantification of the bands were obtained using an Image Station 2000R system (Eastman Kodak Co., Rochester, NY).
Subcellular fractionation
After a 2-h preincubation in KRB, MIN6 cells plated in 10-cm dishes were stimulated with glucose. Cells were washed twice with ice-cold PBS and scraped in 500 μl hypotonic buffer (10 mM HEPES; 10 mM NaCl; 1 mM KH2PO4; 5 mM NaHCO3; 1 mM CaCl2; 0.5 mM MgCl2; 5 mM EDTA; 1 mM PMSF; 10 μg/ml aprotinin; 10 μg/ml leupeptin; 1 μg/ml pepstatin). After a 15-min incubation, cells were dounced (pestle B) 50 times on ice and centrifuged for 5 min at 7500 rpm at 4 C. Supernatant containing cytosol and membranes was collected and preserved for protein content determination and Western blot analysis. The pellet containing nuclei was dounced (pestle B) 30 times in TSE buffer [10 mM Tris (pH 7.5), 300 mM sucrose, 1 mM EDTA (pH 8), 0.1% Nonidet-P40, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 1 μg/ml pepstatin], centrifuged for 5 min at 5000 rpm at 4 C, and rinsed twice in 1 ml TSE buffer. Final pellet containing pure nuclei was dissolved in TSE buffer for protein determination before denaturation in Laemmli buffer and analysis by SDS-PAGE.
Immunostaining for phospho-ERK1/2 in MIN6 cells
Cells were grown on glass coverslips for 3–5 d. Cells were then fixed in 3.7% formaldehyde in PBS for 30 min, washed with 0.1 M Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% Triton X-100 (TBST buffer), and permeabilized for 10 min in methanol at –20 C. After incubation in blocking buffer (5% normal horse serum in TBST buffer) for 1 h at room temperature, cells were exposed overnight at 4 C to the mouse anti-phospho-ERK1/2 antibody (1:400 dilution in 5% BSA TBST buffer). After washing, phospho-ERK1/2 antibody was detected by incubation for 1 h at room temperature with fluorescein isothiocyanate-conjugated donkey antimouse in TBST buffer containing 3% BSA (1:100 dilution). Coverslips were then mounted using a polyvinyl alcohol medium at least 1 h before observation using a dual-photon confocal microscope (Zeiss, Oberkocher, Germany).
Statistical analysis
Results are presented as mean ± SEM. Differences between results were analyzed by Student’s t test (unpaired unless indicated) (*, P < 0.05; **, P < 0.01; ***, P < 0.001).
Results
Increase in amount of activated forms of ERK1/2 in cytoplasm and nuclei of MIN6 cells on glucose stimulation
As a prelude to a study of the role of ERK1/2 activity in phosphorylation of cytoplasmic substrates and insulin secretion, we first attempted to confirm earlier reports by characterizing the effect of glucose on subcellular redistribution of active ERK1/2 using the MIN6 cells (13, 16). ERK1 and ERK2 were barely phosphorylated in MIN6 cells in the standard culture medium. When glucose was withdrawn for 2 h (2-h preincubation in KRB), activities of both enzymes were nearly undetectable (Fig. 1A). When 10 mM glucose was reintroduced, we found that both kinases were rapidly and transiently activated (Fig. 1A). Glucose dose-dependently increased ERK1/2 phosphorylation from 3 to 20 mM. Maximal activation of ERK2 was observed at 10–15 mM glucose (Fig. 1B). It is noteworthy that a noninsulinotropic glucose concentration (3 mM) induced a weak but significant phosphorylation of ERK1/2 (Fig. 1B). Because ERK1/2 activation elicited by glucose has been shown to be dependent on calcium influx through the opening of VDCCs (10, 13, 15, 16), we verified that the glucose-induced ERK1/2 activity was completely abolished when the cells were treated with the VDCC blocker nifedipine (Fig. 1C).
FIG. 1. Glucose transiently activates ERK1/2, which remains in large part in the cytoplasmic fraction. A–D, Western blot assay of activated and total ERK1/2 levels in MIN6 cells. Cells were preincubated in KRB for 2 h and either untreated or exposed to various concentration of glucose for different times as indicated. C, Effect of L-type calcium channel antagonist, nifedipine (2 μM), on ERK1/2 phosphorylation in MIN6 cells after treatment (5 min) with glucose. Nifedipine was added concomitantly with glucose treatment. D, Effect of PKA inhibitor, H89 (2 μM), on ERK1/2 phosphorylation in MIN6 cells after treatment (5 min) with glucose. H89 was added during the 2-h KRB preincubation period and maintained during the glucose treatment. E, Presence of active ERK1/2 in subcellular fractions from MIN6 cells exposed to glucose [nuclear (N) and cytosolic (C) fractions]. Cells were preincubated in KRB for 2 h and either untreated or stimulated (5 min) with glucose. A–E, Representative exposures of five experiments are shown. Total ERK1/2 levels within total cell lysates (A–D) and the different cellular fractions (E) are shown. F, Localization of the active form of ERK1/2 after glucose stimulation. Cells were preincubated in KRB for 2 h and either left untreated (a) or stimulated (5 min) with glucose (10 mM) (b).
Because it has been shown that cAMP and activation of PKA are involved in the mechanism of glucose activation of ERK1/2 in the INS-1 cell line (10), we also investigated the potential role of PKA in the ERK1/2 activation elicited by glucose in MIN6 cells. Cells were stimulated with glucose in either the absence or presence of the PKA inhibitor H89, at the minimal concentration shown to exert a complete suppression of the glucagon-induced ERK1/2 activation (31). As shown in Fig. 1D, H89 inhibited by 70% the glucose-stimulated ERK1/2 activation. The specificity of the H89 treatment was ascertained by the fact that it had no effect on epidermal growth factor-induced ERK1/2 phosphorylation (data not shown).
Because the subcellular localization plays an important role in determining the functions of ERK1 and ERK2 (22), we examined their distribution in subcellular fractions of MIN6 cells before and after exposure to glucose. In the absence of glucose, the immunoreactive ERK1 and ERK2 were associated with the cytoplasmic fraction and absent from the nuclear fraction (Fig. 1E). Treatment of cells with 10 mM glucose caused a substantial increase in the amount of ERK1 and ERK2 in the nuclear fraction, although a large part of active ERK1/2 remained in the cytoplasm. Immunolocalization using confocal microscopy showed an ERK1/2 staining of active forms throughout the cytoplasm as well as in the nucleus of cells exposed to 10 mM glucose (Fig. 1F), consistent with the results of subcellular fractionation.
Glucose-induced ERK1/2 activity regulates synapsin I phosphorylation
The majority of activated ERK1 and ERK2 remained in the MIN6 cells cytoplasm on glucose stimulation (Fig. 1, E and F). Therefore, we evaluated whether ERK1/2 activity elicited by glucose leads to phosphorylation of cytoplasmic substrate(s). In neurons, ERK1/2 serine-phosphorylate synapsin I leading to the recruitment of secretory vesicles to the plasma membrane for neurotransmitter exocytosis (32, 33, 34). Because it has been reported that synapsin I colocalizes with the insulin secretory granules in the MIN6 ? cells cytoplasm and that the phosphorylation/dephosphorylation cycle of synapsin I has been proposed to be implicated in insulin exocytosis (8, 9, 18), we sought to determine whether the ERK1/2 activity regulates the serine-phosphorylation of this cytoplasmic substrate on glucose stimulation. First, we used an immunoprecipitation approach to evaluate the existence of an association between synapsin I and ERK1/2 in the basal state and on glucose stimulation. Second, we measured the serine-phosphorylation level of synapsin I upon glucose stimulation, in either the absence or presence of PD98059, a selective inhibitor of MEK1/2, the upstream ERK1/2 kinase, which displays an IC50 of 2–4 μM and maximal inhibitory effects at 10–100 μM (35, 36). It must be noted that PD98059 is a compound with a highly selective profile, when used at reasonable doses, because no protein kinase has been found to be inhibited by PD98059 at a concentration (50 μM and below) that prevents activation of the ERK1/2 cascade (37). As a control, we verified that the PD98059 treatment completely abolished the ERK1/2 activation elicited by glucose (Fig. 2A). To rule out the possibility that PD98059 may be acting as a moderate metabolic poison on any step of glycolysis or Krebs cycle, thus reducing ATP synthesis and subsequent calcium influx, we verified that the PD98059 treatment did not inhibit the calcium influx induced by glucose (Fig. 2B).
FIG. 2. Glucose phosphorylates synapsin I through the ERK1/2 cascade. MIN6 cells were preincubated for 2 h in KRB as for Fig. 1. A, B, D, E, and F; PD98059 (20 μM) was present (or not) during 2-h KRB preincubation period and the glucose stimulation as indicated. A, MIN6 cells were either untreated or stimulated for 5 min with glucose and then lysed. A typical exposure representative of three experiments is shown. B, 45Ca2+ uptake was measured as described under experimental procedures. After 2 h preincubation in KRB, cells were incubated 5 min with glucose. Data are means ± SEM of four to six determinations. C, D, and E, MIN6 cells were either untreated or stimulated for 5 min with glucose (C and D) and 20 min (E) lysed and subjected to immunoprecipitation (IP) in reduced conditions with goat anti-synapsin I antibody. Immunoprecipitated proteins were blotted with anti-ERK1/2 (A), anti-phospho-ERK1/2 (B), or antiphosphoserine (E) antibodies. Representative exposures of five to six independent experiments are shown. Identical amounts of proteins were analyzed as described in Materials and Methods. As control, relative quantities of ERK1/2 (C and D) and synapsin I (E) in cell lysates are shown. F, Graph showing level of serine-phosphorylation of synapsin I as mean ± SEM of six independent immunoprecipitation experiments deduced from quantitative results obtained from analysis using Kodak Image Station 2000R system program. **, P < 0.01 vs. basal synapsin I phosphorylation.
The immunoprecipitation experiments with synapsin I antibody followed by immunoblotting for ERK1/2 showed no interaction between synapsin I and ERK1/2 in the basal state, whereas a marked association between synapsin I and ERK1/2 was observed after 5 min of glucose stimulation (Fig. 2C), suggesting that only the phosphorylated active form of ERK1/2 interacts with synapsin I. As a confirmation, we found that the phosphorylated form of ERK1/2 was physically associated with synapsin I on glucose stimulation and that this interaction was totally prevented by the PD98059 treatment (Fig. 2D). We next investigated whether the PD98059 treatment inhibited the glucose-induced synapsin I phosphorylation. As seen in Fig. 2, E and F, glucose induced a 2-fold increase in serine-phosphorylation of synapsin I through the ERK1/2 signaling cascade because this phosphorylation was totally prevented by the PD98059 treatment. Considering that the ERK1/2-mediated phosphorylation of synapsin I in neurons is essential for neurotransmitter exocytosis and that the phosphorylation/dephosphorylation cycle of synapsin I is implicated in the exocytosis of insulin secretory granules, we reexamined the potential role of ERK1/2 activity in glucose-induced insulin release.
ERK1/2 activity participates in the glucose-induced insulin secretion in MIN6 cells
We used the same protocol as described for ERK1/2 and synapsin I phosphorylation experiments and studied insulin release in MIN6 cells preincubated for 2 h in KRB. As shown in Fig. 3A, we found that the glucose-induced insulin release was inhibited by 70% when the ERK1/2 activity was totally blocked by PD98059 (Fig. 2A). The PD98059-mediated inhibition of glucose-induced insulin release was reversed when the inhibitor was removed from the glucose stimulation period. We observed that ERK1/2 activated by glucose translocate to the MIN6 cell nucleus (Fig. 1, E and F). Because it has been shown that the glucose-induced ERK1/2 activity controls the insulin gene transcription and that inhibition of protein synthesis by cycloheximide impairs insulin secretion (14, 23, 38), the ERK1/2 cascade may also participate in replenishing the pool of insulin granules. Insulin biosynthesis is a complex regulated process, and complementary methods exist to measure insulin biosynthesis directly. However, no change in the insulin content of MIN6 cells with or without the presence of PD98059 was observed, possibly excluding a defect in insulin biosynthesis (Fig. 3B). Moreover, because we observed that the glucose-induced insulin secretion was almost totally restored when PD98059 was removed during the glucose stimulation period (Fig. 3A), it is unlikely that the inhibition of glucose-induced insulin secretion by blocking ERK1/2 activity with PD98059 was related to a defect in protein synthesis or stability.
FIG. 3. ERK1/2 activity participates in glucose-induced insulin release in MIN6 cells. Cells were preincubated in KRB for 2 (A) or 1 h (C) and either untreated or stimulated with glucose. PD98059 was present (or not) during 2-h KRB preincubation period and stimulation (A and C). Data are means ± SEM of 12 determinations, each performed in triplicate. ***, P < 0.001 vs. glucose-induced insulin secretion. B, MIN6 cells were preincubated in KRB for 2 h in the presence or not of PD98059 and lysed in PBS + 1% Triton X-100 for 30 min. D, Cells were preincubated in KRB for 2 h and either untreated or stimulated (5 min) with glucose in the presence or absence of PD98059, lysed, and levels of ERK1/2 phosphorylation analyzed by Western blotting. A typical exposure of three experiments is shown.
