Portal de Periódicos da CAPES

Calcium-stimulated Phosphorylation of MAP-2 in Pancreatic ॆTC3-cells Is Mediated by Ca2+/Calmodulin-dependent Kinase II

1997; Elsevier BV; Volume: 272; Issue: 43 Linguagem: Inglês

10.1074/jbc.272.43.27464

1083-351X

Kimberly A. Krueger, Harshika S. Bhatt, Michael Landt, Richard A. Easom,

Microtubule and mitosis dynamics

An understanding of the role of CaM kinase II in the pancreatic ॆ-cell is dependent on the identification of its cellular targets. One of the best substrates of CaM kinase II in vitro that could function in secretory events is the microtubule-associated protein, MAP-2. By immunoblot analysis, a high molecular weight protein with electrophoretic properties characteristic of MAP-2, was identified in rat insulinoma ॆTC3 cells and isolated rat islets. In immunoprecipitation experiments employing α-toxin-permeabilized ॆTC3 cells, elevation of intracellular Ca2+ or addition of forskolin, an adenylate cyclase activator, induced significant phosphorylation of MAP-2 in situ. The effect of Ca2+ was rapid, concentration-dependent and closely correlated with activation of CaM kinase II under similar experimental conditions. H-89, a specific and potent inhibitor of cAMP-dependent protein kinase (PKA), prevented forskolin-induced MAP-2 phosphorylation but had little effect on MAP-2 phosphorylation stimulated by elevated Ca2+. Phosphopeptide mapping revealed that the phosphorylation pattern observed in situ upon incubation of the ॆTC3 cells with increased free Ca2+, was strikingly similar to that generated in vitro by CaM kinase II, most notably with regard to the increased phosphate incorporated into one prominent site. These data provide evidence that MAP-2 is phosphorylated by CaM kinase II in the pancreatic ॆ-cell in situ, and that this event may provide an important link in the mediation of Ca2+-dependent insulin secretion. An understanding of the role of CaM kinase II in the pancreatic ॆ-cell is dependent on the identification of its cellular targets. One of the best substrates of CaM kinase II in vitro that could function in secretory events is the microtubule-associated protein, MAP-2. By immunoblot analysis, a high molecular weight protein with electrophoretic properties characteristic of MAP-2, was identified in rat insulinoma ॆTC3 cells and isolated rat islets. In immunoprecipitation experiments employing α-toxin-permeabilized ॆTC3 cells, elevation of intracellular Ca2+ or addition of forskolin, an adenylate cyclase activator, induced significant phosphorylation of MAP-2 in situ. The effect of Ca2+ was rapid, concentration-dependent and closely correlated with activation of CaM kinase II under similar experimental conditions. H-89, a specific and potent inhibitor of cAMP-dependent protein kinase (PKA), prevented forskolin-induced MAP-2 phosphorylation but had little effect on MAP-2 phosphorylation stimulated by elevated Ca2+. Phosphopeptide mapping revealed that the phosphorylation pattern observed in situ upon incubation of the ॆTC3 cells with increased free Ca2+, was strikingly similar to that generated in vitro by CaM kinase II, most notably with regard to the increased phosphate incorporated into one prominent site. These data provide evidence that MAP-2 is phosphorylated by CaM kinase II in the pancreatic ॆ-cell in situ, and that this event may provide an important link in the mediation of Ca2+-dependent insulin secretion. Circumstantial evidence supports a functional role of the multifunctional Ca2+/calmodulin-dependent protein kinase II (CaM kinase II) 1The abbreviations used are: CaM kinase II, Ca2+/calmodulin-dependent protein kinase II; MAP-2, microtubule-associated protein-2; PKA, protein kinase A; TPCK,l-1-tosylamido-2-phenylethyl chloromethyl ketone. in the regulation of insulin secretion from the pancreatic ॆ-cell. Principal within this evidence is the demonstration that glucose, the major physiological regulator of insulin secretion in rodents and humans (1Hedeskov C.J. Physiol. Rev. 1980; 60: 442-509Crossref PubMed Scopus (327) Google Scholar), activates CaM kinase II in isolated rat islets in a concentration-dependent manner (2Wenham R.M. Landt M. Easom R.A. J. Biol. Chem. 1994; 269: 4947-4952Abstract Full Text PDF PubMed Google Scholar) that temporally correlates with the initial and sustained phases of insulin secretion (3Easom R.A. Filler N.R. Ings E.I. Tarpley J. Landt M. Endocrinology. 