Portal de Periódicos da CAPES

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 β-Cells

2002; Elsevier BV; Volume: 277; Issue: 50 Linguagem: Inglês

10.1074/jbc.m209165200

1083-351X

Edith Gomez, Catrin Pritchard, Terence P. Herbert,

Calcium signaling and nucleotide metabolism

Glucagon like peptide-1 (GLP1) is a Gs-coupled receptor agonist that exerts multiple effects on pancreatic β-cells, including the stimulation of insulin gene expression and secretion. In this report, we show that treatment of the mouse pancreatic β-cell line MIN6 with GLP1 leads to the glucose-dependent activation of Erk. These effects are mimicked by forskolin, a direct activator of adenylate cyclase, and blocked by H89, an inhibitor of cAMP-dependent protein kinase. Additionally, we provide evidence that GLP1-stimulated activation of Erk requires an influx of calcium through L-type voltage-gated calcium channels and the activation of calcium/calmodulin-dependent protein kinase II. GLP1-stimulated activation of Erk is blocked by inhibitors of MEK, but GLP1 does not induce the activation of A-Raf, B-Raf, C-Raf, or Ras. Additionally, dominant negative forms of Ras(N17) and Rap1(N17) fail to block GLP1-stimulated activation of Erk. In conclusion, our results indicate that, in the presence of stimulatory concentrations of glucose, GLP1 stimulates the activation of Erk through a mechanism dependent on MEK but independent of both Raf and Ras. This requires 1) the activation of cAMP-dependent protein kinase, 2) an influx of extracellular Ca2+ through L-type voltage-gated calcium channels, and 3) the activation of CaM kinase II. Glucagon like peptide-1 (GLP1) is a Gs-coupled receptor agonist that exerts multiple effects on pancreatic β-cells, including the stimulation of insulin gene expression and secretion. In this report, we show that treatment of the mouse pancreatic β-cell line MIN6 with GLP1 leads to the glucose-dependent activation of Erk. These effects are mimicked by forskolin, a direct activator of adenylate cyclase, and blocked by H89, an inhibitor of cAMP-dependent protein kinase. Additionally, we provide evidence that GLP1-stimulated activation of Erk requires an influx of calcium through L-type voltage-gated calcium channels and the activation of calcium/calmodulin-dependent protein kinase II. GLP1-stimulated activation of Erk is blocked by inhibitors of MEK, but GLP1 does not induce the activation of A-Raf, B-Raf, C-Raf, or Ras. Additionally, dominant negative forms of Ras(N17) and Rap1(N17) fail to block GLP1-stimulated activation of Erk. In conclusion, our results indicate that, in the presence of stimulatory concentrations of glucose, GLP1 stimulates the activation of Erk through a mechanism dependent on MEK but independent of both Raf and Ras. This requires 1) the activation of cAMP-dependent protein kinase, 2) an influx of extracellular Ca2+ through L-type voltage-gated calcium channels, and 3) the activation of CaM kinase II. The peptide hormone GLP1 1The abbreviations used are: GLP1, glucagon-like peptide-1; GLP1-R, GLP1 receptor; MIN6, mouse insulinoma cell line 6; PKA, cAMP-dependent protein kinase; TPA, 12-O-tetradecanoylphorbol 13-acetate; EGF, epidermal growth factor; Erk, extracellular signal-regulated kinase; MEK, mitogen-activated protein kinase/Erk kinase; PKC, protein kinase C; VGCC, voltage-gated calcium channel; KRB, Krebs-Ringer bicarbonate buffer; MEF, mouse embryonic fibroblast; GST, glutathioneS-transferase; PI, phosphatidylinositol; PTP, protein-tyrosine phosphatases; MBP, myelin basic protein; EGFP, enhanced green fluorescent protein; CaM kinase II, calcium/calmodulin-dependent protein kinase II. is secreted from intestinal L-cells in response to oral ingestion of nutrients such as carbohydrates and fats (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar). A major target for GLP1 action is the pancreatic β-cell, where it exerts multiple effects including the stimulation of β-cell proliferation, differentiation, insulin gene transcription, and the potentiation of glucose dependent insulin secretion (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar). These effects are mediated through the binding of GLP1 to its receptor (GLP1-R), a member of the secretin/glucagon/vasoactive intestinal peptide G-protein-coupled receptor superfamily, that can couple to Gαs-containing heterotrimeric G-proteins leading to the activation of adenylate cyclase and the increase in the production of cAMP (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar). However, it has also been reported that GLP1-R can also signal through Gαi and Gαqproteins (2Montrose-Rafizadeh C. Avdonin P. Garant M.J. Rodgers B.D. Kole S. Yang H. Levine M.A. Schwindinger W. Bernier M. Endocrinology. 