We also tested the PD98059 treatment used by Khoo and Cobb (16) to investigate the role of ERK1/2 in the glucose-induced insulin release in INS-1 cells [e.g. cells were preincubated with 100 μM PD98059 for 1 h before addition of glucose, and PD98059 was present throughout the glucose stimulation period], which is quite close to our experimental conditions. On the contrary to what was observed in INS-1 (16), this PD98059 treatment condition blocked the glucose-induced insulin release by 53% in MIN6 cells (Fig. 3C). Again, PD98059-mediated inhibition of glucose-induced insulin release was reversed when the inhibitor was removed from the glucose stimulation period (Fig. 3C). Thus, it appears, from these results, that there is a substantial discrepancy between the data obtained from different ?-cell lines, MIN6 and INS-1. As control, we verified that this PD98059 treatment condition abolished the glucose-stimulated ERK1/2 activation (Fig. 3D).
ERK1 and ERK2 siRNA in MIN6 cells
To strengthen our data showing that ERK1/2 activity is required for glucose-stimulated insulin secretion using pharmacological blockade of ERK1/2 with PD98059 (Fig. 3, A and C), we performed a siRNA knockdown strategy to silence the expression of ERK1 and ERK2 mRNAs in MIN6 cells. We first found that transfection of MIN6 cells using modest amounts of ERK2 siRNA (50 nM) resulted in 100% knockdown of ERK2 and provoked MIN6 cell mortality (data not shown). Because it is well known that ERK2 is a critical cell cycle protein (21, 22, 39), these results raise the possibility that a total depletion of ERK2 in the MIN6 ?-cell might be lethal.
On the other hand, to avoid cell death under ERK2 siRNA, we performed ERK1 siRNA with 50 nM of siRNA duplexes. The ERK1 siRNA-transfected MIN6 cells were viable and no morphologic changes were observed (data not shown). As seen in Fig. 4A, ERK1 and ERK2 expressions were silenced by 70–80 and 50%, respectively, in ERK1 siRNA-transfected MIN6 cells, compared with nontransfected control cells. Therefore, we used this ERK1 siRNA-transfected MIN6 cell model, which phenotypically shows significant reduction of both ERK1 and ERK2 proteins, to evaluate the role of ERK1/2 pathway in insulin secretion stimulated by glucose. Immunoblot analysis of down-regulated MIN6 cells for ERK1 and ERK2 proteins revealed that glucose-induced ERK1 and ERK2 phosphorylation was largely inhibited (Fig. 4B). The specificity of our siRNA approach was ascertained by using ?-actin as internal and loading control (Fig. 4, A and B) or using other similarly sized ERK1/2-related siRNA duplexes that failed to induce any change in the expression of any of the mRNAs studied (ERK1, ERK2, or ?-actin) (data not shown). Interestingly, we found that glucose-stimulated insulin secretion was significantly reduced by 36% in ERK1 siRNA-transfected cells (Fig. 4C), clearly indicating that the ERK1/2 signaling pathway participates in insulin secretion in MIN6 ?-cells.
FIG. 4. ERK1/2 activation and insulin secretion defects in response to glucose in ERK1 and ERK2 siRNA transfected MIN6 cells. A, Western blot analysis of ERK1 and ERK2 expressions in nontransfected cells, incubated with transfection reagent alone as control (control reagent), and in ERK1 siRNA-transfected MIN6 cells. Blot was probed with ERK1 antibody and actin antibody used as specificity and loading control. A typical exposure representative of three independent experiments is shown. B, After transfection procedure, nontransfected cells (control), incubated with transfection reagent alone (control reagent), and ERK1 siRNA-transfected MIN6 cells were preincubated 2 h in KRB. For Western blot analysis, cells were then stimulated with glucose for 5 min, washed with ice-cold PBS, and lysed. Lysates (25 μg protein) were resolved by 10% SDS-PAGE in reduced conditions and blotted with phospho-ERK1/2, ERK1/2, or actin antibody. A typical exposure representative of two experiments is shown. As control, relative quantities of ERK1/2 and actin in cell lysates are also shown. C, For measuring insulin secretion, after transfection cells were preincubated for 2 h KRB and then either untreated or stimulated with glucose. Data are means ± SEM of nine determinations.
Insulinotropic glucose concentration induces phosphorylation of ERK1/2 in pancreatic islets
To confirm, using another model, the role of ERK1/2 in glucose-induced insulin secretion, we first investigated whether an insulinotropic glucose concentration induces phosphorylation of ERK1/2 in pancreatic islets. It has been reported that a 10-min stimulation by glucose had no effect on ERK1/2 activity in rat islets (24). Indeed, using the same experimental procedure (1-h stabilization in KRB supplemented with 2.8 mM glucose), we did not observe any activation of ERK1/2 by 16.7 mM glucose, compared with the basal (2.8 mM glucose) level of ERK1/2 phosphorylation, as reported by Burns et al. (24) (data not shown). Nevertheless, when stabilization was prolonged to at least 2 h (Fig. 5A), we found that, compared with the 1-h stabilized islets, basal ERK1/2 phosphorylation was reduced but not abolished and that 16.7 mM glucose stimulation of islets for 10 min resulted in a significant increase in the doubly phosphorylated active forms of these kinases (Fig. 5A). Pretreatment of islets with PD98059 during the 2-h stabilization period resulted in a deep inhibition of the 2.8 mM as well as the 16.7 mM glucose-induced ERK1/2 phosphorylation (Fig. 5A). Because ERK1/2 activation elicited by glucose is dependent on calcium influx through opening of VDCCs in MIN6 (Fig. 1C) and INS-1 (10, 16), we investigated whether the glucose-induced ERK1/2 activity in pancreatic islets was blocked when the cells were treated with the VDCC blocker nifedipine. As seen in Fig. 5B, nifedipine did not inhibit the basal ERK1/2 phosphorylation, whereas it blocked by 80% the 16.7 mM glucose-induced ERK1/2 phosphorylation. Thus, in pancreatic islets, the ERK1/2 activity elicited by an insulinotropic glucose concentration requires calcium influx via the opening of VDCCs, as observed in MIN6 cells (Fig. 1C) and as reported using another ?-cell line (10, 16). It is noteworthy that under the low, noninsulinotropic 2.8 mM glucose concentration, we found that ERK1/2 are phosphorylated within the islets independently from extracellular calcium entry.
FIG. 5. Glucose stimulates the ERK1/2 activity in isolated rat pancreatic islets. Islets were stabilized in KRB supplemented with glucose (2.8 mM) for 2 h and then incubated with 2.8 or 16.7 mM glucose for 10 min. A, Effect of PD98059 (20 μM) on ERK1/2 phosphorylation islets after treatment with glucose. PD98059 was present during the 2-h stabilization period and maintained during the 10-min glucose stimulation. B, Effect of nifedipine (2 μM) on ERK1/2 phosphorylation in islets after treatment with glucose. Nifedipine was added during the 10-min glucose stimulation period. A and B, The most representative blots obtained from three to five independent experiments are shown.
Role of ERK1/2 in the glucose-induced insulin secretion from islets of Langerhans
We next evaluated the potential involvement of ERK1/2 signaling cascade in glucose-induced insulin release from isolated pancreatic islets. We first examined the effect of PD98059 treatment on glucose-induced insulin release during a static incubation. As shown in Fig. 6A, 2.8 and 16.7 mM glucose-induced insulin releases were reduced by 60 and 62%, respectively when ERK1/2 was blocked by PD98059 treatment. We verified that a PD98059 treatment had no effect on insulin content of the islets (Fig. 6B) and that the VDCC blocker nifedipine did not inhibit the 2.8 mM glucose-basal secretion, whereas it was efficient in blocking the 16.7 mM glucose-induced insulin secretion (Fig. 6A). These data demonstrate that under low, noninsulinotropic (2.8 mM) glucose concentration, ERK1 and ERK2 are active within the islets independently from extracellular calcium entry and are involved in basal insulin secretion, whereas an insulinotropic (16.7 mM) glucose concentration, there is a significant increase in the phosphorylated active forms of ERK1/2, induced via calcium influx, which participate in the glucose-induced insulin release.
FIG. 6. Glucose-induced insulin secretion from isolated rat islets requires the ERK1/2 activity. A, Islets were stabilized for 2 h in KRB supplemented with glucose (2.8 mM) and then stimulated with glucose (2.8 or 16.7 mM) for 10 min. Islets were treated with PD98059 (20 μM) or nifedipine (2 μM) as for Fig. 5. Data are means ± SEM of 13 determinations, each performed in duplicate. *, P < 0.05 vs. glucose-induced insulin secretion. B, Insulin content determination as described in Materials and Methods.
ERK1/2 participate in both phases of glucose-stimulated insulin secretion
Because static incubation fails to reveal dynamics of insulin release in terms of biphasic secretion, we determined the impact of PD98059 treatment on both phases of glucose-stimulated insulin secretion using the pancreatic islet perifusion model. Isolated islets were stabilized for 2 h in the absence or the presence of PD98059 and then transferred to the perifusion chambers for an additional 20-min perifusion of 2.8 mM glucose (one chamber containing untreated control islets, the other chamber containing the PD98059-treated islets), followed by 16.7 mM glucose stimulation for 30 min in the absence (open circles) or presence of PD98059 (filled squares). As seen in Fig. 7A, basal insulin secretion was inhibited during the initial 20-min perifusion period with PD98059, and the 16.7 mM glucose-induced insulin release was significantly reduced in the PD98059-treated islets. This marked inhibition of 16.7 mM glucose-induced insulin secretion occurred throughout the time course and was present during both the first and second phases of insulin release. The subsequent removal of 16.7 mM glucose resulted in a sharp decrease in insulin secretion near to basal levels within 10–20 min, in the absence or presence of PD98059. Insulin released during the first phase, as measured by evaluation of the area under the curve (AUC) calculated classically, from the initial insulin rise (min 3 after the beginning of the glucose stimulation) to the lowest point between the first and second rise in insulin release (min 7), was inhibited by about 55% (Fig. 7B). The amount of insulin secreted during the second phase, starting from min 8 to the end of stimulation, was inhibited by 38% (Fig. 7C). Because the basal (2.8 mM glucose) secretion was altered by the PD98059 treatment, we calculated the AUC with the average of the –10 to 0 min as the baseline. The results indicate 27 (nonsignificant) and 49% (significant) inhibitions for the first and second phases, respectively (Fig. 7, D and E). We can conclude that ERK1/2 activity participates in the basal as well as first and second phases of glucose-induced insulin release in rat pancreatic islets.
FIG. 7. Role of ERK1/2 activity in the two phases of the glucose-induced insulin secretion from isolated rat islets. A, Islets were stabilized for 2 h in KRB with glucose (2.8 mM) either in the absence or presence of PD98059 and then transferred to perifusion chambers in water bath at 37 C. First, islets were perifused for 20 min with KRB containing 2.8 mM glucose. At time 0, glucose was raised to 16.7 mM for 30 min and then lowered to 2.8 mM for 20 min. PD98059 was maintained from the beginning of the stabilization to the end of stimulation (filled squares, PD98059-treated islets). Data are mean ± SEM. Each perifusion was repeated three times and insulin secretion was quantified in duplicate. *, P < 0.05; **, P < 0.01 (paired Student’s t test, the pairs being the two chambers of the same water bath). B and C, Insulin secretion was calculated as AUC. D and E, Insulin secretion was calculated as AUC with the average of the –10 to 0 min as the baseline. *, P < 0.05 (paired Student’s t test, the pairs being the two chambers of the same water bath).
Discussion
In this report, we show that glucose-activated ERK1/2 remain in large part in the ?-cell cytoplasm. Supporting the concept that the ERK1/2 cascade is a significant regulator of glucose-dependent cytoplasmic events, we demonstrate that ERK1/2 participate in the glucose-mediated phosphorylation of synapsin I, a protein known to be associated with insulin granules and proposed to be involved in the granule translocation that precedes insulin exocytosis (8, 9, 20, 32, 33, 34). Moreover, using MIN6 cells and perifused isolated pancreatic islets, we provide evidence that ERK1/2 activity also contributes to optimal glucose-induced insulin secretion. Therefore, we propose that the glucose-induced activation of the ERK1/2 signaling pathway in ?-cells participates in the amplification of the glucose action on insulin secretion by controlling activation of one or several cytosolic proteins including synapsin I.