1997; 138: 2359-2364Crossref PubMed Scopus (30) Google Scholar). Other data utilizing pharmacological inhibitors (i.e.KN-62, KN-93) of this enzyme have also implicated CaM kinase II in the regulation of insulin secretion (4Wenham R.M. Landt M. Walters S.M. Hidaka H. Easom R.A. Biochem. Biophys. Res. Commun. 1992; 189: 128-133Crossref PubMed Scopus (39) Google Scholar, 5Niki I. Okazaki K. Saitoh M. Niki A. Niki H. Tamagawa T. Iguchi A. Hidaka H. Biochem. Biophys. Res. Commun. 1993; 191: 255-261Crossref PubMed Scopus (74) Google Scholar), although conclusions made from such studies are complicated by nonspecific effects demonstrated by these drugs (6Clyne C. Nguyen A. Rainey W. Endocr. Res. 1995; 21: 259-265Crossref PubMed Scopus (16) Google Scholar, 7Li G. Hidaka H. Wollheim C.B. Mol. Pharmacol. 1992; 42: 489-498PubMed Google Scholar). Another study that reports the inability of KN-62 to inhibit Ca2+-induced insulin secretion from the permeabilized ॆ-cell (7Li G. Hidaka H. Wollheim C.B. Mol. Pharmacol. 1992; 42: 489-498PubMed Google Scholar) argues, however, against a role of CaM kinase II in the insulin secretory process. Irrespective of the relevance of CaM kinase II to the ॆ-cell secretory process, the understanding of the physiological consequence of the activation of CaM kinase II is dependent on the identification of target substrates in the ॆ-cell. A large number of cellular proteins are phosphorylated by CaM kinase II in vitro (8Braun A.P. Schulman H. Annu. Rev. Physiol. 1995; 57: 417-445Crossref PubMed Scopus (738) Google Scholar), but relatively few of these have been proven as legitimate substratesin situ. Prominent among this latter group, however, is the microtubule-associated protein-2 (MAP-2), which has been shown to be phosphorylated by CaM kinase II in GH3 cells (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar) or hippocampal slices (10Diaz-Nido J. Montoro R.J. Lopez-Barneo J. Avila J. Eur. J. Neurosci. 1993; 5: 818-824Crossref PubMed Scopus (20) Google Scholar) stimulated with depolarizing concentrations of potassium. MAP-2 is a member of a larger family of microtubule-associated proteins that have the capacity to regulate reversible polymerization and stability of microtubules through their affinity for tubulin (11Sloboda R. Dentler W. Rosenbaum J. Biochemistry. 1976; 15: 4497-4505Crossref PubMed Scopus (302) Google Scholar) as well as their interaction with other cellular structures such as actin (12Sattilaro R. Biochemistry. 1986; 25: 2003-2009Crossref PubMed Scopus (99) Google Scholar). This regulatory capacity is in turn controlled by the phosphorylation state of MAP-2, at least in vitro (13Raffaelli N. Yamauchi P.S. Purich D.L. FEBS Lett. 1992; 296: 21-24Crossref PubMed Scopus (26) Google Scholar). Although a minimal extent of MAP-2 phosphorylation appears to be essential for MAP-2 function (14Brugg B. Matus A. J. Cell Biol. 1991; 114: 735-743Crossref PubMed Scopus (177) Google Scholar), phosphorylation by specific kinases in vitrohas resulted in reduced affinity to microtubules, reduced rate and extent of assembly, accentuated disassembly, and reduced interaction of microtubules with actin filaments (15Maccioni R.B. Cambiazo V. Physiol. Rev. 1995; 75: 835-864Crossref PubMed Scopus (336) Google Scholar). In optimal conditions, isolated MAP-2 has been demonstrated to incorporate phosphate to the level of 46 mol/mol of MAP-2 (16Tsuyama S. Terayama Y. Matsuyama S. J. Biol. Chem. 1987; 262: 10886-10892Abstract Full Text PDF PubMed Google Scholar). Although MAP-2 is phosphorylated by multiple protein kinases including the phospholipid-dependent protein kinase C (17Ainsztein A.M. Purich D.L. J. Biol. Chem. 1994; 269: 28465-28471Abstract Full Text PDF PubMed Google Scholar) and the cAMP-dependent protein kinase (PKA) (18Theurkauf W.E. Vallee R.B. J. Biol. Chem. 1982; 257: 3284-3290Abstract Full Text PDF PubMed Google Scholar), MAP-2 is considered one of the best substrates for CaM kinase II with the stoichiometry of phosphorylation reported to be from 5 to over 20 mol of phosphate/mol of MAP-2 (19Yamamoto H. Fukunaga K. Goto S. Tanaka E. Miyamoto E. J. Neurochem. 1985; 44: 759-768Crossref PubMed Scopus (157) Google Scholar). Based on the established involvement of the microtubule network in insulin secretion (20Somers G. Blondel B. Orci L. Malaisse W.J. Endocrinology. 