1999; 140: 1132-1140Google Scholar). In pancreatic β-cells, the binding of GLP1 to its receptor not only increases cAMP but also induces a rise in intracellular free Ca2+ levels through the closure of ATP sensitive K+ channels, an increase in Ca2+ influx through L-type voltage-gated calcium channels (L-type VGCC), and Ca2+-dependent calcium release from intracellular stores (3Holz G.G. Leech C.A. Habener J.F. J. Biol. Chem. 1995; 270: 17749-17757Google Scholar, 4Holz G.G. Leech C.A. Heller R.S. Castonguay M. Habener J.F. J. Biol. Chem. 1999; 274: 14147-14156Google Scholar, 5Bode H.P. Moormann B. Dabew R. Goke B. Endocrinology. 1999; 140: 3919-3927Google Scholar). Additionally, a number of kinases are activated upon GLP1 binding to their receptor, including phosphatidylinositol 3-kinase (PI 3-kinase) and the extracellular regulated kinase, Erk (2Montrose-Rafizadeh C. Avdonin P. Garant M.J. Rodgers B.D. Kole S. Yang H. Levine M.A. Schwindinger W. Bernier M. Endocrinology. 1999; 140: 1132-1140Google Scholar, 6Buteau J. Foisy S. Rhodes C.J. Carpenter L. Biden T.J. Prentki M. Diabetes. 2001; 50: 2237-2243Google Scholar, 7Buteau J. Roduit R. Susini S. Prentki M. Diabetologia. 1999; 42: 856-864Google Scholar). In a number of cell types, Erk activation has been shown to be important in many cellular processes including mitosis, cell differentiation, apoptosis, and the modulation of gene expression. Therefore, Erk activation is likely to be important in mediating a number of the effects induced by GLP1 in pancreatic β-cells. However, the mechanism by which GLP1 activates Erk in pancreatic β-cells is poorly understood. Typically, Erk1 and Erk2 are activated upon phosphorylation by the dual specificity tyrosine serine kinases, MEK1 and MEK2, which are themselves activated by one of several Raf isoforms, Raf1 (C-Raf), B-Raf, and A-Raf (8Kolch W. Biochem. J. 2000; 351: 289-305Google Scholar). The mechanism of Raf activation is complex and not fully understood; however, its activation is known to require the binding of the small G-proteins Ras or Rap1 in their GTP ligated forms (8Kolch W. Biochem. J. 2000; 351: 289-305Google Scholar). In many cell types, drugs or hormones that increase intracellular cAMP inhibit Erk signaling (9Chen J. Iyengar R. Science. 1994; 263: 1278-1281Google Scholar, 10Wu J. Dent P. Jelinek T. Wolfman A. Weber M.J. Sturgill T.W. Science. 1993; 262: 1065-1069Google Scholar, 11Cook S.J. McCormick F. Science. 1993; 262: 1069-1072Google Scholar, 12Qiu W. Zhuang S. von Lintig F.C. Boss G.R. Pilz R.B. J. Biol. Chem. 2000; 275: 31921-31929Google Scholar). However, in other cell types, such as neuronal cells, increased intracellular cAMP leads to an increase in Erk activation (13Vossler M.R. Yao H. York R.D. Pan M.G. Rim C.S. Stork P.J. Cell. 1997; 89: 73-82Google Scholar, 14Dugan L.L. Kim J.S. Zhang Y. Bart R.D. Sun Y. Holtzman D.M. Gutmann D.H. J. Biol. Chem. 1999; 274: 25842-25848Google Scholar). cAMP activation or inhibition of Erk is thought to occur through the activation of either cAMP-dependent protein kinase (PKA) or the cAMP-responsive Ras guanine nucleotide exchange factor, Epac (15Stork P.J. Schmitt J.M. Trends Cell Biol. 2002; 12: 258-266Google Scholar, 16de Rooij J. Zwartkruis F.J. Verheijen M.H. Cool R.H. Nijman S.M. Wittinghofer A. Bos J.L. Nature. 1998; 396: 474-477Google Scholar). In cells that mainly express Raf1, the activation of Rap1 is thought to inhibit Erk activation (17Schmitt J.M. Stork P.J. Mol. Cell. Biol. 2001; 21: 3671-3683Google Scholar, 18Naor Z. Benard O. Seger R. Trends Endocrinol. Metab. 2000; 11: 91-99Google Scholar). However, in other cells types, the activation of Rap1 leads to the activation of B-Raf, which in turn activates Erk (12Qiu W. Zhuang S. von Lintig F.C. Boss G.R. Pilz R.B. J. Biol. Chem. 2000; 275: 31921-31929Google Scholar, 13Vossler M.R. Yao H. York R.D. Pan M.G. Rim C.S. Stork P.J. Cell. 1997; 89: 73-82Google Scholar, 14Dugan L.L. Kim J.S. Zhang Y. Bart R.D. Sun Y. Holtzman D.M. Gutmann D.H. J. Biol. Chem. 1999; 274: 25842-25848Google Scholar). In this report, we provide evidence that, in the mouse pancreatic β-cell line MIN6, GLP1 stimulates the activation of Erk through a mechanism dependent of MEK but independent of both Raf and Ras. This requires 1) the activation of PKA, 2) an influx of extracellular Ca2+ through L-type voltage-gated calcium channels, and 3) the activation of CaM kinase II. PD098059, KN-93, KN-62, nifedipine, and Bay K8644 were all purchased from Calbiochem. U0126 was purchased from