In both MIN6 cells and isolated islets, we confirm that glucose induces phosphorylation of ERK1/2 via calcium entry and PKA activation. As reported by Briaud et al. in INS-1 cells (10), it is probable that glucose-induced rise in cytoplasmic [Ca2+] leads to cAMP production and subsequent PKA activation that mediate in part ERK1/2 activation in MIN6 cells. The majority of activated forms of ERK1/2 upon the glucose challenge remained in the cytoplasm interacting with synapsin I and, possibly, with other protein(s) involved in exocytosis. In the basal state, we demonstrate that ERK1/2 do not interact with synapsin I. On the other hand, we observed a physical interaction between synapsin I and the activated forms of ERK1/2 on glucose stimulation, leading to synapsin I phosphorylation. It has already been demonstrated that synapsin I is phosphorylated by PKA and CAMKII in glucose-stimulated ?-cells (18, 20). Our conditions (20-min glucose stimulation in the absence of any phosphatase inhibitor) allow us to specifically detect the serine-phosphorylated state of synapsin I, which is not dependent on PKA nor CAMKII, because the former is quickly reversed by dephosphorylation by calcineurin (18), whereas the latter returns to the basal value within 15 min (20). Thus, it is not surprising that the PD98059 treatment totally prevented the serine-phosphorylation of synapsin I. Phosphorylation of ERK1/2-specific sites of synapsin I in ?-cells appears to be more sustained than other phosphorylation sites, as shown in other cell types (32). The rapid and transient activation of ERK1/2 by glucose in ?-cells is sufficient to induce a sustained phosphorylation of synapsin I.
Another major finding of this paper is that, in both MIN6 cells and isolated islets, the glucose-induced ERK1/2 activation is required for an optimal glucose-induced insulin secretion because the observed inhibition of the glucose-induced insulin secretion when ERK1/2 activity was blocked by a 2-h treatment with PD98059 was not due to a defect in insulin content. In addition, the rapid reversibility of the inhibition indicates that the effect is not related to a defect in protein synthesis or disappearance of short half-life proteins. The results of the siRNA-mediated ERK1/2 partial knockdown fully confirmed those obtained using pharmacological blockade. However, we cannot rule out the possibility that the glucose-induced insulin release observed in ERK1 siRNA-transfected cells may result not only from a dysfunction of phosphorylation/activation of synapsin I and/or other unknown protein(s) of the insulin exocytotic process but also from a modification in the insulin gene transcription. It must also be noted that 100% knockdown of ERK2 provoked MIN6 cell mortality raising the possibility that ERK2 is a critical cell cycle protein in pancreatic ?-cells.
Using INS-1 cells, Khoo and Cobb (16) reported that glucose-induced insulin release occurs normally, even if the ERK1/2 activity was blocked by PD98059. Here, using MIN6 cells, we show that the same PD98059 treatment inhibited the glucose-induced insulin secretion. Thus, there is an apparent discrepancy with the data obtained from different cell lines. This discrepancy could be attributed to their differences in glucose sensitivity. Indeed, it is noteworthy that INS-1 cells from rats and MIN6 cells from mice display different insulin secretion profiles (40, 41, 42). Therefore, differences in the glucose metabolism, a key step in promoting insulin release, or subtle differences in the recruitment of intracellular molecular mechanisms responsible for the generation of the first- and second-phase insulin release could explain in part the observed discrepancy between the cellular models.
Using static incubation of pancreatic islets, we confirmed the involvement of the ERK1/2 signaling cascade on insulin release on a 10-min 16.7 mM glucose challenge. However, we observed that insulin release became insensitive to the PD98059 treatment with increasing times of static glucose stimulation (30–90 min) and that long-lasting (90 min) stimulation of insulin release by 16.7 mM glucose was not significantly affected by PD98059 (data not shown), which is fully consistent with the work of Burns et al. (24). It is well known that local interactions between the cells in the islet exert an important level of control in the physiological regulation of insulin secretion. It is likely that, during static incubation, accumulated insulin within the islet might suppress the secretion of pancreatic peptides, such as miniglucagon from -cells or somatostatin from -cells which have been both reported to be efficient inhibitors of insulin secretion (26, 43, 44). The suppression of an inhibitory tone on ?-cell due to miniglucagon and/or somatostatin, within the islet, might further enhance the insulinotropic glucose effect and might render negligible over the time the fraction of the glucose-induced insulin secretion due to ERK1/2. Because insulin receptors are expressed and functional in ?-cells, it has been reported that insulin regulates its own secretion in a regulatory feedback loop (45, 46). Therefore, accumulated insulin within the islet could also modulate its own secretion through mobilization of parallel mechanisms, implying other kinases, interfering with the inhibition of ERK1/2 signaling pathway. Unlike the results observed in static incubations, in the islets perifusion experiments, in which there is no accumulation of islet peptides at the vicinity of the ?-cells, PD98059 treatment exerted a persistent inhibition of the glucose-induced insulin release throughout the 30-min glucose stimulation period.
Additionally, whatever the calculation method used for measuring AUC in the perifusion experiments, our data suggest that ERK1/2 activity participates in the basal as well as the first and the second phases of the glucose-induced insulin secretion. These observations are compatible with the concept that ERK1/2 activated by glucose participate in the regulation of cytoskeleton recast and exocytotic machinery for optimal basal and stimulated insulin secretion through phosphorylation of proteins such as synapsin I. The precise role of ERK1/2 activity in the basal and stimulated insulin secretion from MIN6 cells and islets deserves now to be investigated in more detail.
In addition, glucagon-like peptide-1 and glucose-dependent insulinotropic peptide activate, in ?-cells, the ERK1/2 signaling cascade that regulates cell survival, proliferation, and insulin gene transcription (10, 28, 47, 48, 49). Our data suggest that ERK1/2 activation by incretin peptides may be an additional mechanism that regulates insulin granule exocytosis through phosphorylation of cytosolic protein(s).
In summary, our data provide new insights in the regulation by glucose of the ?-cell physiology. We demonstrate that ERK1/2, among others kinases such as PKA or CAMKII, represent an additional transduction signal that amplifies the glucose effect on insulin secretion. Hyperglycemic exposure to the ?-cells, which is a key feature of type 2 diabetes, has been reported to contribute to the deterioration of ?-cell function, a concept referred to as glucotoxicity. It is now interesting to study the potential role of the ERK1/2 signaling pathway in the progressive loss of ?-cell insulin secretion due to long-term hyperglycemia throughout the course of type 2 diabetes. Recently it has been reported that IL-1?-stimulated ERK1/2 activation was a mediator of the long-term deleterious effects of glucose on ?-cell function (50).
In conclusion, our results establish that the ERK1/2 activity is necessary for an optimal glucose-induced insulin secretion, a discovery that has now to be taken into account for understanding in more detail the molecular mechanisms the impairment of which leads to ?-cell defects involved in type 2 diabetes.
References
Docherty K, Steiner DF 1996 Molecular and cellular biology of the ? cell. In: Portes D, Sherwin R, eds. Ellenberg Rifkin’s diabetes mellitus. 5th ed. Stamford, CT: Elsevier Science Publishing; 29–48
Matschinsky FM 1996 Banting Lecture 1995. A lesson in metabolic regulation inspired by the glucokinase glucose sensor paradigm. Diabetes 45:223–241
Henquin JC, Ishiyama N, Nenquin M, Ravier MA, Jonas JC 2002 Signals and pools underlying biphasic insulin secretion. Diabetes 51(Suppl 1):S60–S67
Barg S, Eliasson L, Renstr?m E, Rorsman P 2002 A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse ?-cell. Diabetes 51(Suppl 1):74–82
Ohara-Imaizumi M, Nakamichi Y, Tanaka T, Ishida H, Nagamatsu S 2002 Imaging exocytosis of single insulin secretory granules with evanescent wave microscopy: distinct behavior of granule motion in biphasic insulin release. J Biol Chem 277:3805–3808
Somers G, Blondel B, Orci L, Malaisse WJ 1979 Motile events in pancreatic endocrine cells. Endocrinology 104:255–264
Li G, Rungger-Brandle E, Just I, Jonas JC, Aktories K, Wollheim CB 1994 Effect of disruption of actin filament by clostridium botulinum C2 toxin on insulin secretion in HIT-T15 cells and pancreatic islet. Mol Biol Cell 5:1199–1213
Matsumoto K, Kukunaga K, Miyazaki JI, Shichiri M, Miyamoto E 1995 Ca2+/calmodulin-dependent protein kinase II and synapsin I-like protein in mouse insulinoma MIN6 cells. Endocrinology 136:3784–3793
Matsumoto K, Ebihara K, Yamamoto H, Tabuchi H, Fukunaga K, Yasunami M, Ohkubo H, Shichiri M, Miyamoto E 1999 Cloning from insulinoma cells of synapsin I associated with insulin secretory granules. J Biol Chem 274:2053–2059
Briaud I, Lingohr MK, Dickson LM, Wrede CE, Rhodes CJ 2003 Differential activation mechanisms of Erk-1/2 and p70(S6K) by glucose in pancreatic ?-cells. Diabetes 52:974–983
Wenham RM, Landt M, Easom RA 1994 Glucose activates the multifunctional calcium/calmodulin dependent protein kinase II in isolated rat pancreatic islets. J Biol Chem 269:4947–4952
Knutson KL, Hoenig M 1994 Identification and subcellular characterization of protein kinase C isoforms in insulinoma ?-cells and whole islets. Endocrinology 135:881–886
Benes C, Roisin MP, Van Tan H, Creuzet C, Miyazaki J, Fagard R 1998 Rapid activation and nuclear translocation of mitogen-activated protein kinases in response to physiological concentration of glucose in the MIN6 pancreatic ? cell line. J Biol Chem 273:15507–15513
Benes C, Poitout V, Marie JC, Martin-Perez J, Roisin MP, Fagard R 1999 Mode of regulation of the extracellular signal-regulated kinases in the pancreatic ?-cell line MIN6 and their implication in the regulation of insulin gene transcription. Biochem J 340:219–225
Fr?din M, Sekine N, Roche E, Filloux C, Prentki M, Wolheim CB, Van Obberghen E 1995 Glucose, other secretagogues, and nerve growth factor stimulate mitogen-activated protein kinase in the insulin secreting ?-cell line, INS-1. J Biol Chem 270:7882–7889
Khoo S, Cobb M 1997 Activation of mitogen-activating protein kinase by glucose is not required for insulin secretion. Proc Natl Acad Sci USA 94:5599–5604
Jones P, Persaud S 1998 Protein kinases, protein phosphorylation and the regulation of insulin secretion from pancreatic ?-cells. Endocr Rev 19:429–461
Lester LB, Faux MC, Nauert JB, Scott JD 2001 Targeted protein kinase A and PP-2B regulate insulin secretion through reversible phosphorylation. Endocrinology 142:1218–1227
Mendez CF, Leibiger IB, Leibiger B, Hoy M, Gromada J, Berggren PO, Bertorello AM 2004 Rapid association of protein kinase C with insulin granules is essential for insulin exocytosis. J Biol Chem 278:44753–44757
Yamamoto H, Matsumoto K, Araki E, Miyamoto E 2003 New aspects of neurotransmitter release and exocytosis: involvement of Ca(2+)/calmodulin-dependent phosphorylation of synapsin I in insulin exocytosis. J Pharmacol Sci 93:30–34
Hunter T 1995 Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling. Cell 80:225–236
Volmat V, Pouyssegur J 2001 Spatiotemporal regulation of the p42/p44 MAPK pathway. Biol Cell 93:71–79
Khoo S, Griffen SC, Xia Y, Baer RJ, German MS, Cobb MH 2003 Regulation of insulin gene transcription by extracellular-signal regulated protein kinases (ERK) 1 and 2 in pancreatic ? cells. J Biol Chem 278:32969–32977
Burns CJ, Howell SL, Jones PM, Persaud SJ 1997 Glucose-stimulated insulin secretion from rat islets of Langerhans is independent of mitogen-activated protein kinase activation. Biochem Biophys Res Commun 239:447–450
Lacy PE, Kostianovsky M 1967 Method for the isolation of intact islets of Langerhans from the rat pancreas. Diabetes 16:35–39
Dalle S, Smith P, Blache P, Le-Nguyen D, Le Brigand L, Bergeron F, Ashcroft FM, Bataille D 1999 Miniglucagon (glucagon 19–29), a potent and efficient inhibitor of secretagogue-induced insulin release through a Ca2+ pathway. J Biol Chem 274:10869–10876
Jhala US, Canettieri G, Screaton RA, Kulkarni RN, Krajewski S, Reed J, Walker J, Lin X, White M, Montminy M 2003 cAMP promotes pancreatic ?-cell survival via CREB-mediated induction of IRS2. Genes Dev 17: 1575–1580