1979; 104: 255-264Crossref PubMed Scopus (54) Google Scholar, 21Lacy P.E. Walker M.M. Fink C.J. Diabetes. 1972; 21: 987-998Crossref PubMed Scopus (284) Google Scholar, 22Lacy P.E. Finke E.H. Codilla R.C. Lab. Invest. 1975; 33: 570-576PubMed Google Scholar, 23Pipeleers D.G. Pipeleers-Marichal M.A. Kipnis D.M. Science. 1976; 191: 88-91Crossref PubMed Scopus (59) Google Scholar) and the suspected association of CaM kinase II with the cytoskeleton of the ॆ-cell (24Harrison D.E. Ashcroft S.J.H. Biochim. Biophys. Acta. 1982; 714: 313-319Crossref PubMed Scopus (44) Google Scholar), it was of interest to evaluate the potential of this enzyme to phosphorylate MAP-2 in these cells. Preliminary studies have established that CaM kinase II can be efficiently activated by Ca2+ in the permeabilized ॆ-cell. Therefore, to counter the inherent problem of a high level of basal MAP-2 phosphorylation, this model has been chosen to permit the study of phosphate incorporation from a high specific activity radionucleotide pool on a 舠silent舡 background. The correlation of MAP-2 phosphorylation to CaM kinase II activation and CaM kinase II activation to glucose-induced secretion, supports the hypothesis that a calcium-induced phosphorylation of MAP-2 by CaM kinase II may function as an important intermediate step in insulin secretion. ॆTC3 cells were obtained from Dr. Shimon Efrat (Albert Einstein College of Medicine, New York). RPMI 1640, glutamine, antibiotics, trypsin/EDTA, and fetal bovine serum were purchased from Life Technologies, Inc. Protein A-Sepharose, monoclonal anti-MAP-2 (clone HM-2), purified bovine brain MAP-2, and α-hemolysin (Staphylococcus aureus α-toxin) were purchased from Sigma. From Worthington, ribonuclease A and TPCK-treated trypsin were acquired. K252a was purchased from LC Laboratories (Woburn, MA); H-89 and KN-93 were obtained from Calbiochem. Forskolin was purchased from Research Biochemicals International (Natick, MA). [γ-32P]ATP was purchased from NEN Life Science Products. Autocamtide-2, sequence KKALRRQETVDAL (25Hanson P.I. Kapiloff M.S. Lou L.L. Rosenfeld M.G. Schulman H. Neuron. 1989; 3: 59-70Abstract Full Text PDF PubMed Scopus (237) Google Scholar), was synthesized by Bio-Synthesis, Inc. (Lewisville, TX). Anti-MAP-2 polyclonal antibody was raised against a heat-stable preparation of rat brain MAP-2 prepared by the method of Fellous et al.(26Fellous A. Francon J. Lennon A.M. Nunez J. Eur. J. Biochem. 1977; 78: 167-174Crossref PubMed Scopus (261) Google Scholar); the resulting antisera were purified to an IgG fraction enriched in anti-MAP-2 by chromatography on MAP-2-agarose. Mouse recombinant Ca2+/calmodulin protein kinase IIα was generously provided by Dr. Roger Colbran (Vanderbilt University Medical Center, Nashville, TN). cAMP-dependent protein kinase catalytic subunit from bovine heart was donated by Dr. Ben Harris (University of North Texas Health Science Ctr., Fort Worth, TX). All other chemicals were of the finest reagent grade available. ॆTC3 cells were cultured in RPMI 1640 medium supplemented with 2 mml-glutamine, 107 fetal bovine serum, 100 ॖg/ml penicillin, and 50 ॖg/ml streptomycin at 37 °C under an atmosphere of 57 CO2. In preparation for permeabilization, ॆTC3 cells were detached (Trypsin/EDTA) and equilibrated in suspension in culture medium for a minimum of 2 h. Following a brief centrifugation, the cells were washed twice with Ca2+-free Krebs-Ringer bicarbonate/Hepes buffer (25 mm Hepes, pH 7.4, 115 mm NaCl, 24 mm NaHCO3, 5 mm KCl, and 1 mm MgCl2) containing 1 mm EGTA, 6 mm glucose, and 0.17 bovine serum albumin. After counting, permeabilization was initiated by the addition of S. aureus toxin, α-hemolysin, to a concentration of 125–200 units/106 cell/0.1 ml Ca2+-free permeabilization buffer (20 mm Hepes, pH 7.0, 140 mm potassium glutamate, 5 mm NaCl, 4 mm MgSO4, 1 mm EGTA, and 300 ॖm Na2ATP). Permeabilization was conducted at 37 °C for 15 min with the efficiency monitored by visualizing trypan blue accessibility to >607 and then terminated by the addition of ice-cold Ca2+-free permeabilization buffer (washing twice). Cells were resuspended in permeabilization buffer containing 0.05 ॖm Ca2+ and placed on ice prior to experimental treatments. Free Ca2+ concentrations in incubation buffers were determined using a Ca2+ electrode (Orion) calibrated against known standards as described by Bers (27Bers D.M. Am. J. Physiol. 