Promega. All other chemicals (unless stated) were obtained from Sigma. In this study, MIN6 cells were used between passages 25 and 35 at ∼80% confluence. MIN6 cells were grown in Dulbecco's modified Eagle's medium containing 25 mm glucose supplemented with 15% heat-inactivated fetal calf serum, 100 μg/ml streptomycin, 100 units/ml penicillin sulfate, and 75 μm β-mercaptoethanol, equilibrated with 5% CO2, 95% air at 37 °C. Prior to treatment, the medium was removed and the cells washed twice with HEPES-balanced Krebs-Ringer bicarbonate buffer (115 mm NaCl, 5 mm KCl, 10 mm NaHCO3, 2.5 mmMgCl2, 2.5 mm CaCl2, 20 mm HEPES, pH 7.4) containing 0.5% bovine serum albumin (KRB buffer). The cells were then incubated for 1 h at 37 °C in KRB buffer containing 1 mm glucose prior to incubation in KRB buffer containing 2.8 or 16.7 mm glucose and the test substances for the times indicated in the figure legends (details of treatments are provided in the figure legends). When cells were treated with elevated extracellular K+concentration, the K+ concentration in the KRB was increased to 50 mm and the Na+ concentration decreased to 70 mm to maintain isotonicity. In the calcium-free experiments, after preincubation the cells were incubated for 15 min in a calcium-free KRB buffer containing 1 mmEGTA and stimulated in the same buffer. All treatments were stopped by the addition of ice-cold lysis buffer containing 1% Triton, 10 mm β-glycerophosphate, 50 mm Tris-HCl, pH 7.5, 1 mm EDTA, 1 mm EGTA, 1 mmsodium orthovanadate, 1 mm benzamidine HCl, 0.2 mm phenylmethylsulfonyl fluoride, 1 μg/ml each of leupeptin and pepstatin, 0.1% β-mercaptoethanol, and 50 mm sodium fluoride. The lysates were then centrifuged for 10 min at 16,000 × g. The supernatants were kept, and total protein concentrations were determined by the Bradford assay (Bio-Rad) using bovine serum albumin as standard. The protein lysates were stored at −80 °C until further analysis. SDS-PAGE and Western blotting were performed as described previously (19Herbert T.P. Kilhams G.R. Batty I.H. Proud C.G. J. Biol. Chem. 2000; 275: 11249-11256Google Scholar). Rabbit anti-phospho-Erk1/2 antibody recognizing only the activated forms of Erk was purchased from New England Biolabs. Mouse anti-Ras antibody was purchased from Transduction Laboratories. Detection was by horseraddish peroxidase-linked anti-rabbit/mouse secondary antibodies and enhanced chemiluminescence (Amersham Biosciences). Mouse embryonic fibroblasts (MEFs) wild type (+/+) or knock-out for B-Raf(−/−) were used as positive and negative controls for this assay. MEFs were serum-starved in Dulbecco's modified Eagle's medium containing 0.5% (v/v) fetal calf serum for 24 h prior to stimulation by 10 ng/ml EGF for 5 min. Details of the MIN6 cell treatments are provided in the legend to Fig. 5b (and see section above “Cell Culture and Treatment”). Protein lysates were prepared as described by Maraiset al. (20Marais R. Light Y. Paterson H.F. Mason C.S. Marshall C.J. J. Biol. Chem. 1997; 272: 4378-4383Google Scholar). Raf proteins were immunoprecipitated for 2 h at 4 °C from 0.1 (for B-Raf) to 1 mg (for C-Raf and A-Raf) of cell extract with 2 μg of anti-Raf-1 antibody (Transduction Laboratories), 4 μg of anti-B-raf rabbit polyclonal antibody (21Mason C.S. Springer C.J. Cooper R.G. Superti-Furga G. Marshall C.J. Marais R. EMBO J. 1999; 18: 2137-2148Google Scholar), and 2 μg of anti-A-Raf mouse monoclonal antibody (Transduction Laboratories). The activity of each Raf protein was assessed by using the Raf kinase cascade assay using glutathione S-transferase (GST)-MEK, GST-ERK, and MBP as sequential substrates together with [γ-32P]ATP (20Marais R. Light Y. Paterson H.F. Mason C.S. Marshall C.J. J. Biol. Chem. 1997; 272: 4378-4383Google Scholar). After incubation at 30 °C for 30 min under constant agitation, the kinase reactions were stopped by addition of Laemmli sample buffer and boiled 5 min at 100 °C. The proteins were separated on a 15% SDS-polyacrylamide gel. The gel was fixed, dried, and the phosphorylation state of MBP, reflecting the activation state of the upstream kinase Raf, revealed by autoradiography. GST-Raf-1-Ras binding domain was expressed in Escherichia coli, affinity-purified using glutathione-Sepharose beads (Amersham Biosciences), and then used to measure Ras activation in MIN6 cells and MEF according to the protocol found in the Upstate Biotechnology website, www.upstatebiotech.com (Lake Placid, NY). Briefly, serum-starved MEF or glucose-starved MIN6 cells (see “Cell Culture and Treatment for Details”) were stimulated with different agonists for 5 min and immediately lysed. Equal amounts of supernatant were incubated with Sepharose-coupled GST-Raf-1-Ras binding domain at 4 °C for 30 min. The bound proteins were separated on a 15% SDS-polyacrylamide gel, and the activated Ras was then visualized by immunoblotting using mouse anti-Ras antibody (Transduction Laboratories). Recombinant adenovirus containing Rap1aN17 (AdRapN17.EGFP) was prepared using the Adeasy system (22He T.C. Zhou S. da Costa L.T. Yu J. Kinzler K.W. Vogelstein B. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 2509-2514Google Scholar). The adenoviral and shuttle vectors were provided by Prof. Tong-Chuan He, Howard Hughes Medical Institute and Johns Hopkins Oncology Center, and the pMT2Ha-RapN17 plasmid was provided by Dr. J. L. Bos, University Medical Centre Utrecht, Utrecht University, Utrecht, Netherlands. Briefly, the RapN17 sequence was amplified by PCR from the pMT2Ha-RapN17 plasmid as a HindIII/EcoRV fragment and ligated into a shuttle vector, pAd-Track.CMV, under the control of the cytomegalovirus (CMV) promoter. Co-transformation and recombination of the above construct with the adenoviral vector pAdEasy-1 resulted in a recombinant adenoviral plasmid, pAdRapN17.EGFP. The adenoviral plasmid was then transfected into HEK 293 cells, and adenovirus AdRapN17.EGFP was obtained by extracting the cells 7–10 days after transfection. EGFP expression was used to monitor the production of the virus following several round of amplification into HEK 293 cells. The control adenovirus expressing EGFP (AdEmpty.EGFP, a gift from Dr. J. Carter, University of Leicester) and the recombinant adenovirus AdRasN17 containing RasN17 (a gift from Prof. B. Khan and Dr. C. Sutherland) were amplified in a similar manner. To infect MIN6 cells, AdRapN17.EGFP, AdRasN17, or AdEmpty.EGFP were mixed with culture medium, and cells were exposed to the viruses with 100–200 multiplicity of infection for 24 h prior to perform the experiment. Under these conditions, >90% of cells were infected, as determined by EGFP expression. To investigate the mechanism by which GLP1 activates Erk, we initially characterized the activation of Erk in response to GLP1 in the pancreatic β-cell line MIN6, a cell line that synthesizes and secretes insulin in response to physiologically relevant glucose concentrations (23Ishihara H. Asano T. Tsukuda K. Katagiri H. Inukai K. Anai M. Kikuchi M. Yazaki Y. Miyazaki J.I. Oka Y. Diabetologia. 1993; 36: 1139-1145Google Scholar, 24Skelly R.H. Schuppin G.T. Ishihara H. Oka Y. Rhodes C.J. Diabetes. 1996; 45: 37-43Google Scholar). MIN6 cells were preincubated for 1 h in KRB supplemented with 1 mm glucose. The cells were then incubated in 2.8 or 16.7 mm glucose in the presence of absence of either GLP1 or forskolin, an activator of adenylate cyclase. Samples were taken over a 30-min time course and the activation state of Erk investigated using a phospho-specific antibody against the activated forms of Erk. No activation of Erk was detected when MIN6 cells were incubated in KRB supplemented with 2.8 mmglucose (Fig. 1). However, when MIN6 cells were incubated in KRB supplemented with 2.8 mmglucose in the presence of GLP1, a transient increase in the activation of Erk was detected at 5 min (Fig. 1). Incubation of cells in KRB supplemented with 16.7 mm glucose led to an activation of Erk compared with cells incubated at 2.8 mm glucose (Fig. 1). This is in agreement with previous reports indicating that glucose can activate Erk in MIN6 cells (25Benes C. Roisin M.P. Van Tan H. Creuzet C. Miyazaki J. Fagard R. J. Biol. Chem. 1998; 273: 15507-15513Google Scholar, 26Benes C. Poitout V. Marie J.C. Martin-Perez J. Roisin M.P. Fagard R. Biochem. J. 1999; 340: 219-225Google Scholar). However, GLP1 treatment in the presence of 16.7 mm glucose led to a robust, rapid, and sustained increase in Erk activation, with maximal activation between 5 and 15 min. GLP1 stimulation of Erk was mimicked by forskolin at either 2.8 or 16.7 mm glucose (Fig. 1). As GLP1 or forskolin can activate adenylate cyclase, and both activate Erk via a glucose-dependent mechanism, the activation of Erk by GLP1 or forskolin is likely to be mediated through similar pathways. GLP1 or forskolin potentiates glucose-dependent insulin secretion (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar), therefore glucose-dependent GLP1 activation of Erk could be mediated by an autocrine effect of insulin. To investigate this possibility, MIN6 cells were incubated in KRB supplemented with 2.8 or 16.7 mm glucose and treated with 1 μm insulin in the presence or absence of GLP1. Insulin treatment did not lead to Erk activation and had no