Gomez E, Pritchard C, Herbert TP 2002 cAMP-dependent protein kinase and Ca2+ influx through L-type voltage-gated calcium channels mediate Raf-independent activation of extracellular regulated kinase in response to glucagon-like peptide-1 in pancreatic ?-cells. J Biol Chem 277:48146–48151
Kervran A, Rieutort M, Guillaume M 1976 A simultaneous radioimmunoassay for growth hormone and insulin in the plasma of rats and rabbits. Diabetes Metab 2:67–72
Dalle S, Imamura T, Rose DW, Worrall DS, Ugi S, Hupfeld CJ, Olefsky JM 2002 Insulin induces heterologous desensitization of G-protein-coupled receptor and insulin-like growth factor I signaling by downregulating ?-arrestin-1. Mol Cell Biol 22:6272–6285
Dalle S, Longuet C, Costes S, Broca C, Faruque O, Fontes G, el Hani H, Bataille D 2004 Glucagon promotes cAMP-response element-binding protein phosphorylation via activation of ERK1/2 in MIN6 cell line and isolated islets of Langerhans. J Biol Chem 279:20345–20355
Matsubara M, Kusubata M, Ishiguroi K, Uchidai T, Titani K, Taniguchi H 1996 Site-specific phosphorylation of synapsin I by mitogen-activated protein kinase and Cdk5 and its effects on physiological functions. J Biol Chem 271:21108–21113
Jovanovic JN, Czernik AJ, Fienberg AA, Greengard P, Sihra TS 2000 Synapsins as mediators of BDNF-enhanced neurotransmitter release. Nat Neurosci 3:323–329
Jovanovic JN, Benfenati F, Siow YL, Sihra TS, Sanghera JS, Pelech SL, Greengard P, Czernik AJ 1996 Neurotrophins stimulate phosphorylation of synapsin I by MAP kinase and regulate synapsin I-actin interactions. Proc Natl Acad Sci USA 93:3679–3683
Gould GW, Cuenda A, Thomson FJ, Cohen P 1995 The activation of distinct mitogen-activated protein kinase cascades is required for the stimulation of 2-deoxyglucose uptake by interleukin-1 and insulin-like growth factor-1 in KB cells. Biochem J 311:735–738
Pang L, Sawada T, Decker SJ, Saltiel AR 1995 Inhibition of MAP kinase kinase blocks the differentiation of PC-12 cells induced by nerve growth factor. J Biol Chem 270:13585–13588
Davies SP, Reddy H, Caivano M, Cohen P 2000 Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351:96–105
Garcia-Barrado MJ, Ravier MA, Rolland JF, Gilon P, Nenquin M, Henquin JC 2001 Inhibition of protein synthesis sequentially impairs distinct steps of stimulus-secretion coupling in pancreatic ? cells. Endocrinology 142:299–307
Squires MS, Nixon PM, Cook SJ 2002 Cell-cycle arrest by PD184352 requires inhibition of extracellular signal-regulated kinases (ERK) 1/2 but not ERK5/BMK1. Biochem J 366:673–680
Asfari M, Janjic D, Meda P, Li G, Halban PA, Wollheim CB 1992 Establishment of 2-mercaptoethanol-dependent differentiated insulin-secreting cell lines. Endocrinology 130:167–178
Ishihara H, Asano T, Tsukuda K, Katagiri H, Inukai K, Anai M, Kikuchi, M, Yazaki Y, Miyazaki J, Oka Y 1994 Overexpression of hexokinase I but not GLUT1 glucose transporter alters concentration dependence of glucose-stimulated insulin secretion in pancreatic ?-cell line MIN6. J Biol Chem 269:3081–3087
Poitout V, Olson LK, Robertson RP 1996 Insulin-secreting cell lines: classification, characteristics and potential applications. Diabetes Metab 22:7–14
Dalle S, Fontes G, Lajoix AD, LeBrigand L, Gross R, Ribes G, Dufour M, Barry L, LeNguyen D, Bataille D 2002 Miniglucagon (glucagon 19–29): a novel regulator of pancreatic islet physiology. Diabetes 51:406–412
Sharp GW 1996 Mechanisms of inhibition of insulin release. Am J Physiol 271:1781–1799
Aspinwall CA, Qian WJ, Roper MG, Kulkarni RN, Kahn CR, Kennedy RT 2000 Roles of insulin receptor substrate-1, phosphatidylinositol 3-kinase, and release of intracellular Ca2+ stores in insulin-stimulated insulin secretion in ?-cells. J Biol Chem 275:22311–22338
Leibiger IB, Leibiger B, Berggren PO 2002 Insulin feedback action on pancreatic ?-cell function. FEBS Lett 532:1–6
Brubaker PL, Drucker DJ 2004 Glucagon-like peptides regulate cell proliferation and apoptosis in the pancreas, gut, and central nervous system. Endocrinology 145:2653–2659
Trümper A, Trümper K, H?rsh D 2002 Mechanisms of mitogenic and anti-apoptotic signaling by glucose-dependent insulinotropic polypeptide in ?(INS-1)-cells. J Endocrinol 174:233–246
Ehses JA, Pelech SL, Pederson RA, McIntosh CH 2002 Glucose-dependent insulinotropic polypeptide activates the Raf-Mek1/2-ERK1/2 module via a cyclic AMP/cAMP-dependent protein kinase/Rap1-mediated pathway. J Biol Chem 277:37088–37097
Maedler K, Storling J, Sturis J, Zuellig RA, Spinas GA, Arkhammar POG, Mandrup-Poulsen T, Donath MY 2004 Glucose- and interleukin-1?-induced ?-cell apoptosis requires Ca2+ influx and extracellular signal-regulated kinase (ERK)1/2 activation and is prevented by a sulfonylurea receptor1/inwardly rectifying K+ channel 6.2 (SUR/KIR6.2) selective potassium channel opener in human islets. Diabetes 53:1706–1713(Christine Longuet, Christ)
Address all correspondence and requests for reprints to: Dr. Stéphane Dalle, Unité Institut National de la Santé et de la Recherche Médecale U376, Centre Hospitalier Universitaire Arnaud de Villeneuve, 371 Rue du Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France. E-mail: dalle@montp.inserm.fr.
Abstract
The p44/p42 MAPKs (ERK1/2) cascade regulates ?-cell nuclear events, which modulates cell differentiation and gene transcription, whereas its implication in processes occurring in the cytoplasm, such as activation of the exocytotic machinery, is still unclear. Using the MIN6 ?-cell line and isolated rat islets of Langerhans, we investigated whether glucose, by activating the ERK1/2 cascade, induces phosphorylation of cytoplasmic proteins implicated in exocytosis of insulin granules such as synapsin I. We observed that the majority of ERK1/2 activity induced by glucose remains in the cytoplasm and physically interacts with synapsin I, allowing phosphorylation of the substrate. Therefore, we reexamined the potential requirement of ERK1/2 for insulin secretion. Blocking activation of ERK1/2 using MEK1/2, the MAPK kinase inhibitor PD98059 or using small interfering RNA-mediated silencing of ERK1 and ERK2 expressions resulted in partial inhibition of glucose-induced insulin release, indicating that ERK1/2 pathway participates also in the regulation of insulin secretion. Moreover, using the pancreatic islet perifusion model, we found that the ERK1/2 activity participates in the first and second phases of insulin release induced by glucose. Taken together, our results demonstrate new aspects of the glucose-dependent actions of ERK1/2 in ?-cells exerted on cytoplasmic proteins, including synapsin I, and participating in the overall glucose-induced insulin secretion.
Introduction
INCREASING CONCENTRATIONS OF circulating glucose, in the 3–20 mM range, stimulate insulin secretion by generating and amplifying signals in pancreatic ?-cells (1, 2, 3). After transportation into the ?-cell by an insulin-insensitive facilitated transporter, glucose is metabolized, producing ATP (1, 2, 3). The increase in the ATP to ADP ratio induces the closure of ATP-dependent potassium channels, which leads to membrane depolarization, promoting opening of L-type voltage-dependent calcium channels (VDCCs) (1, 2, 3). The resulting calcium entry generates a rapid and sustained rise in cytoplasmic-free calcium concentrations [cytoplasmic (Ca2+)], which trigger the biphasic insulin release (1, 2, 3). The first phase requires a rapid and marked elevation of cytoplasmic [Ca2+] and corresponds to exocytosis of a limited readily releasable pool of insulin secretory granules docked to the plasma membrane (1, 2, 3, 4). The sustained second phase of secretion requires signal(s) additional to the cytoplasmic [Ca2+] rise and corresponds to the recruitment to the plasma membrane of a reserve pool of secretory granules by activation/rearrangement of the components of the exocytotic machinery (1, 2, 3, 4, 5, 6, 7, 8, 9). However, the amplifying signaling pathways, which increase the efficiency of calcium on insulin exocytosis, remain in a large part unidentified (1, 2, 3).
Glucose-induced calcium entry has been reported to activate various kinases in the ?-cells, including protein kinase A (PKA) via calcium influx-dependent cAMP production (10), calmodulin kinase II (CAMKII) (11), protein kinase C (12), and p44/42 MAPKs (also called ERK1/2) (10, 13, 14, 15, 16, 17). PKA, protein kinase C, and CAMKII have been shown to be involved in the glucose-induced insulin secretion (18, 19, 20). ERK1/2 are the terminal kinases in a ubiquitous and pleiotropic three-kinase cascade, consisting of Raf isoforms that activate MEK1/2, MAPK kinase which, in turn, activate ERK1 and ERK2. ERK1/2 have been implicated in many cellular events, including proliferation, differentiation, survival, and secretion (21, 22). In the ?-cells, glucose-activated ERK1/2 have been described, so far, to regulate insulin gene transcription (14, 23).
Over its physiological concentration range, glucose activates ERK1/2 in pancreatic ?-cells (10, 13, 14, 15, 16, 17). Signals from extracellular calcium entry, cAMP production, and intracellular calcium stores were suggested to mediate the glucose effects on ERK1/2 (10, 13, 14, 15, 16, 17), besides those observed on insulin release. Thus, increased ERK1/2 activity by glucose may correlate with insulin secretion. Nevertheless, studies have shown a lack of relationship between ERK1/2 activity and insulin release (13, 16, 24). On the other hand, using the insulinoma (INS)-1 ?-cell line, it has been shown that the glucose-activated ERK1/2 partially translocate to purified nuclear fractions, whereas the majority of active ERK1/2 was found in the cytoplasm of glucose-stimulated cells (16). Thus, it is possible that ERK1/2 also regulate glucose-dependent cytoplasmic events involved in the ?-cell biology. To date, no cytoplasmic function for ERK1/2 has been reported in ?-cells.
To address this issue, we used the MIN6 cells and isolated rat pancreatic islets to examine the potential requirement of ERK1/2 activity in phosphorylation and activation of cytoplasmic targets, and we reexamined the potential requirement of ERK1/2 for insulin secretion. We studied synapsin I, a well-known ERK1/2 substrate in neurons, present in the ?-cell cytoplasm and expressed in normal rat islets (8, 9). Synapsin I is associated with the insulin-secretory granules in MIN6 cells (9), and phosphorylation of synapsin I by PKA and CAMKII has been proposed to be implicated in the glucose-induced insulin release (8, 9, 18, 20). We observed that, upon glucose stimulation, ERK1/2 associate with synapsin I and that the glucose-induced ERK1/2 activity leads to phosphorylation of synapsin I. On the other hand, we found that glucose-induced insulin release from MIN6 cells and isolated rat pancreatic islets is partially dependent on ERK1/2 activity. Thus, our data demonstrate that ERK1/2 activity is a significant regulator of glucose-dependent cytoplasmic events, including phosphorylation of synapsin I, and suggest, for the first time, a significant involvement of ERK1/2 in the insulin secretory process.
Materials and Methods
Animals
Male Wistar rats (Iffa-Credo, l’Arbresle, France) had free access to laboratory chow and water. They were housed, handled, and killed according to the rules of the Centre National de la Recherche Scientifique Animal Care and Use Committee.
Materials
Mouse anti-ERK1/2 antibody was obtained from Transduction Laboratories (BD Biosciences Pharmingen, San Diego, CA). Mouse anti-phospho-ERK1/2 (p44/42 MAPK) antibody, which selectively recognizes the doubly phosphorylated, active forms of these kinases, and rabbit anti-synapsin I antibody were purchased from Cell Signaling (New England Biolabs, Beverly, MA). Goat anti-synapsin I antibody, horseradish peroxidase-linked donkey antirabbit, or antimouse antibodies, and protein A/G-plus agarose were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse antiphosphoserine antibody was obtained from Abcam (Cambridge, UK). Fluorescein isothiocyanate-conjugated donkey antimouse antibody was purchased from Vector Laboratories (Burlingame, CA). DMEM and fetal calf serum (FCS) were purchased from Invitrogen Life Technologies (Grand Island, NY). Nitrocellulose transfer membranes (Protran) were obtained from Schleicher & Schuell (Dassel, Germany). PD98059 was purchased from Calbiochem (La Jolla, CA). 45CaCl2 (2 mCi/ml) was from Amersham Pharmacia Biotech (Amersham, UK). All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO).
Pancreatic islets preparation and MIN6 cells culture
Pancreatic islets were isolated from fed male Wistar rats weighing 280–320 g on the day the rats were killed. Islets were isolated by collagenase digestion followed by Ficoll gradient separation, with a method adapted from (25). The MIN6 cell line (passages 18–24) was maintained as described previously (26, 27, 28). Cultures were never allowed to become completely confluent. MIN6 cells used in the experiments were tested for the dose-dependent insulin response to glucose, which always was in the physiological range (26).