1982; 242: C404-C408Crossref PubMed Google Scholar). The permeabilization of ॆTC3 by α-toxin induces the formation in the plasma membrane of pores of defined diameter (∼2 nm) permitting ions and nucleotide access to the intracellular space without the loss of intracellular proteins (28Ahnert-Hilger G. Celis J.E. Cell Biology: A Laboratory Handbook. 3. Academic Press, San Diego1994: 77-82Google Scholar). Pancreatic islets were isolated from male Wistar rats (Harlan Sprague-Dawley, Indianapolis, IN) by collagenase P (Boehringer Mannheim) digestion and subsequent enrichment by centrifugation on a Ficoll gradient as described previously (2Wenham R.M. Landt M. Easom R.A. J. Biol. Chem. 1994; 269: 4947-4952Abstract Full Text PDF PubMed Google Scholar). Immunoblot analyses were performed on nitrocellulose membranes using a Western-LightTM protein detection kit (Tropix, Bedford, MA). Incubations with primary antibodies (rabbit polyclonal or monoclonal anti-MAP-2) were conducted overnight at 4 °C in blocking buffer. For the determination of CaM kinase II activation, 5 × 105 permeabilized cells were incubated in buffer (500 ॖl) containing varying concentrations of free Ca2+ for 1 min at 37 °C. CaM kinase II activity was assayed in sonicated homogenates using autocamtide-2 as substrate by a method described previously (29Babb E.L. Tarpley J. Landt M. Easom R.A. Biochem. J. 1996; 317: 167-172Crossref PubMed Scopus (16) Google Scholar).32Pi incorporation into autocamtide-2 was determined by Cerenkov radiation (Beckman). The activity of CaM kinase II in the absence of Ca2+/calmodulin (autonomous activity) expressed as percentage of total activity in the presence of Ca2+ was used as a measure of enzyme activation. Immunoprecipitation conditions were optimized for specific activity of [γ-32P]ATP, cell number, MAP-2 antibody/protein A ratio, and degree of permeabilization. Permeabilized ॆTC3 cells (approximately 2 × 106/condition) were preincubated at 37 °C for 5 min in 0.05 ॖmCa2+ permeabilization buffer, including kinase inhibitors when appropriate. The cells were then pelleted, resuspended in 200 ॖl of either 0.05 ॖm or 5–10 ॖmCa2+ permeabilization buffer with 300 ॖm[γ-32P]ATP (specific activity, 0.333 Ci/mmol) containing kinase inhibitors or activators when appropriate, and incubated at 37 °C for the indicated times. Phosphorylation was terminated by brief centrifugation (8,000 × g), washing with ice-cold phosphate-buffered saline (137 mmNaCl, 2.7 mm KCl, 4.3 mmNa2HPO4, 1.47 mmKH2PO4, pH 7.2) supplemented with phosphatase inhibitors (50 mm NaF, 10 mm sodium pyrophosphate), and finally resuspension of the cells in 300 ॖl of ice-cold RIPA buffer (0.01 m sodium phosphate, pH 7.2, 0.15m NaCl, 17 Nonidet P-40, 17 sodium deoxycholate, 0.17 SDS, 1 mm dithiothreitol) containing phosphatase and protease inhibitors (50 mm NaF, 10 mm sodium pyrophosphate, 1 mm phenylmethylsulfonyl fluoride, 50 ॖg/ml leupeptin). The cells were lysed in this RIPA buffer for 45 min at 4 °C on a rotating platform before clarification by centrifugation (12 min at 100,000 × g at 4 °C). The supernatant was transferred to a clean tube and incubated with polyclonal anti-rat MAP-2 antibody (1:100 dilution) for 2 h at 4 °C. Preswelled and washed protein A-Sepharose was added (25 ॖl), and the incubation continued for another 2 h at 4 °C on the rotating platform. The immune complexes bound to protein A-Sepharose were pelleted by centrifugation (3 min at 8,000 × g at 4 °C), and the pellets were washed twice with RIPA buffer. The immunoprecipitation pellets were resuspended in 35 ॖl of 2 × SDS sample buffer (124 mm Tris-HCl, pH 6.7, 6 mm SDS, 47 2-mercaptoethanol, 107 glycerol, 0.0077 bromphenol blue) and boiled for 10 min. Dissociated protein A-Sepharose was removed by centrifugation, and a portion (20 ॖl) of the supernatant was subjected to SDS-polyacrylamide electrophoresis on a 57 gel. Selected gels were silver stained to verify equality of protein loading. Dried gels were developed by autoradiography and32P-incorporation into MAP-2 quantified by densitometry using Optimas 4.0 and Scanalytics, ZERODscan 1.0, video imaging software. Purified MAP-2 (20 ॖg) was