stimulatory effect on GLP1 induced Erk activation. This indicates that GLP1 activation of Erk is not mediated by insulin (results not shown). Since glucose-dependent GLP1 activation of Erk is mimicked by forskolin, it is likely to be mediated through the activation of adenylate cyclase. To determine the role of PKA in GLP1 activation of Erk, MIN6 cells were treated with GLP1 or as a control forskolin, in KRB supplemented with 16.7 mm glucose in either the presence or absence of the PKA inhibitor, H89, at a concentration shown to block forskolin stimulation of Erk. H89 inhibited both GLP1- and forskolin-stimulated Erk activation (Fig. 2). As GLP1-stimulated activation of Erk is mimicked by forskolin, an activator of adenylate cyclase, and inhibited by H89, an inhibitor of PKA, it is probable that GLP1-stimulated activation of Erk is mediated by the activation of adenylate cyclase, leading to an increase in intracellular cAMP and the subsequent activation of PKA. Glucose metabolism in pancreatic β-cells leads to an increase in intracellular calcium through an influx of calcium through L-type VGCC, which is augmented by hormones such as GLP1 (27Holz G.G. Habener J.F. Trends Biochem. Sci. 1992; 17: 388-393Google Scholar, 28Holz G.G. Kuhtreiber W.M. Habener J.F. Nature. 1993; 361: 362-365Google Scholar). This rise in intracellular Ca2+ regulates a number of important pancreatic β-cell functions including insulin secretion (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar). To determine whether Ca2+ influx is an important determinant in GLP1-stimulated Erk activation, MIN6 cells were incubated at 16.7 mm glucose in the presence or absence of EGTA to chelate extracellular calcium (Fig. 3a). GLP1 treatment led to the activation of Erk, which was inhibited by the presence EGTA (Fig. 3a). Therefore, GLP1 requires an influx of extracellular calcium to stimulate Erk. To investigate whether an increase in intracellular calcium is sufficient for GLP1-stimulated Erk activation, MIN6 cells were treated with the Ca2+ ionophore, ionomycin, to artificially raise intracellular calcium concentrations, in the presence or absence of GLP1 at 2.8 or 16.7 mm glucose (Fig. 3b). Treatment of MIN6 cells with ionomycin in the presence of 2.8 or 16.7 mm glucose had no effect on Erk activation. Treatment of MIN6 with ionomycin and GLP1 at 2.8 mm glucose also had no effect on Erk activation (Fig. 3b). Additionally, GLP1-stimulated Erk activation in the presence of 16.7 mm glucose was unaffected by the presence of ionomycin (Fig. 3b). These results indicate that an influx of extracellular calcium is not sufficient for GLP1-stimulated Erk activation. However, it is possible that the mode of calcium entry is important in GLP1-stimulated Erk activation and that Ca2+must enter through L-type VGCC to be effective in signaling to Erk. To investigate this possibility, MIN6 cells were treated with GLP1 at 16.7 mm glucose in the presence or absence of nifedipine, an L-type VGCC blocker. GLP1-stimulated activation of Erk was inhibited by the presence of nifedipine (Fig. 3c). Therefore, GLP1 stimulation of Erk is mediated through an influx of extracellular calcium through L-type VGCC. To determine whether Ca2+influx through L-type VGCC was sufficient for Erk activation, cells were incubated in the presence of 1) increased extracellular K+ concentration, which leads to the depolarization of the membrane, the opening of L-type VGCC, and the influx of extracellular Ca2+ (Fig. 3d) or 2) Bay K8644, an L-type VGCC agonist (Fig. 3e). Incubation of MIN6 cells in 50 mm extracellular K+ led to a strong stimulation of Erk (Fig. 3d). To demonstrate that this was mediated through Ca2+ influx through L-type VGCC, the experiment was also conducted in the presence of nifedipine or EGTA. Both treatments inhibited K+-induced Erk activation (Fig. 3d). This is in agreement with previous reports demonstrating that K+-induced Erk activation in MIN6 cells requires the influx of Ca2+ through L-type VGCC (25Benes C. Roisin M.P. Van Tan H. Creuzet C. Miyazaki J. Fagard R. J. Biol. Chem. 1998; 273: 15507-15513Google Scholar, 26Benes C. Poitout V. Marie J.C. Martin-Perez J. Roisin M.P. Fagard R. Biochem. J. 1999; 340: 219-225Google Scholar). Additionally, treatment of MIN6 cells with Bay K8644 alone was sufficient to induce Erk activation (Fig. 3e). Taken together, these results indicate that an influx of Ca2+through L-Type VGCC is sufficient for Erk activation and is necessary for GLP1-stimulated Erk activation. However, it is unclear whether the influx of Ca2+ through L-type VGCC is both necessary and sufficient for GLP1-stimulated