Insulin secretion from static incubation of isolated pancreatic islets
Isolated islets were stabilized for 2 h at 37 C in HEPES-balanced Krebs-Ringer buffer (KRB) [119 mM NaCl; 4 mM KCl; 1.2 mM KH2PO4; 1.2 mM MgSO4; 2.5 mM CaCl2; 20 mM HEPES, pH 7.2)] containing 0.1% BSA (KRB) and 2.8 mM glucose and then stimulated (3 islets/tube) by 16.7 mM glucose in static incubation. When used, nifedipine (2 μM) was added during the 2.8- and 16.7-mM glucose stimulation periods, whereas PD98059 (20 μM) was added from the beginning of the stabilization period and maintained until the end of the stimulation. To prevent an effect related to the toxicity of the solvent-carrier dimethyl sulfoxide, control islets were always treated with dimethyl sulfoxide in the same conditions of concentration and time as for the PD98059-treated islets. At the end of stimulation, islets were pelleted by centrifugation and lysed in acid-ethanol for assessment of insulin content. Insulin secretion was quantified by RIA using [125I]porcine insulin, rat insulin (Novo Nordisk, Copenhagen, Denmark) as standard, and the guinea pig antiporcine insulin antibody 41 (29). Results are presented as insulin secreted (nanograms per milliliter) normalized to islet insulin content extracted with acid-ethanol overnight at –20 C.
Perifusion of isolated pancreatic islets
For perifusion experiments, after a 2-h static incubation in KRB containing 2.8 mM glucose in the presence or the absence of PD98059, batches of 30 islets were placed in perifusion chambers (two chambers per bath) at 37 C and perifused at a flow rate of 1.5 ml/min with KRB containing various concentrations of glucose as indicated in the presence or the absence of PD98059. Islets were equilibrated for 20 min in KRB supplemented with 2.8 mM glucose and then stimulated for 30 min with 16.7 mM glucose. The 16.7-mM glucose stimulation period was followed by a 20-min perifusion in 2.8 mM glucose.
Insulin secretion from MIN6 cells
MIN6 cells were plated in 24-well plates at a density of 1 x 106 cells/well for 3–5 d. Insulin secretion studies were performed using a static incubation method in 5% CO2 at 37 C with cells (70–80% confluence) as described previously (26). For the experiments with PD98059, MIN6 cells were preincubated for 2 h in KRB in the absence or presence of PD98059 before stimulation of insulin secretion with 10 mM glucose, and PD98059 was present throughout the incubation period unless otherwise indicated. For determination of insulin content, MIN6 cells were lysed for 30 min at 4 C in PBS + 1% Triton X-100.
Measurement of 45Ca2+ influx
MIN6 cells were plated in 24-well plates at a density of 1 x 106 cells/well for 3–5 d. Twenty-four hours before the experiments, the culture medium was changed. On the day of the experiment, the cells were preincubated for 2 h in 500 μl KRB in 5% CO2 at 37 C. The preincubation solution was then replaced by 250 μl KRB containing 8 μCi/ml 45CaCl2 and the test agents. The reaction, developed at 37 C, was stopped by aspiration of the medium. The cells were rapidly washed three times with 1 ml ice-cold buffer (135 mM NaCl, 5 mM KCl, 2.5 mM CaCl2, 1 mM LaCl3, 10 mM HEPES). The cells were then solubilized in 1 ml KRB containing 0.1% Triton X-100 for 1 h at room temperature. An aliquot of the solution (100 μl) was then assayed for 45Ca2+ content in a ?-counter after addition of a liquid scintillation medium (complete phase combining system, Amersham Pharmacia Biotech).
Small interfering (si)RNA-mediated silencing of endogenous expression of ERK1 and ERK2 in MIN6 cells
Expression of either ERK1 or ERK2 was specifically silenced in MIN6 cells using mouse 20–25 nucleotide prevalidated siRNA duplexes purchased from Santa Cruz Biotechnology. On the day before transfection, MIN6 cells were resuspended in 12-well plates in appropriate growth medium with 15% FCS but without antibiotics. Cells were then grown overnight to reach 40–50% confluency. The day of the experiment, siRNA complexes were prepared and transfection was performed according to the manufacturer’s instructions. Concentration of ERK1 and ERK2 siRNAs for MIN6 cell transfection required optimization (data not shown). Briefly, 3.6 μl of 10 μM of ERK1 or ERK2 siRNA (final concentration 50 nM) was mixed gently with 60 μl siRNA transfection medium (solution A), and 3.6 μl siRNA transfection reagent was mixed with 14.5 μl siRNA transfection medium (solution B). After a 5-min incubation at room temperature, solution A and solution B were combined and incubated at room temperature for 20 min to form transfectant-siRNA complex. Medium from MIN6 cells was then removed and replaced by 600 μl fresh growth medium with 15% FCS but without antibiotics. Transfectant-siRNA complexes were added dropwise while gently rocking the 12-well plates. MIN6 cells were transfected with ERK1 or ERK2 siRNA for at least 5 h at 37 C before switching to fresh growth medium with 15% FCS including antibiotics. Then 24, 48, 72, and 96 h after transfection, cells were lysed and expressions of either ERK1 or ERK2 were assayed by Western blot. The amounts of siRNA, siRNA transfection reagent, and siRNA transfection medium were proportionally scaled down to the surface area of cell culture (24-well plates format) to perform insulin secretion.
Western blotting and immunoprecipitation
Pancreatic islets were stabilized for 2 h in KRB containing 2.8 mM glucose in the absence or the presence of PD98059 as described for insulin release experiments, washed, and further incubated in groups of 100 islets for 10 min at 37 C in KRB supplemented with effectors as described in figure legends. After the 10-min incubation, islets were rapidly centrifuged and 1 μl/islet of cold lysis buffer [50 mM HEPES; 1% Nonidet-P40; 2 mM Na3VO4; 100 mM NaF; 10 mM PyrPO4; 4 mM EDTA; 1 mM phenylmethylsulfonyl fluoride (PMSF); 1 μg/ml leupeptin; 1 μg/ml aprotinin] was added. Islets were sonicated (10 sec), centrifuged at 14,000 rpm for 30 min at 4 C to remove insoluble materials, and supernatant were stored at –20 C until use for subsequent protein determination by Bradford assay and Western blotting.
For experiments with MIN6 cells, after a 2-h preincubation in KRB without glucose, 6-well plates MIN6 cells (70–80% confluence) were incubated in KRB containing various glucose concentrations as indicated in figure legends. Cells were then washed with ice-cold PBS, and lysis buffer (50 mM HEPES, 1 mM EDTA, 1% Triton X-100, 0.1% sodium dodecyl sulfate, 1 mM Na3VO4, 1 mM PMSF, 30 mM PyroPO4, 10 mM NaF, and 1 mg/ml bacitracin) was added. Cell lysates were centrifuged at 14,000 rpm for 30 min at 4 C to remove insoluble materials, and protein content of the supernatant was determined by Bradford assay. For immunoprecipitation, the supernatants (400–800 μg total protein) were incubated overnight at 4 C with the primary antibody. Immunocomplexes were precipitated from the supernatant with protein A/G-plus agarose for 4 h at 4 C and washed three times with ice-cold cell lysis buffer. For Western blot analysis, MIN6 cells, islet lysates (25–50 μg protein/lane), and immunoprecipitates were denatured by boiling (3 min) in Laemmli sample buffer containing 100 mM dithiothreitol and resolved by SDS-PAGE as previously described (30). Immunoprecipitation and Western blot analysis were carried out with an equal amount of total proteins. Visualization and quantification of the bands were obtained using an Image Station 2000R system (Eastman Kodak Co., Rochester, NY).
Subcellular fractionation
After a 2-h preincubation in KRB, MIN6 cells plated in 10-cm dishes were stimulated with glucose. Cells were washed twice with ice-cold PBS and scraped in 500 μl hypotonic buffer (10 mM HEPES; 10 mM NaCl; 1 mM KH2PO4; 5 mM NaHCO3; 1 mM CaCl2; 0.5 mM MgCl2; 5 mM EDTA; 1 mM PMSF; 10 μg/ml aprotinin; 10 μg/ml leupeptin; 1 μg/ml pepstatin). After a 15-min incubation, cells were dounced (pestle B) 50 times on ice and centrifuged for 5 min at 7500 rpm at 4 C. Supernatant containing cytosol and membranes was collected and preserved for protein content determination and Western blot analysis. The pellet containing nuclei was dounced (pestle B) 30 times in TSE buffer [10 mM Tris (pH 7.5), 300 mM sucrose, 1 mM EDTA (pH 8), 0.1% Nonidet-P40, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin, and 1 μg/ml pepstatin], centrifuged for 5 min at 5000 rpm at 4 C, and rinsed twice in 1 ml TSE buffer. Final pellet containing pure nuclei was dissolved in TSE buffer for protein determination before denaturation in Laemmli buffer and analysis by SDS-PAGE.
Immunostaining for phospho-ERK1/2 in MIN6 cells
Cells were grown on glass coverslips for 3–5 d. Cells were then fixed in 3.7% formaldehyde in PBS for 30 min, washed with 0.1 M Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% Triton X-100 (TBST buffer), and permeabilized for 10 min in methanol at –20 C. After incubation in blocking buffer (5% normal horse serum in TBST buffer) for 1 h at room temperature, cells were exposed overnight at 4 C to the mouse anti-phospho-ERK1/2 antibody (1:400 dilution in 5% BSA TBST buffer). After washing, phospho-ERK1/2 antibody was detected by incubation for 1 h at room temperature with fluorescein isothiocyanate-conjugated donkey antimouse in TBST buffer containing 3% BSA (1:100 dilution). Coverslips were then mounted using a polyvinyl alcohol medium at least 1 h before observation using a dual-photon confocal microscope (Zeiss, Oberkocher, Germany).
Statistical analysis
Results are presented as mean ± SEM. Differences between results were analyzed by Student’s t test (unpaired unless indicated) (*, P < 0.05; **, P < 0.01; ***, P < 0.001).
Results
Increase in amount of activated forms of ERK1/2 in cytoplasm and nuclei of MIN6 cells on glucose stimulation
As a prelude to a study of the role of ERK1/2 activity in phosphorylation of cytoplasmic substrates and insulin secretion, we first attempted to confirm earlier reports by characterizing the effect of glucose on subcellular redistribution of active ERK1/2 using the MIN6 cells (13, 16). ERK1 and ERK2 were barely phosphorylated in MIN6 cells in the standard culture medium. When glucose was withdrawn for 2 h (2-h preincubation in KRB), activities of both enzymes were nearly undetectable (Fig. 1A). When 10 mM glucose was reintroduced, we found that both kinases were rapidly and transiently activated (Fig. 1A). Glucose dose-dependently increased ERK1/2 phosphorylation from 3 to 20 mM. Maximal activation of ERK2 was observed at 10–15 mM glucose (Fig. 1B). It is noteworthy that a noninsulinotropic glucose concentration (3 mM) induced a weak but significant phosphorylation of ERK1/2 (Fig. 1B). Because ERK1/2 activation elicited by glucose has been shown to be dependent on calcium influx through the opening of VDCCs (10, 13, 15, 16), we verified that the glucose-induced ERK1/2 activity was completely abolished when the cells were treated with the VDCC blocker nifedipine (Fig. 1C).
FIG. 1. Glucose transiently activates ERK1/2, which remains in large part in the cytoplasmic fraction. A–D, Western blot assay of activated and total ERK1/2 levels in MIN6 cells. Cells were preincubated in KRB for 2 h and either untreated or exposed to various concentration of glucose for different times as indicated. C, Effect of L-type calcium channel antagonist, nifedipine (2 μM), on ERK1/2 phosphorylation in MIN6 cells after treatment (5 min) with glucose. Nifedipine was added concomitantly with glucose treatment. D, Effect of PKA inhibitor, H89 (2 μM), on ERK1/2 phosphorylation in MIN6 cells after treatment (5 min) with glucose. H89 was added during the 2-h KRB preincubation period and maintained during the glucose treatment. E, Presence of active ERK1/2 in subcellular fractions from MIN6 cells exposed to glucose [nuclear (N) and cytosolic (C) fractions]. Cells were preincubated in KRB for 2 h and either untreated or stimulated (5 min) with glucose. A–E, Representative exposures of five experiments are shown. Total ERK1/2 levels within total cell lysates (A–D) and the different cellular fractions (E) are shown. F, Localization of the active form of ERK1/2 after glucose stimulation. Cells were preincubated in KRB for 2 h and either left untreated (a) or stimulated (5 min) with glucose (10 mM) (b).