phosphorylated by the PKA catalytic subunit, or mouse recombinant CaM kinase IIα as described (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar) with the following exceptions; the PKA mixture was without exogenously added CaCl2, and the reaction volume of 50 ॖl contained [γ-32P]ATP (2 Ci/mmol) and 500 ng of kinase. Reactions proceeded for 18 min at 30 °C and were terminated by rapid chilling on ice. For phosphopeptide mapping, 32P-labeled MAP-2 was eluted from gel slices by incubation in 50 mmNH4HCO3, pH 7.3–7.6, initially supplemented with 17 ॆ-mercaptoethanol and 0.17 SDS for 18 h at 25 °C, and then without supplement for a further 3 h. The eluates were pooled, and the eluted MAP-2 was precipitated by the addition of a final concentration of 167 trichloroacetic acid (for 1 h on ice) in the presence of 20 ॖg heat-denatured RNase as carrier. In vitro phosphorylated MAP-2 was similarly precipitated at this step. The precipitate was resuspended in oxidizing solution (50 ॖl of performic acid) and then digested by the addition of 10 ॖg TPCK-treated trypsin for 18 h at 37 °C and then another 10 ॖg for a further 2.5 h. After repeated lyophilizing, the proteolytic digests were resuspended in electrophoresis buffer (2.57 formic acid and 7.87 glacial acetic acid, v/v) and spotted onto cellulose thin-layer plates. Two-dimensional separation of phosphopeptides by electrophoresis and chromatography was performed on a HTLE 7000 thin-layer electrophoresis apparatus (C. B. S. Scientific, La Jolla, CA) as described (30van der Geer P. Luo K. Sefton B.M. Hunter T. Celis J. Cell Biology: A Laboratory Handbook.3. Academic Press, San Diego1994: 422-448Google Scholar) except that the electrophoresis and chromatography steps were conducted at 1.3 kV for 25 min and for 14 h using a phosphochromatography buffer (37.57n-butanol, 257 pyridine, 7.57 glacial acetic acid, v/v), respectively. Data are expressed as the mean ± S.E. determined from at least three independent observations unless otherwise stated. Differences were assessed statistically through the employment of the most appropriate tests, either a two-way or one-way parametric ANOVA with Dunnett's multiple range test or with an independent t test (SAS Institute, Cary, NC). p < 0.05 indicates statistical significance. MAP-2 has been extensively characterized in mammalian brain where it is concentrated in dendritic processes (31Olmstedt J.B. Annu. Rev. Cell Biol. 1986; 2: 421-457Crossref PubMed Scopus (335) Google Scholar, 32De Camilli P. Miller P.E. Levitt P. Walter U. Greengard P. Neuroscience. 1984; 11: 761-817Crossref PubMed Scopus (144) Google Scholar, 33Caceres A. Payne M.R. Binder L.I. Steward O. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 1738-1742Crossref PubMed Scopus (159) Google Scholar) accounting for as much as 17 of the total cytoplasmic protein. In contrast, MAP-2 levels are much lower in non-neuronal tissues (34Valdivia M.M. Avila J. Coll J. Colaco C. Sandoval I.V. Biochem. Biophys. Res. Commun. 1982; 105: 1241-1249Crossref PubMed Scopus (26) Google Scholar) but demonstrated to be expressed in secretory cells, rat glioma (35Koszka C. Leichtfried F.E. Wiche G. Eur. J. Cell. Biol. 1985; 38: 149-156PubMed Google Scholar), pituitary and PC12 (34Valdivia M.M. Avila J. Coll J. Colaco C. Sandoval I.V. Biochem. Biophys. Res. Commun. 1982; 105: 1241-1249Crossref PubMed Scopus (26) Google Scholar). By immunoblot analysis using a polyclonal anti-MAP-2 antibody, ॆTC3 cells were demonstrated to express a high molecular weight protein (M r> 205 kDa) of electrophoretic mobility indistinguishable from MAP-2 purified from bovine brain (Fig.1 A, lane 1 versus lane 4). This MAP-2-like protein was immunoprecipitated from ॆTC3 cell homogenates by this antibody as indicated by its disappearance from ॆ-cell homogenates (Fig. 1 A, lane 2) and its appearance in protein A-sedimented immunoprecipitates (Fig.1 A, lane 3). Immunoblot analysis of this immunoprecipitate using a monoclonal anti-MAP-2 antibody confirmed the identity of this high molecular weight protein as MAP-2 (Fig.1 B, lane 2). That MAP-2 expression in ॆTC3 cells was not an artifact of ॆ-cell transformation was supported by the presence of immunoreactive immunoprecipitable MAP-2 in isolated rat islets (Fig. 1 B, lanes 3 and 4). It was noted however, that ॆTC3 cells express only