Erk activation. Indeed, GLP1 may be providing additional signals, which are important for GLP1-stimulated Erk activation. GLP1 stimulation of Erk requires an influx of extracellular calcium through L-type VGCC. Many of the effects of calcium on intracellular signaling are mediated through Ca2+-binding proteins such as calmodulin. Therefore, we initially investigated the role of calmodulin in GLP1-stimulated Erk activation. MIN6 cells were preincubated for 45 min with W7, a competitive inhibitor of calmodulin, prior to incubation in KRB containing 16.7 mm glucose in the presence of GLP1 (Fig. 4a). Treatment of cells with W7 led to a dose-dependent inhibition of GLP1-stimulated Erk activation indicating that calmodulin is likely to be required for GLP1-stimulated Erk activation (Fig. 4a). Given the important role calmodulin plays in the activation of calcium/calmodulin-dependent kinases (CaM kinases) and the role CaM kinase II plays in insulin secretion in pancreatic β-cells, we investigated the potential role of CaM kinase II in GLP1-mediated Erk activation using selective inhibitors of CaM kinase II, KN62 and KN93 (29Williams C.L. Phelps S.H. Porter R.A. Biochem. Pharmacol. 1996; 51: 707-715Google Scholar, 30Fan G.H. Wang L.Z. Qiu H.C. Ma L. Pei G. Mol. Pharmacol. 1999; 56: 39-45Google Scholar). MIN6 cells were incubated with either KN62 or KN93 for 30 min or 1 h, respectively, prior to incubation in KRB supplemented with 16.7 mm glucose and GLP1 (Fig. 4,b and c). Treatment of cells with either KN62 or KN93 caused a dose-dependent inhibition of GLP1-stimulated Erk activation (Fig. 4, b and c). These results indicate that calmodulin and the activation of CaM kinase II may be required for glucose-dependent GLP1 activation of Erk. To further investigate the signaling pathway by which GLP1 stimulates Erk activation, we investigated the role of MEK, the upstream kinase of Erk, using two structurally distinct specific inhibitors of MEK, PD098059 and U0126 (Fig. 5a). MIN6 cells were pretreated with either PD098059 or U0126 prior to treatment with either GLP1 or forskolin in the presence of 16.7 mm glucose, or 1 μm TPA as control. The activation of Erk was monitored using a phospho-specific antibody to Erk (Fig. 5a). Both PD098059 and U0126 inhibited GLP1- and forskolin-stimulated Erk activation, indicating that the activation of MEK is necessary for GLP1- and forskolin-induced Erk activation (Fig. 5a). MEK is phosphorylated and activated by several isoforms of Raf including Raf1 (C-Raf), B-Raf, and A-Raf. Therefore, we measured the activation of all three endogenous isoforms of Raf upon treatment with GLP1 using the immunoprecipitation kinase cascade assay (Fig. 5b) (20Marais R. Light Y. Paterson H.F. Mason C.S. Marshall C.J. J. Biol. Chem. 1997; 272: 4378-4383Google Scholar). MIN6 cells were treated for the times indicated in Fig. 5b, with 10 nm GLP1 or 10 μm forskolin in the presence of 16.7 mm glucose or 1 μm TPA as control. Protein lysates were prepared, and Raf proteins were immunoprecipitated using antibodies specific to the three Raf isoforms (A-Raf, B-Raf, and C-Raf). The immunoprecipitates were incubated with GST-MEK, GST-ERK, MBP, and [γ-32P]ATP and the activity of Raf assessed by monitoring the phosphorylation state of MBP. MEFs (immortalized mouse embryonic fibroblasts) wild type (+/+) or knock-out for B-raf (−/−), treated or not with 10 ng/ml EGF, were used as positive and negative controls for this experiment (Fig. 5b) (31Huser M. Luckett J. Chiloeches A. Mercer K. Iwobi M. Giblett S. Sun X.M. Brown J. Marais R. Pritchard C. EMBO J. 2001; 20: 1940-1951Google Scholar). In agreement with previous studies (31Huser M. Luckett J. Chiloeches A. Mercer K. Iwobi M. Giblett S. Sun X.M. Brown J. Marais R. Pritchard C. EMBO J. 2001; 20: 1940-1951Google Scholar), B-Raf had elevated basal activity in unstimulated cells, whereas both Raf-1 and A-Raf were inactive (data not shown). As expected, no B-Raf activity was detected in the MEF B-raf(−/−) cells. EGF stimulated the activities of the three Raf isoforms (B-Raf, A-Raf, and C-Raf) in the B-raf(+/+) MEFs and the activities of A-Raf and C-Raf in the B-raf knock-out cells. In the MIN6 cell line, TPA stimulation led to the activation of both A-Raf and C-Raf kinases (Fig. 5b). In contrast, conditions that lead to the activation of Erk, glucose plus GLP-1 or glucose plus forskolin, did not lead to the activation of any of the three Raf isoforms, A-Raf, B-Raf, or C-Raf (Fig. 5b). To further support these findings we investigated the effect of overexpressing dominant negative Rap1(N17) or Ras(N17) on GLP1 stimulation of Erk. MIN6 cells were mock-infected or infected with recombinant adenoviruses expressing either empty vector (AdEmpty.EGFP), Rap1N17 (AdRap1N17.EGFP), or RasN17 (AdRasN17). 