Because it has been shown that cAMP and activation of PKA are involved in the mechanism of glucose activation of ERK1/2 in the INS-1 cell line (10), we also investigated the potential role of PKA in the ERK1/2 activation elicited by glucose in MIN6 cells. Cells were stimulated with glucose in either the absence or presence of the PKA inhibitor H89, at the minimal concentration shown to exert a complete suppression of the glucagon-induced ERK1/2 activation (31). As shown in Fig. 1D, H89 inhibited by 70% the glucose-stimulated ERK1/2 activation. The specificity of the H89 treatment was ascertained by the fact that it had no effect on epidermal growth factor-induced ERK1/2 phosphorylation (data not shown).
Because the subcellular localization plays an important role in determining the functions of ERK1 and ERK2 (22), we examined their distribution in subcellular fractions of MIN6 cells before and after exposure to glucose. In the absence of glucose, the immunoreactive ERK1 and ERK2 were associated with the cytoplasmic fraction and absent from the nuclear fraction (Fig. 1E). Treatment of cells with 10 mM glucose caused a substantial increase in the amount of ERK1 and ERK2 in the nuclear fraction, although a large part of active ERK1/2 remained in the cytoplasm. Immunolocalization using confocal microscopy showed an ERK1/2 staining of active forms throughout the cytoplasm as well as in the nucleus of cells exposed to 10 mM glucose (Fig. 1F), consistent with the results of subcellular fractionation.
Glucose-induced ERK1/2 activity regulates synapsin I phosphorylation
The majority of activated ERK1 and ERK2 remained in the MIN6 cells cytoplasm on glucose stimulation (Fig. 1, E and F). Therefore, we evaluated whether ERK1/2 activity elicited by glucose leads to phosphorylation of cytoplasmic substrate(s). In neurons, ERK1/2 serine-phosphorylate synapsin I leading to the recruitment of secretory vesicles to the plasma membrane for neurotransmitter exocytosis (32, 33, 34). Because it has been reported that synapsin I colocalizes with the insulin secretory granules in the MIN6 ? cells cytoplasm and that the phosphorylation/dephosphorylation cycle of synapsin I has been proposed to be implicated in insulin exocytosis (8, 9, 18), we sought to determine whether the ERK1/2 activity regulates the serine-phosphorylation of this cytoplasmic substrate on glucose stimulation. First, we used an immunoprecipitation approach to evaluate the existence of an association between synapsin I and ERK1/2 in the basal state and on glucose stimulation. Second, we measured the serine-phosphorylation level of synapsin I upon glucose stimulation, in either the absence or presence of PD98059, a selective inhibitor of MEK1/2, the upstream ERK1/2 kinase, which displays an IC50 of 2–4 μM and maximal inhibitory effects at 10–100 μM (35, 36). It must be noted that PD98059 is a compound with a highly selective profile, when used at reasonable doses, because no protein kinase has been found to be inhibited by PD98059 at a concentration (50 μM and below) that prevents activation of the ERK1/2 cascade (37). As a control, we verified that the PD98059 treatment completely abolished the ERK1/2 activation elicited by glucose (Fig. 2A). To rule out the possibility that PD98059 may be acting as a moderate metabolic poison on any step of glycolysis or Krebs cycle, thus reducing ATP synthesis and subsequent calcium influx, we verified that the PD98059 treatment did not inhibit the calcium influx induced by glucose (Fig. 2B).
FIG. 2. Glucose phosphorylates synapsin I through the ERK1/2 cascade. MIN6 cells were preincubated for 2 h in KRB as for Fig. 1. A, B, D, E, and F; PD98059 (20 μM) was present (or not) during 2-h KRB preincubation period and the glucose stimulation as indicated. A, MIN6 cells were either untreated or stimulated for 5 min with glucose and then lysed. A typical exposure representative of three experiments is shown. B, 45Ca2+ uptake was measured as described under experimental procedures. After 2 h preincubation in KRB, cells were incubated 5 min with glucose. Data are means ± SEM of four to six determinations. C, D, and E, MIN6 cells were either untreated or stimulated for 5 min with glucose (C and D) and 20 min (E) lysed and subjected to immunoprecipitation (IP) in reduced conditions with goat anti-synapsin I antibody. Immunoprecipitated proteins were blotted with anti-ERK1/2 (A), anti-phospho-ERK1/2 (B), or antiphosphoserine (E) antibodies. Representative exposures of five to six independent experiments are shown. Identical amounts of proteins were analyzed as described in Materials and Methods. As control, relative quantities of ERK1/2 (C and D) and synapsin I (E) in cell lysates are shown. F, Graph showing level of serine-phosphorylation of synapsin I as mean ± SEM of six independent immunoprecipitation experiments deduced from quantitative results obtained from analysis using Kodak Image Station 2000R system program. **, P < 0.01 vs. basal synapsin I phosphorylation.
The immunoprecipitation experiments with synapsin I antibody followed by immunoblotting for ERK1/2 showed no interaction between synapsin I and ERK1/2 in the basal state, whereas a marked association between synapsin I and ERK1/2 was observed after 5 min of glucose stimulation (Fig. 2C), suggesting that only the phosphorylated active form of ERK1/2 interacts with synapsin I. As a confirmation, we found that the phosphorylated form of ERK1/2 was physically associated with synapsin I on glucose stimulation and that this interaction was totally prevented by the PD98059 treatment (Fig. 2D). We next investigated whether the PD98059 treatment inhibited the glucose-induced synapsin I phosphorylation. As seen in Fig. 2, E and F, glucose induced a 2-fold increase in serine-phosphorylation of synapsin I through the ERK1/2 signaling cascade because this phosphorylation was totally prevented by the PD98059 treatment. Considering that the ERK1/2-mediated phosphorylation of synapsin I in neurons is essential for neurotransmitter exocytosis and that the phosphorylation/dephosphorylation cycle of synapsin I is implicated in the exocytosis of insulin secretory granules, we reexamined the potential role of ERK1/2 activity in glucose-induced insulin release.
ERK1/2 activity participates in the glucose-induced insulin secretion in MIN6 cells
We used the same protocol as described for ERK1/2 and synapsin I phosphorylation experiments and studied insulin release in MIN6 cells preincubated for 2 h in KRB. As shown in Fig. 3A, we found that the glucose-induced insulin release was inhibited by 70% when the ERK1/2 activity was totally blocked by PD98059 (Fig. 2A). The PD98059-mediated inhibition of glucose-induced insulin release was reversed when the inhibitor was removed from the glucose stimulation period. We observed that ERK1/2 activated by glucose translocate to the MIN6 cell nucleus (Fig. 1, E and F). Because it has been shown that the glucose-induced ERK1/2 activity controls the insulin gene transcription and that inhibition of protein synthesis by cycloheximide impairs insulin secretion (14, 23, 38), the ERK1/2 cascade may also participate in replenishing the pool of insulin granules. Insulin biosynthesis is a complex regulated process, and complementary methods exist to measure insulin biosynthesis directly. However, no change in the insulin content of MIN6 cells with or without the presence of PD98059 was observed, possibly excluding a defect in insulin biosynthesis (Fig. 3B). Moreover, because we observed that the glucose-induced insulin secretion was almost totally restored when PD98059 was removed during the glucose stimulation period (Fig. 3A), it is unlikely that the inhibition of glucose-induced insulin secretion by blocking ERK1/2 activity with PD98059 was related to a defect in protein synthesis or stability.
FIG. 3. ERK1/2 activity participates in glucose-induced insulin release in MIN6 cells. Cells were preincubated in KRB for 2 (A) or 1 h (C) and either untreated or stimulated with glucose. PD98059 was present (or not) during 2-h KRB preincubation period and stimulation (A and C). Data are means ± SEM of 12 determinations, each performed in triplicate. ***, P < 0.001 vs. glucose-induced insulin secretion. B, MIN6 cells were preincubated in KRB for 2 h in the presence or not of PD98059 and lysed in PBS + 1% Triton X-100 for 30 min. D, Cells were preincubated in KRB for 2 h and either untreated or stimulated (5 min) with glucose in the presence or absence of PD98059, lysed, and levels of ERK1/2 phosphorylation analyzed by Western blotting. A typical exposure of three experiments is shown.
We also tested the PD98059 treatment used by Khoo and Cobb (16) to investigate the role of ERK1/2 in the glucose-induced insulin release in INS-1 cells [e.g. cells were preincubated with 100 μM PD98059 for 1 h before addition of glucose, and PD98059 was present throughout the glucose stimulation period], which is quite close to our experimental conditions. On the contrary to what was observed in INS-1 (16), this PD98059 treatment condition blocked the glucose-induced insulin release by 53% in MIN6 cells (Fig. 3C). Again, PD98059-mediated inhibition of glucose-induced insulin release was reversed when the inhibitor was removed from the glucose stimulation period (Fig. 3C). Thus, it appears, from these results, that there is a substantial discrepancy between the data obtained from different ?-cell lines, MIN6 and INS-1. As control, we verified that this PD98059 treatment condition abolished the glucose-stimulated ERK1/2 activation (Fig. 3D).
ERK1 and ERK2 siRNA in MIN6 cells
To strengthen our data showing that ERK1/2 activity is required for glucose-stimulated insulin secretion using pharmacological blockade of ERK1/2 with PD98059 (Fig. 3, A and C), we performed a siRNA knockdown strategy to silence the expression of ERK1 and ERK2 mRNAs in MIN6 cells. We first found that transfection of MIN6 cells using modest amounts of ERK2 siRNA (50 nM) resulted in 100% knockdown of ERK2 and provoked MIN6 cell mortality (data not shown). Because it is well known that ERK2 is a critical cell cycle protein (21, 22, 39), these results raise the possibility that a total depletion of ERK2 in the MIN6 ?-cell might be lethal.
On the other hand, to avoid cell death under ERK2 siRNA, we performed ERK1 siRNA with 50 nM of siRNA duplexes. The ERK1 siRNA-transfected MIN6 cells were viable and no morphologic changes were observed (data not shown). As seen in Fig. 4A, ERK1 and ERK2 expressions were silenced by 70–80 and 50%, respectively, in ERK1 siRNA-transfected MIN6 cells, compared with nontransfected control cells. Therefore, we used this ERK1 siRNA-transfected MIN6 cell model, which phenotypically shows significant reduction of both ERK1 and ERK2 proteins, to evaluate the role of ERK1/2 pathway in insulin secretion stimulated by glucose. Immunoblot analysis of down-regulated MIN6 cells for ERK1 and ERK2 proteins revealed that glucose-induced ERK1 and ERK2 phosphorylation was largely inhibited (Fig. 4B). The specificity of our siRNA approach was ascertained by using ?-actin as internal and loading control (Fig. 4, A and B) or using other similarly sized ERK1/2-related siRNA duplexes that failed to induce any change in the expression of any of the mRNAs studied (ERK1, ERK2, or ?-actin) (data not shown). Interestingly, we found that glucose-stimulated insulin secretion was significantly reduced by 36% in ERK1 siRNA-transfected cells (Fig. 4C), clearly indicating that the ERK1/2 signaling pathway participates in insulin secretion in MIN6 ?-cells.
FIG. 4. ERK1/2 activation and insulin secretion defects in response to glucose in ERK1 and ERK2 siRNA transfected MIN6 cells. A, Western blot analysis of ERK1 and ERK2 expressions in nontransfected cells, incubated with transfection reagent alone as control (control reagent), and in ERK1 siRNA-transfected MIN6 cells. Blot was probed with ERK1 antibody and actin antibody used as specificity and loading control. A typical exposure representative of three independent experiments is shown. B, After transfection procedure, nontransfected cells (control), incubated with transfection reagent alone (control reagent), and ERK1 siRNA-transfected MIN6 cells were preincubated 2 h in KRB. For Western blot analysis, cells were then stimulated with glucose for 5 min, washed with ice-cold PBS, and lysed. Lysates (25 μg protein) were resolved by 10% SDS-PAGE in reduced conditions and blotted with phospho-ERK1/2, ERK1/2, or actin antibody. A typical exposure representative of two experiments is shown. As control, relative quantities of ERK1/2 and actin in cell lysates are also shown. C, For measuring insulin secretion, after transfection cells were preincubated for 2 h KRB and then either untreated or stimulated with glucose. Data are means ± SEM of nine determinations.