a single form of MAP-2 in contrast to the characteristic doublet of MAP-2 (comprised of MAP-2A and -2B) observed in neurons (36Schulman H. J. Cell Biol. 1984; 99: 11-19Crossref PubMed Scopus (134) Google Scholar) and demonstrated here in islet immunoprecipitates (Fig. 1 B, lane 4). These findings, coupled to the ensuing demonstration that this high molecular weight protein was capable of being phosphorylated by kinases known to phosphorylate MAP-2 in vitro (see below), established that ॆTC3 cells express MAP-2. Despite previous inferences to the presence of MAPs in the pancreatic ॆ-cell (23Pipeleers D.G. Pipeleers-Marichal M.A. Kipnis D.M. Science. 1976; 191: 88-91Crossref PubMed Scopus (59) Google Scholar), this study is believed to be the first demonstration that these cells express MAP-2. Closer scrutiny of immunoblot analyses indicate that ॆ-cells express MAP-2 to a lower extent (by a factor of 50–60) relative to whole brain extract and therefore similar to estimates from other non-neuronal tissues (34Valdivia M.M. Avila J. Coll J. Colaco C. Sandoval I.V. Biochem. Biophys. Res. Commun. 1982; 105: 1241-1249Crossref PubMed Scopus (26) Google Scholar). The major objective of this study was to evaluate whether MAP-2 serves as a substrate for CaM kinase II in the pancreatic ॆ-cell. Sequence analysis of brain MAP-2 has identified 13 potential phosphorylation sites for CaM kinase II based on the published consensus sequence RXX(S/T) (37Kosik K.S. Orecchio L.D. Bakalis S. Duffy L. Neve R.L. J. Neurochem. 1988; 51: 587-598Crossref PubMed Scopus (40) Google Scholar). At least five of these sites have been demonstrated to be phosphorylated by this kinasein vitro (38Goldenring J.R. Vallano M.L. DeLorenzo R.J. J. Neurochem. 1985; 45: 900-905Crossref PubMed Scopus (63) Google Scholar) and a similar number of sites observed in stimulated GH3 cells in situ (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar). However, MAP-2 also serves as a prominent substrate for PKA (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar, 39Theurkauf W.E. Vallee R.B. J. Biol. Chem. 1983; 258: 7883-7886Abstract Full Text PDF PubMed Google Scholar) and other known protein kinases (40Tsuyama S. Bramblett G.T. Huang K.P. Flavin M. J. Biol. Chem. 1986; 261: 4110-4116Abstract Full Text PDF PubMed Google Scholar). Therefore, to circumvent anticipated difficulties in the detection of increased phosphate incorporation into MAP-2 as the result of the activation of selective protein kinases on a high background level of basal phosphorylation (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar), ॆTC3 cells were permeabilized with α-toxin, and radiolabeled [γ-32P]ATP only introduced during incubation periods. This method of permeabilization was chosen to minimize the loss of intracellular proteins (28Ahnert-Hilger G. Celis J.E. Cell Biology: A Laboratory Handbook. 3. Academic Press, San Diego1994: 77-82Google Scholar). In the presence of 0.05 ॖm Ca2+ (to mimic the intracellular concentration of a resting ॆ-cell (41Gilon P. Nenquin M. Henquin J.-C. Biochem. J. 1995; 311: 259-267Crossref PubMed Scopus (42) Google Scholar))32Pi was incorporated into MAP-2 in a time-dependent manner (Fig.2). This response likely reflected the activity of protein kinases involved in the maintenance of basal phosphorylation levels of MAP-2, which are thought to be required for the retention of protein function (14Brugg B. Matus A. J. Cell Biol. 1991; 114: 735-743Crossref PubMed Scopus (177) Google Scholar). On elevation of the Ca2+ concentration to 5 ॖm (to promote the activation of CaM kinase II) the extent of 32Piincorporation into MAP-2 was significantly increased; at the optimal time of 1 min, 5 ॖm Ca2+ increased32Pi incorporation into MAP-2 by 326 ± 767 relative to time 0 and by 1637 relative to control cells incubated in the presence of 0.05 ॖm Ca2+. An autoradiogram of immunoprecipitated MAP-2 under these experimental conditions is shown in Fig. 2 A. The phosphorylation of MAP-2 was also dependent on Ca2+concentration. Thus, Ca2+ concentrations of 0.5 ॖm or greater were required to induce detectable MAP-2 phosphorylation (Fig. 3 A), and half-maximal phosphorylation was achieved at approximately 0.8 ॖm Ca2+. As demonstrated in Fig.3 B, similar Ca2+ concentrations were required to activate CaM kinase II under identical conditions. Again increases in free Ca2+ concentration beyond 0.5 ॖm were required to induce kinase activation, and half-maximal activation was achieved at approximately 1 ॖm Ca2+, consistent with the known low affinity of this enzyme for Ca2+/calmodulin relative to other Ca2+-activated kinases (42Hanson P.I. Schulman H. Annu. Rev. Biochem. 