24 h post-infection, the cells were treated for 5 min with 16.7 mm glucose alone or 16.7 mm glucose plus GLP1 or EGF and the activation status of Erk determined (Fig. 5c). The overexpression of dominant negative Rap1 had no detectable inhibitory effect on the ability of GLP1 or EGF to stimulate Erk activation (Fig. 5c). The overexpression of dominant negative Ras had no detectable inhibitory effect on the ability of GLP1 to stimulate Erk activation but effectively inhibited EGF-stimulated Erk activation (Fig. 5c). Therefore, GLP1-stimulated Erk activation appears to be independent of both Ras and Raf. As an adjunct to the above, we also investigated the activation of Ras in response to agonist stimulation. Ras activation was assayed with a GST-Raf1-Ras binding domain fusion protein, which associates exclusively with the GTP-bound form of Ras, but not with the GDP-bound form of Ras in vitro. Stimulation of MIN6 cells with GLP1 did not lead to an increase in the binding of GTP-bound Ras to GST-Raf1-Ras binding domain compared with cells incubated in 16.7 mm glucose alone (Fig. 5d, panel i). This indicates that Ras is not activated by GLP1 (Fig. 5d,panel i). In contrast, EGF or TPA treatment of MIN6 cells and EGF treatment of wild type MEFs caused a significant increase in the activation of Ras as shown by the increased amount of the GTP-bound form of Ras bound to the GST-Raf1-Ras binding domain (Fig. 5d, panel i). Taken together, these data indicate that glucose-dependent GLP1 stimulation of Erk is independent of the activation of A-Raf, B-Raf, C-Raf, or Ras. In pancreatic β-cells, GLP1 activates the extracellular regulated kinase, Erk (2Montrose-Rafizadeh C. Avdonin P. Garant M.J. Rodgers B.D. Kole S. Yang H. Levine M.A. Schwindinger W. Bernier M. Endocrinology. 1999; 140: 1132-1140Google Scholar, 6Buteau J. Foisy S. Rhodes C.J. Carpenter L. Biden T.J. Prentki M. Diabetes. 2001; 50: 2237-2243Google Scholar, 7Buteau J. Roduit R. Susini S. Prentki M. Diabetologia. 1999; 42: 856-864Google Scholar), via a poorly understood mechanism. In this report, we provide evidence that, in the pancreatic β-cell line MIN6, glucose-dependent GLP1 activation of Erk is mediated by an influx of calcium through L-type VGCC and requires both the activation of PKA and calcium/calmodulin-dependent protein kinase II (CaM kinase II). Furthermore, we show that GLP1 stimulates the activation of Erk via a mechanism independent of both Raf and Ras but dependent on the activation of MEK (Fig. 6). Glucose metabolism in pancreatic β-cells leads to an influx of calcium through L-type VGCC, which is augmented by hormones such as GLP1 (27Holz G.G. Habener J.F. Trends Biochem. Sci. 1992; 17: 388-393Google Scholar, 28Holz G.G. Kuhtreiber W.M. Habener J.F. Nature. 1993; 361: 362-365Google Scholar). The rise in intracellular Ca2+ regulates a number of important pancreatic β-cell functions including insulin secretion (1Perfetti R. Merkel P. Eur. J. Endocrinol. 2000; 143: 717-725Google Scholar). Glucose-dependent GLP1 stimulation of Erk also requires Ca2+ entry specifically through L-type VGCC (Fig. 3). The entry of Ca2+ through L-type VGCC is also required for glucose- or K+-induced activation of Erk in MIN6 cells (25Benes C. Roisin M.P. Van Tan H. Creuzet C. Miyazaki J. Fagard R. J. Biol. Chem. 1998; 273: 15507-15513Google Scholar, 26Benes C. Poitout V. Marie J.C. Martin-Perez J. Roisin M.P. Fagard R. Biochem. J. 1999; 340: 219-225Google Scholar). Indeed, it is both necessary and sufficient for Erk activation, as Bay K8644, an L-type VGCC agonist, activates Erk and nifedipine, an L-type VGCC blocker, inhibits Erk activation induced by an increase in extracellular K+ concentration (Fig. 3,d and e). Since the influx of Ca2+through L-type VGCC alone is sufficient for the activation of Erk, and GLP1 augments glucose-stimulated Ca2+ influx through L-type VGCC (4Holz G.G. Leech C.A. Heller R.S. Castonguay M. Habener J.F. J. Biol. Chem. 1999; 274: 14147-14156Google Scholar), it is possible that the increase influx of calcium is the principle signal by which GLP1 induces glucose-dependent Erk activation. However, it is equally possible that GLP1 activates additional signaling pathways that are important in the activation of Erk. In neurons, the influx of calcium through L-type VGCC also leads to the activation of Erk. This mode of Ca2+ entry is critical in determining which signaling pathway is activated and is thought to be mediated through the local increase in Ca2+concentration around the mouth of the L-type VGCC resulting in the binding of Ca2+/calmodulin to the L-type VGCC and the subsequent activation of Erk (32Dolmetsch R.E. Pajvani U. Fife K. Spotts J.M. Greenberg M.E. Science. 