Insulinotropic glucose concentration induces phosphorylation of ERK1/2 in pancreatic islets
To confirm, using another model, the role of ERK1/2 in glucose-induced insulin secretion, we first investigated whether an insulinotropic glucose concentration induces phosphorylation of ERK1/2 in pancreatic islets. It has been reported that a 10-min stimulation by glucose had no effect on ERK1/2 activity in rat islets (24). Indeed, using the same experimental procedure (1-h stabilization in KRB supplemented with 2.8 mM glucose), we did not observe any activation of ERK1/2 by 16.7 mM glucose, compared with the basal (2.8 mM glucose) level of ERK1/2 phosphorylation, as reported by Burns et al. (24) (data not shown). Nevertheless, when stabilization was prolonged to at least 2 h (Fig. 5A), we found that, compared with the 1-h stabilized islets, basal ERK1/2 phosphorylation was reduced but not abolished and that 16.7 mM glucose stimulation of islets for 10 min resulted in a significant increase in the doubly phosphorylated active forms of these kinases (Fig. 5A). Pretreatment of islets with PD98059 during the 2-h stabilization period resulted in a deep inhibition of the 2.8 mM as well as the 16.7 mM glucose-induced ERK1/2 phosphorylation (Fig. 5A). Because ERK1/2 activation elicited by glucose is dependent on calcium influx through opening of VDCCs in MIN6 (Fig. 1C) and INS-1 (10, 16), we investigated whether the glucose-induced ERK1/2 activity in pancreatic islets was blocked when the cells were treated with the VDCC blocker nifedipine. As seen in Fig. 5B, nifedipine did not inhibit the basal ERK1/2 phosphorylation, whereas it blocked by 80% the 16.7 mM glucose-induced ERK1/2 phosphorylation. Thus, in pancreatic islets, the ERK1/2 activity elicited by an insulinotropic glucose concentration requires calcium influx via the opening of VDCCs, as observed in MIN6 cells (Fig. 1C) and as reported using another ?-cell line (10, 16). It is noteworthy that under the low, noninsulinotropic 2.8 mM glucose concentration, we found that ERK1/2 are phosphorylated within the islets independently from extracellular calcium entry.
FIG. 5. Glucose stimulates the ERK1/2 activity in isolated rat pancreatic islets. Islets were stabilized in KRB supplemented with glucose (2.8 mM) for 2 h and then incubated with 2.8 or 16.7 mM glucose for 10 min. A, Effect of PD98059 (20 μM) on ERK1/2 phosphorylation islets after treatment with glucose. PD98059 was present during the 2-h stabilization period and maintained during the 10-min glucose stimulation. B, Effect of nifedipine (2 μM) on ERK1/2 phosphorylation in islets after treatment with glucose. Nifedipine was added during the 10-min glucose stimulation period. A and B, The most representative blots obtained from three to five independent experiments are shown.
Role of ERK1/2 in the glucose-induced insulin secretion from islets of Langerhans
We next evaluated the potential involvement of ERK1/2 signaling cascade in glucose-induced insulin release from isolated pancreatic islets. We first examined the effect of PD98059 treatment on glucose-induced insulin release during a static incubation. As shown in Fig. 6A, 2.8 and 16.7 mM glucose-induced insulin releases were reduced by 60 and 62%, respectively when ERK1/2 was blocked by PD98059 treatment. We verified that a PD98059 treatment had no effect on insulin content of the islets (Fig. 6B) and that the VDCC blocker nifedipine did not inhibit the 2.8 mM glucose-basal secretion, whereas it was efficient in blocking the 16.7 mM glucose-induced insulin secretion (Fig. 6A). These data demonstrate that under low, noninsulinotropic (2.8 mM) glucose concentration, ERK1 and ERK2 are active within the islets independently from extracellular calcium entry and are involved in basal insulin secretion, whereas an insulinotropic (16.7 mM) glucose concentration, there is a significant increase in the phosphorylated active forms of ERK1/2, induced via calcium influx, which participate in the glucose-induced insulin release.
FIG. 6. Glucose-induced insulin secretion from isolated rat islets requires the ERK1/2 activity. A, Islets were stabilized for 2 h in KRB supplemented with glucose (2.8 mM) and then stimulated with glucose (2.8 or 16.7 mM) for 10 min. Islets were treated with PD98059 (20 μM) or nifedipine (2 μM) as for Fig. 5. Data are means ± SEM of 13 determinations, each performed in duplicate. *, P < 0.05 vs. glucose-induced insulin secretion. B, Insulin content determination as described in Materials and Methods.
ERK1/2 participate in both phases of glucose-stimulated insulin secretion
Because static incubation fails to reveal dynamics of insulin release in terms of biphasic secretion, we determined the impact of PD98059 treatment on both phases of glucose-stimulated insulin secretion using the pancreatic islet perifusion model. Isolated islets were stabilized for 2 h in the absence or the presence of PD98059 and then transferred to the perifusion chambers for an additional 20-min perifusion of 2.8 mM glucose (one chamber containing untreated control islets, the other chamber containing the PD98059-treated islets), followed by 16.7 mM glucose stimulation for 30 min in the absence (open circles) or presence of PD98059 (filled squares). As seen in Fig. 7A, basal insulin secretion was inhibited during the initial 20-min perifusion period with PD98059, and the 16.7 mM glucose-induced insulin release was significantly reduced in the PD98059-treated islets. This marked inhibition of 16.7 mM glucose-induced insulin secretion occurred throughout the time course and was present during both the first and second phases of insulin release. The subsequent removal of 16.7 mM glucose resulted in a sharp decrease in insulin secretion near to basal levels within 10–20 min, in the absence or presence of PD98059. Insulin released during the first phase, as measured by evaluation of the area under the curve (AUC) calculated classically, from the initial insulin rise (min 3 after the beginning of the glucose stimulation) to the lowest point between the first and second rise in insulin release (min 7), was inhibited by about 55% (Fig. 7B). The amount of insulin secreted during the second phase, starting from min 8 to the end of stimulation, was inhibited by 38% (Fig. 7C). Because the basal (2.8 mM glucose) secretion was altered by the PD98059 treatment, we calculated the AUC with the average of the –10 to 0 min as the baseline. The results indicate 27 (nonsignificant) and 49% (significant) inhibitions for the first and second phases, respectively (Fig. 7, D and E). We can conclude that ERK1/2 activity participates in the basal as well as first and second phases of glucose-induced insulin release in rat pancreatic islets.
FIG. 7. Role of ERK1/2 activity in the two phases of the glucose-induced insulin secretion from isolated rat islets. A, Islets were stabilized for 2 h in KRB with glucose (2.8 mM) either in the absence or presence of PD98059 and then transferred to perifusion chambers in water bath at 37 C. First, islets were perifused for 20 min with KRB containing 2.8 mM glucose. At time 0, glucose was raised to 16.7 mM for 30 min and then lowered to 2.8 mM for 20 min. PD98059 was maintained from the beginning of the stabilization to the end of stimulation (filled squares, PD98059-treated islets). Data are mean ± SEM. Each perifusion was repeated three times and insulin secretion was quantified in duplicate. *, P < 0.05; **, P < 0.01 (paired Student’s t test, the pairs being the two chambers of the same water bath). B and C, Insulin secretion was calculated as AUC. D and E, Insulin secretion was calculated as AUC with the average of the –10 to 0 min as the baseline. *, P < 0.05 (paired Student’s t test, the pairs being the two chambers of the same water bath).
Discussion
In this report, we show that glucose-activated ERK1/2 remain in large part in the ?-cell cytoplasm. Supporting the concept that the ERK1/2 cascade is a significant regulator of glucose-dependent cytoplasmic events, we demonstrate that ERK1/2 participate in the glucose-mediated phosphorylation of synapsin I, a protein known to be associated with insulin granules and proposed to be involved in the granule translocation that precedes insulin exocytosis (8, 9, 20, 32, 33, 34). Moreover, using MIN6 cells and perifused isolated pancreatic islets, we provide evidence that ERK1/2 activity also contributes to optimal glucose-induced insulin secretion. Therefore, we propose that the glucose-induced activation of the ERK1/2 signaling pathway in ?-cells participates in the amplification of the glucose action on insulin secretion by controlling activation of one or several cytosolic proteins including synapsin I.
In both MIN6 cells and isolated islets, we confirm that glucose induces phosphorylation of ERK1/2 via calcium entry and PKA activation. As reported by Briaud et al. in INS-1 cells (10), it is probable that glucose-induced rise in cytoplasmic [Ca2+] leads to cAMP production and subsequent PKA activation that mediate in part ERK1/2 activation in MIN6 cells. The majority of activated forms of ERK1/2 upon the glucose challenge remained in the cytoplasm interacting with synapsin I and, possibly, with other protein(s) involved in exocytosis. In the basal state, we demonstrate that ERK1/2 do not interact with synapsin I. On the other hand, we observed a physical interaction between synapsin I and the activated forms of ERK1/2 on glucose stimulation, leading to synapsin I phosphorylation. It has already been demonstrated that synapsin I is phosphorylated by PKA and CAMKII in glucose-stimulated ?-cells (18, 20). Our conditions (20-min glucose stimulation in the absence of any phosphatase inhibitor) allow us to specifically detect the serine-phosphorylated state of synapsin I, which is not dependent on PKA nor CAMKII, because the former is quickly reversed by dephosphorylation by calcineurin (18), whereas the latter returns to the basal value within 15 min (20). Thus, it is not surprising that the PD98059 treatment totally prevented the serine-phosphorylation of synapsin I. Phosphorylation of ERK1/2-specific sites of synapsin I in ?-cells appears to be more sustained than other phosphorylation sites, as shown in other cell types (32). The rapid and transient activation of ERK1/2 by glucose in ?-cells is sufficient to induce a sustained phosphorylation of synapsin I.
Another major finding of this paper is that, in both MIN6 cells and isolated islets, the glucose-induced ERK1/2 activation is required for an optimal glucose-induced insulin secretion because the observed inhibition of the glucose-induced insulin secretion when ERK1/2 activity was blocked by a 2-h treatment with PD98059 was not due to a defect in insulin content. In addition, the rapid reversibility of the inhibition indicates that the effect is not related to a defect in protein synthesis or disappearance of short half-life proteins. The results of the siRNA-mediated ERK1/2 partial knockdown fully confirmed those obtained using pharmacological blockade. However, we cannot rule out the possibility that the glucose-induced insulin release observed in ERK1 siRNA-transfected cells may result not only from a dysfunction of phosphorylation/activation of synapsin I and/or other unknown protein(s) of the insulin exocytotic process but also from a modification in the insulin gene transcription. It must also be noted that 100% knockdown of ERK2 provoked MIN6 cell mortality raising the possibility that ERK2 is a critical cell cycle protein in pancreatic ?-cells.
Using INS-1 cells, Khoo and Cobb (16) reported that glucose-induced insulin release occurs normally, even if the ERK1/2 activity was blocked by PD98059. Here, using MIN6 cells, we show that the same PD98059 treatment inhibited the glucose-induced insulin secretion. Thus, there is an apparent discrepancy with the data obtained from different cell lines. This discrepancy could be attributed to their differences in glucose sensitivity. Indeed, it is noteworthy that INS-1 cells from rats and MIN6 cells from mice display different insulin secretion profiles (40, 41, 42). Therefore, differences in the glucose metabolism, a key step in promoting insulin release, or subtle differences in the recruitment of intracellular molecular mechanisms responsible for the generation of the first- and second-phase insulin release could explain in part the observed discrepancy between the cellular models.
Using static incubation of pancreatic islets, we confirmed the involvement of the ERK1/2 signaling cascade on insulin release on a 10-min 16.7 mM glucose challenge. However, we observed that insulin release became insensitive to the PD98059 treatment with increasing times of static glucose stimulation (30–90 min) and that long-lasting (90 min) stimulation of insulin release by 16.7 mM glucose was not significantly affected by PD98059 (data not shown), which is fully consistent with the work of Burns et al. (24). It is well known that local interactions between the cells in the islet exert an important level of control in the physiological regulation of insulin secretion. It is likely that, during static incubation, accumulated insulin within the islet might suppress the secretion of pancreatic peptides, such as miniglucagon from -cells or somatostatin from -cells which have been both reported to be efficient inhibitors of insulin secretion (26, 43, 44). The suppression of an inhibitory tone on ?-cell due to miniglucagon and/or somatostatin, within the islet, might further enhance the insulinotropic glucose effect and might render negligible over the time the fraction of the glucose-induced insulin secretion due to ERK1/2. Because insulin receptors are expressed and functional in ?-cells, it has been reported that insulin regulates its own secretion in a regulatory feedback loop (45, 46). Therefore, accumulated insulin within the islet could also modulate its own secretion through mobilization of parallel mechanisms, implying other kinases, interfering with the inhibition of ERK1/2 signaling pathway. Unlike the results observed in static incubations, in the islets perifusion experiments, in which there is no accumulation of islet peptides at the vicinity of the ?-cells, PD98059 treatment exerted a persistent inhibition of the glucose-induced insulin release throughout the 30-min glucose stimulation period.
Additionally, whatever the calculation method used for measuring AUC in the perifusion experiments, our data suggest that ERK1/2 activity participates in the basal as well as the first and the second phases of the glucose-induced insulin secretion. These observations are compatible with the concept that ERK1/2 activated by glucose participate in the regulation of cytoskeleton recast and exocytotic machinery for optimal basal and stimulated insulin secretion through phosphorylation of proteins such as synapsin I. The precise role of ERK1/2 activity in the basal and stimulated insulin secretion from MIN6 cells and islets deserves now to be investigated in more detail.
In addition, glucagon-like peptide-1 and glucose-dependent insulinotropic peptide activate, in ?-cells, the ERK1/2 signaling cascade that regulates cell survival, proliferation, and insulin gene transcription (10, 28, 47, 48, 49). Our data suggest that ERK1/2 activation by incretin peptides may be an additional mechanism that regulates insulin granule exocytosis through phosphorylation of cytosolic protein(s).