1992; 61: 559-601Crossref PubMed Scopus (662) Google Scholar). The similarity of these Ca2+ dependence profiles is consistent with a functional association of Ca2+-dependent activation of CaM kinase II with the phosphorylation of ॆ-cell MAP-2 and is further substantiated by virtually identical Ca2+-dependence of CaM kinase-mediated phosphorylation of brain MAP-2 conducted in vitro (36Schulman H. J. Cell Biol. 1984; 99: 11-19Crossref PubMed Scopus (134) Google Scholar). The maintenance of a minimal level of cAMP is required to support glucose-induced insulin secretion from fluorescence-activated cell sorter-purified ॆ-cells (43Wang J.-L. Easom R.A. Hughes J.H. McDaniel M.L. Biochem. Biophys. Res. Commun. 1990; 171: 424-430Crossref PubMed Scopus (20) Google Scholar, 44Schuit F.C. Pipeleers D.G. Endocrinology. 1985; 117: 834-840Crossref PubMed Scopus (155) Google Scholar), and other studies have localized an effect of cAMP to potentiate Ca2+-induced insulin secretion to some distal step of the secretory process (45Hughes S.J. Christie M.R. Ashcroft S.J.H. Mol. Cell. Endocrinol. 1987; 50: 231-236Crossref PubMed Scopus (24) Google Scholar). Since MAP-2 may also serve as a substrate for PKA (36Schulman H. J. Cell Biol. 1984; 99: 11-19Crossref PubMed Scopus (134) Google Scholar) in the pancreatic ॆ-cell, it was important to determine to what extent Ca2+-induced phosphorylation of MAP-2 was contributed by the activation of this kinase. To this end, permeabilized cells were incubated in buffer containing 0.05 or 5 ॖm Ca2+ supplemented with forskolin (10 ॖm), a known activator of adenylate cyclase and/or H-89 (5 ॖm), a specific inhibitor of PKA (46Chijiwa T. Mishima A. Hagiwara M. Sano M. Hayashi K. Inoue T. Naito K. Toshioka T. Hidaka H. J. Biol. Chem. 1990; 265: 5267-5272Abstract Full Text PDF PubMed Google Scholar) (Fig. 4). In the presence of basal concentrations of Ca2+ (0.05 ॖm), forskolin induced a significant phosphorylation of MAP-2 (160 ± 137 relative to control, p = 0.004), which was totally abrogated by the inclusion of 5 ॖm H-89 (Fig. 4). As anticipated, forskolin had no significant effect on the activation state of CaM kinase II in these cell preparations (data not shown). In contrast, MAP-2 phosphorylation induced by stimulatory concentrations of Ca2+ (5 ॖm) was only modestly (227) reduced in the presence of H-89, an effect that was not statistically significant (p = 0.48) (Fig. 4). Accordingly, H-89 (5 ॖm) had only modest effects on CaM kinase II activity in ॆTC3 cell homogenates or on CaM kinase II-mediated phoshorylation of purified MAP-2 in vitro (∼157 inhibition in either case, data not shown). These observations demonstrate that the activation of PKA is capable of inducing MAP-2 phosphorylation in permeabilized ॆTC3 cells. This activation may contribute, although not significantly, to MAP-2 phosphorylation induced by 5 ॖmCa2+. A logical explanation is provided by the demonstrated presence in the ॆ-cell of calmodulin-dependent adenylate cyclase and phosphodiesterase activities that could mediate Ca2+-dependent modulations of intracellular cAMP concentrations (47Sharp G.W.G. Diabetologia. 1979; 16: 287-297Crossref PubMed Scopus (167) Google Scholar). Attempts to support the hypothesis that Ca2+-induced phosphorylation of MAP-2 was mediated by CaM kinase II via the use of putative inhibitors of this enzyme, KN-93 and K252a were thwarted by observed nonspecific effects of these compounds. Although KN-93 and K252a both abolished Ca2+-induced phosphorylation of MAP-2, these compounds also significantly suppressed forskolin-induced phosphorylation of MAP-2 (data not shown). In light of the inability of forskolin to affect the activation state of CaM kinase II, it was reasoned that these effects must reflect a lack of specificity of these compounds in situ. Therefore, in the absence of selective inhibitors of CaM kinase II, specific phosphorylation sites targeted in response to Ca2+ were determined by two-dimensional tryptic phosphopeptide