2001; 294: 333-339Google Scholar). As the calmodulin agonist W7 blocks GLP1-stimulated Erk activation (Fig. 4a), it is possible that GLP1 stimulates Erk in pancreatic β-cells by a mechanism similar to that in neurons. Selective inhibitors of CaM kinase II, KN62 and KN93, inhibit GLP1-stimulated activation of Erk (Fig. 4, b andc), but these inhibitors have limited specificity and can interfere with both K+ and Ca2+ channel activity (33Bhatt H.S. Conner B.P. Prasanna G. Yorio T. Easom R.A. Biochem. Pharmacol. 2000; 60: 1655-1663Google Scholar). However, it has previously been demonstrated that CaM kinase II is activated by glucose in pancreatic β-cells (34Easom R.A. Filler N.R. Ings E.M. Tarpley J. Landt M. Endocrinology. 1997; 138: 2359-2364Google Scholar) and that the activation of CaM kinase II can mediate Erk activation in other cell types such as aortic and vascular smooth muscle cells (35Muthalif M.M. Benter I.F. Uddin M.R. Malik K.U. J. Biol. Chem. 1996; 271: 30149-30157Google Scholar, 36Ginnan R. Singer H.A. Am. J. Physiol. 2002; 282: C754-C761Google Scholar). Therefore, it is probable that CaM kinase II mediates GLP1-stimulated Erk activation in MIN6 cells. Whether GLP1 potentiates glucose-stimulated activation of CaM kinase II and that this is both necessary and sufficient for GLP1 activation of Erk or whether glucose activation of CaM kinase II is necessary but GLP1 provides an additional input essential for the activation of Erk is unclear. In vascular smooth muscle cells, Ca2+-dependent activation of Erk requires the activation of CaM kinase II. In this case, Erk activation is dependent on the non-receptor proline-rich tyrosine kinase (PYK2) activation, the transactivation of the EGF receptor, and the subsequent activation of Raf (36Ginnan R. Singer H.A. Am. J. Physiol. 2002; 282: C754-C761Google Scholar). However, in our report we show that GLP1-stimulated Erk activation is independent of Raf, as no activation of all three isoforms of Raf was detected in anin vitro kinase assay (Fig. 5b), and dominant negative forms of Ras or Rap1 had no effect on GLP1-stimulated Erk activation (Fig. 5c). Moreover, GLP1-stimulated Erk activation did not lead to the activation of Ras, as no increase in the association of GTP-bound Ras to Raf was detected upon GLP1 treatment compared with control (Fig. 5d). This Raf-independent activation of Erk does not require the activation of the classical PKCs or PI 3-kinase, as the general classical PKC inhibitor bisindolylmaleimide and inhibitors or PI 3-kinase, wortmannin or LY294002, failed to block GLP1-stimulated activation of Erk (results not shown). GLP1-stimulated Erk activation does require the activation of MEK as it is blocked by two structurally distinct inhibitors of MEK, PD098059 and U0126 (Fig. 5a). It is unlikely that CaM kinase II directly phosphorylates MEK at either Ser-218 or Ser-222, two sites important in MEK activation, as neither of these sites lie within the known optimal substrate recognition motif for CaM kinase II (37White R.R. Kwon Y.G. Taing M. Lawrence D.S. Edelman A.M. J. Biol. Chem. 1998; 273: 3166-3172Google Scholar). However, CaM kinase II could phosphorylate an intermediate kinase necessary for MEK activation. One possible mechanism by which Erk could be activated by cAMP in a Raf-independent mechanism is through the regulation of binding of the protein-tyrosine phosphatases (PTP), PTP-SL, STEP, and HePTP to Erk (38Saxena M. Williams S. Tasken K. Mustelin T. Nat. Cell Biol. 1999; 1: 305-311Google Scholar, 39Blanco-Aparicio C. Torres J. Pulido R. J. Cell Biol. 1999; 147: 1129-1136Google Scholar). When these protein-tyrosine phosphatases are bound to Erk, Erk is maintained in an inactive state; however, upon their activation by PKA, Erk is released and becomes activated (38Saxena M. Williams S. Tasken K. Mustelin T. Nat. Cell Biol. 1999; 1: 305-311Google Scholar, 39Blanco-Aparicio C. Torres J. Pulido R. J. Cell Biol. 1999; 147: 1129-1136Google Scholar). We thank Prof. Barbara Kahn (Beth Israel Deaconess Medical Center, Boston) and Dr. Calum Sutherland (Dundee University, Dundee, UK) for generously providing the recombinant adenovirus expressing RasN17 and Prof. J. L. Bos for providing a pMT2-HaRapN17 construct. We would also like to thank Prof. Richard Marais (Institute of Cancer Research, London, UK) for recombinant proteins and the anti-B-Raf antibody used in the Raf kinase assays.