In summary, our data provide new insights in the regulation by glucose of the ?-cell physiology. We demonstrate that ERK1/2, among others kinases such as PKA or CAMKII, represent an additional transduction signal that amplifies the glucose effect on insulin secretion. Hyperglycemic exposure to the ?-cells, which is a key feature of type 2 diabetes, has been reported to contribute to the deterioration of ?-cell function, a concept referred to as glucotoxicity. It is now interesting to study the potential role of the ERK1/2 signaling pathway in the progressive loss of ?-cell insulin secretion due to long-term hyperglycemia throughout the course of type 2 diabetes. Recently it has been reported that IL-1?-stimulated ERK1/2 activation was a mediator of the long-term deleterious effects of glucose on ?-cell function (50).
In conclusion, our results establish that the ERK1/2 activity is necessary for an optimal glucose-induced insulin secretion, a discovery that has now to be taken into account for understanding in more detail the molecular mechanisms the impairment of which leads to ?-cell defects involved in type 2 diabetes.
References
Docherty K, Steiner DF 1996 Molecular and cellular biology of the ? cell. In: Portes D, Sherwin R, eds. Ellenberg Rifkin’s diabetes mellitus. 5th ed. Stamford, CT: Elsevier Science Publishing; 29–48
Matschinsky FM 1996 Banting Lecture 1995. A lesson in metabolic regulation inspired by the glucokinase glucose sensor paradigm. Diabetes 45:223–241
Henquin JC, Ishiyama N, Nenquin M, Ravier MA, Jonas JC 2002 Signals and pools underlying biphasic insulin secretion. Diabetes 51(Suppl 1):S60–S67
Barg S, Eliasson L, Renstr?m E, Rorsman P 2002 A subset of 50 secretory granules in close contact with L-type Ca2+ channels accounts for first-phase insulin secretion in mouse ?-cell. Diabetes 51(Suppl 1):74–82
Ohara-Imaizumi M, Nakamichi Y, Tanaka T, Ishida H, Nagamatsu S 2002 Imaging exocytosis of single insulin secretory granules with evanescent wave microscopy: distinct behavior of granule motion in biphasic insulin release. J Biol Chem 277:3805–3808
Somers G, Blondel B, Orci L, Malaisse WJ 1979 Motile events in pancreatic endocrine cells. Endocrinology 104:255–264
Li G, Rungger-Brandle E, Just I, Jonas JC, Aktories K, Wollheim CB 1994 Effect of disruption of actin filament by clostridium botulinum C2 toxin on insulin secretion in HIT-T15 cells and pancreatic islet. Mol Biol Cell 5:1199–1213
Matsumoto K, Kukunaga K, Miyazaki JI, Shichiri M, Miyamoto E 1995 Ca2+/calmodulin-dependent protein kinase II and synapsin I-like protein in mouse insulinoma MIN6 cells. Endocrinology 136:3784–3793
Matsumoto K, Ebihara K, Yamamoto H, Tabuchi H, Fukunaga K, Yasunami M, Ohkubo H, Shichiri M, Miyamoto E 1999 Cloning from insulinoma cells of synapsin I associated with insulin secretory granules. J Biol Chem 274:2053–2059
Briaud I, Lingohr MK, Dickson LM, Wrede CE, Rhodes CJ 2003 Differential activation mechanisms of Erk-1/2 and p70(S6K) by glucose in pancreatic ?-cells. Diabetes 52:974–983
Wenham RM, Landt M, Easom RA 1994 Glucose activates the multifunctional calcium/calmodulin dependent protein kinase II in isolated rat pancreatic islets. J Biol Chem 269:4947–4952
Knutson KL, Hoenig M 1994 Identification and subcellular characterization of protein kinase C isoforms in insulinoma ?-cells and whole islets. Endocrinology 135:881–886
Benes C, Roisin MP, Van Tan H, Creuzet C, Miyazaki J, Fagard R 1998 Rapid activation and nuclear translocation of mitogen-activated protein kinases in response to physiological concentration of glucose in the MIN6 pancreatic ? cell line. J Biol Chem 273:15507–15513
Benes C, Poitout V, Marie JC, Martin-Perez J, Roisin MP, Fagard R 1999 Mode of regulation of the extracellular signal-regulated kinases in the pancreatic ?-cell line MIN6 and their implication in the regulation of insulin gene transcription. Biochem J 340:219–225
Fr?din M, Sekine N, Roche E, Filloux C, Prentki M, Wolheim CB, Van Obberghen E 1995 Glucose, other secretagogues, and nerve growth factor stimulate mitogen-activated protein kinase in the insulin secreting ?-cell line, INS-1. J Biol Chem 270:7882–7889
Khoo S, Cobb M 1997 Activation of mitogen-activating protein kinase by glucose is not required for insulin secretion. Proc Natl Acad Sci USA 94:5599–5604
Jones P, Persaud S 1998 Protein kinases, protein phosphorylation and the regulation of insulin secretion from pancreatic ?-cells. Endocr Rev 19:429–461
Lester LB, Faux MC, Nauert JB, Scott JD 2001 Targeted protein kinase A and PP-2B regulate insulin secretion through reversible phosphorylation. Endocrinology 142:1218–1227
Mendez CF, Leibiger IB, Leibiger B, Hoy M, Gromada J, Berggren PO, Bertorello AM 2004 Rapid association of protein kinase C with insulin granules is essential for insulin exocytosis. J Biol Chem 278:44753–44757
Yamamoto H, Matsumoto K, Araki E, Miyamoto E 2003 New aspects of neurotransmitter release and exocytosis: involvement of Ca(2+)/calmodulin-dependent phosphorylation of synapsin I in insulin exocytosis. J Pharmacol Sci 93:30–34
Hunter T 1995 Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling. Cell 80:225–236
Volmat V, Pouyssegur J 2001 Spatiotemporal regulation of the p42/p44 MAPK pathway. Biol Cell 93:71–79
Khoo S, Griffen SC, Xia Y, Baer RJ, German MS, Cobb MH 2003 Regulation of insulin gene transcription by extracellular-signal regulated protein kinases (ERK) 1 and 2 in pancreatic ? cells. J Biol Chem 278:32969–32977
Burns CJ, Howell SL, Jones PM, Persaud SJ 1997 Glucose-stimulated insulin secretion from rat islets of Langerhans is independent of mitogen-activated protein kinase activation. Biochem Biophys Res Commun 239:447–450
Lacy PE, Kostianovsky M 1967 Method for the isolation of intact islets of Langerhans from the rat pancreas. Diabetes 16:35–39
Dalle S, Smith P, Blache P, Le-Nguyen D, Le Brigand L, Bergeron F, Ashcroft FM, Bataille D 1999 Miniglucagon (glucagon 19–29), a potent and efficient inhibitor of secretagogue-induced insulin release through a Ca2+ pathway. J Biol Chem 274:10869–10876
Jhala US, Canettieri G, Screaton RA, Kulkarni RN, Krajewski S, Reed J, Walker J, Lin X, White M, Montminy M 2003 cAMP promotes pancreatic ?-cell survival via CREB-mediated induction of IRS2. Genes Dev 17: 1575–1580
Gomez E, Pritchard C, Herbert TP 2002 cAMP-dependent protein kinase and Ca2+ influx through L-type voltage-gated calcium channels mediate Raf-independent activation of extracellular regulated kinase in response to glucagon-like peptide-1 in pancreatic ?-cells. J Biol Chem 277:48146–48151
Kervran A, Rieutort M, Guillaume M 1976 A simultaneous radioimmunoassay for growth hormone and insulin in the plasma of rats and rabbits. Diabetes Metab 2:67–72
Dalle S, Imamura T, Rose DW, Worrall DS, Ugi S, Hupfeld CJ, Olefsky JM 2002 Insulin induces heterologous desensitization of G-protein-coupled receptor and insulin-like growth factor I signaling by downregulating ?-arrestin-1. Mol Cell Biol 22:6272–6285
Dalle S, Longuet C, Costes S, Broca C, Faruque O, Fontes G, el Hani H, Bataille D 2004 Glucagon promotes cAMP-response element-binding protein phosphorylation via activation of ERK1/2 in MIN6 cell line and isolated islets of Langerhans. J Biol Chem 279:20345–20355
Matsubara M, Kusubata M, Ishiguroi K, Uchidai T, Titani K, Taniguchi H 1996 Site-specific phosphorylation of synapsin I by mitogen-activated protein kinase and Cdk5 and its effects on physiological functions. J Biol Chem 271:21108–21113
Jovanovic JN, Czernik AJ, Fienberg AA, Greengard P, Sihra TS 2000 Synapsins as mediators of BDNF-enhanced neurotransmitter release. Nat Neurosci 3:323–329
Jovanovic JN, Benfenati F, Siow YL, Sihra TS, Sanghera JS, Pelech SL, Greengard P, Czernik AJ 1996 Neurotrophins stimulate phosphorylation of synapsin I by MAP kinase and regulate synapsin I-actin interactions. Proc Natl Acad Sci USA 93:3679–3683
Gould GW, Cuenda A, Thomson FJ, Cohen P 1995 The activation of distinct mitogen-activated protein kinase cascades is required for the stimulation of 2-deoxyglucose uptake by interleukin-1 and insulin-like growth factor-1 in KB cells. Biochem J 311:735–738
Pang L, Sawada T, Decker SJ, Saltiel AR 1995 Inhibition of MAP kinase kinase blocks the differentiation of PC-12 cells induced by nerve growth factor. J Biol Chem 270:13585–13588
Davies SP, Reddy H, Caivano M, Cohen P 2000 Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351:96–105
Garcia-Barrado MJ, Ravier MA, Rolland JF, Gilon P, Nenquin M, Henquin JC 2001 Inhibition of protein synthesis sequentially impairs distinct steps of stimulus-secretion coupling in pancreatic ? cells. Endocrinology 142:299–307
Squires MS, Nixon PM, Cook SJ 2002 Cell-cycle arrest by PD184352 requires inhibition of extracellular signal-regulated kinases (ERK) 1/2 but not ERK5/BMK1. Biochem J 366:673–680
Asfari M, Janjic D, Meda P, Li G, Halban PA, Wollheim CB 1992 Establishment of 2-mercaptoethanol-dependent differentiated insulin-secreting cell lines. Endocrinology 130:167–178
Ishihara H, Asano T, Tsukuda K, Katagiri H, Inukai K, Anai M, Kikuchi, M, Yazaki Y, Miyazaki J, Oka Y 1994 Overexpression of hexokinase I but not GLUT1 glucose transporter alters concentration dependence of glucose-stimulated insulin secretion in pancreatic ?-cell line MIN6. J Biol Chem 269:3081–3087
Poitout V, Olson LK, Robertson RP 1996 Insulin-secreting cell lines: classification, characteristics and potential applications. Diabetes Metab 22:7–14
Dalle S, Fontes G, Lajoix AD, LeBrigand L, Gross R, Ribes G, Dufour M, Barry L, LeNguyen D, Bataille D 2002 Miniglucagon (glucagon 19–29): a novel regulator of pancreatic islet physiology. Diabetes 51:406–412
Sharp GW 1996 Mechanisms of inhibition of insulin release. Am J Physiol 271:1781–1799
Aspinwall CA, Qian WJ, Roper MG, Kulkarni RN, Kahn CR, Kennedy RT 2000 Roles of insulin receptor substrate-1, phosphatidylinositol 3-kinase, and release of intracellular Ca2+ stores in insulin-stimulated insulin secretion in ?-cells. J Biol Chem 275:22311–22338
Leibiger IB, Leibiger B, Berggren PO 2002 Insulin feedback action on pancreatic ?-cell function. FEBS Lett 532:1–6
Brubaker PL, Drucker DJ 2004 Glucagon-like peptides regulate cell proliferation and apoptosis in the pancreas, gut, and central nervous system. Endocrinology 145:2653–2659
Trümper A, Trümper K, H?rsh D 2002 Mechanisms of mitogenic and anti-apoptotic signaling by glucose-dependent insulinotropic polypeptide in ?(INS-1)-cells. J Endocrinol 174:233–246
Ehses JA, Pelech SL, Pederson RA, McIntosh CH 2002 Glucose-dependent insulinotropic polypeptide activates the Raf-Mek1/2-ERK1/2 module via a cyclic AMP/cAMP-dependent protein kinase/Rap1-mediated pathway. J Biol Chem 277:37088–37097
Maedler K, Storling J, Sturis J, Zuellig RA, Spinas GA, Arkhammar POG, Mandrup-Poulsen T, Donath MY 2004 Glucose- and interleukin-1?-induced ?-cell apoptosis requires Ca2+ influx and extracellular signal-regulated kinase (ERK)1/2 activation and is prevented by a sulfonylurea receptor1/inwardly rectifying K+ channel 6.2 (SUR/KIR6.2) selective potassium channel opener in human islets. Diabetes 53:1706–1713(Christine Longuet, Christ)