analysis. Through in vitro incubation with recombinant enzyme, six major and several minor phosphorylation sites for CaM kinase II on purified brain MAP-2 were identified (Fig.5 A), which is consistent with previous reports (9Jefferson A.B. Schulman H. J. Biol. Chem. 1991; 266: 346-354Abstract Full Text PDF PubMed Google Scholar). Although initial studies were conducted using a neuronally expressed isoform of CaM kinase II, i.e. CaM kinase IIα, similar phosphopeptide patterns were generated from MAP-2 phosphorylated by a δ2 isoform recently demonstrated to be prominently expressed in ॆ-cells (48Mohlig M. Wolter S. Mayer P. Lang J. Osterhoff M. Horn P.A. Schatz H. Pfeiffer A. Endocrinology. 1997; 138: 2577-2584Crossref PubMed Scopus (37) Google Scholar). All of the major CaM kinase II sites were evident in digests made from MAP-2 that had been immunoprecipitated from ॆTC3 cells stimulated in the presence of 5 ॖm Ca2+ (Fig. 5 B, arrowheads) as verified by comigration with in vitro generated phosphopeptides (Fig. 5 C). Not only do these data suggest that structural features of neuronal MAP-2 surrounding these phosphorylation sites are equivalent in the pancreatic ॆ-cell protein but further imply that functional regulation of MAP-2 asserted by CaM kinase II-specific phosphorylation may also be conserved. Comparison of phosphopeptide digests generated from MAP-2 phosphorylated in the presence of basal (0.05 ॖm) or stimulatory (5 ॖm) Ca2+ concentrations revealed significant differences. A representative experiment is illustrated in Fig. 6. Although some variation was observed between experiments, characteristic of most analyses was a marked (780 ± 1407 over control) Ca2+-induced phosphorylation of a site central to the phosphopeptide map (Fig. 6, large open circle). Interestingly, this spot corresponded to the site most responsive toin vitro phosphorylation by purified CaM kinase II (Fig.5 A) providing compelling evidence that MAP-2 serves as a substrate for this enzyme in ॆTC3 cells. In the indicated experiment, Ca2+ induced the net phosphorylation of other sites (labeled by a small 舠o舡) that corresponded to CaM kinase II-specific sites (cf. Fig. 5 A), but significant differences in phosphate incorporation into these sites was not uniformly observed in all experiments. It is possible that these additional sites are not as readily available to the enzyme in situ relative to in vitro conditions, which suggests that they are secondary to the site described above. These data therefore demonstrate that CaM kinase II phosphorylates at least one site on MAP-2 establishing this protein as a substrate for this enzyme in the ॆ-cell. Ca2+ induced several changes in the phosphorylation of MAP-2 that cannot be ascribed to CaM kinase II. One such change was characterized by a net dephosphorylation (Fig. 6, cross symbol) implicating the action of a Ca2+-dependent phosphatase, e.g.calcineurin, as has been previously reported (49Goto S. Yamamoto H. Fukunaga K. Iwasa T. Matsukado Y. Miyamoto E. J. Neurochem. 1985; 45: 276-283Crossref PubMed Scopus (150) Google Scholar, 50Ferreira A. Kincaid R. Kosik K.S. Mol. Biol. Cell. 1993; 4: 1225-1238Crossref PubMed Scopus (89) Google Scholar). Ca2+also induced the phosphorylation of sites of similar migration to major sites targeted by PKA in vitro (Fig. 6, indicated by 舠p舡) that were clearly distinct from sites targeted by CaM kinase II (Fig. 7). This suggests that these may represent cAMP-induced phosphorylation events consistent with the ability of H-89 to modestly inhibit MAP-2 phosphorylation. Alternatively, they could represent sites phosphorylated by other Ca2+-sensitive protein kinases such as protein kinase C (51Easom R.A. Landt M. Colca J.R. Hughes J.H. Turk J. McDaniel M.L. J. Biol. Chem. 1990; 265: 14938-14946Abstract Full Text PDF PubMed Google Scholar) or MAP kinase (52Frodin M. Sekine N. Roche E. Filloux C. Prentki M. Wollheim C.B. Van Obberghen E. J. Biol. Chem. 1995; 270: 7882-7889Abstract Full Text Full Text PDF PubMed Scopus (199) Google Scholar). To what extent the function of MAP-2 is dependent on phosphorylation at multiple sites targeted by distinct kinases is not clear although it is likely that the site of phosphate incorporation rather than the overall amount is the critical factor for the specific regulation of MAP-2 (14Brugg B. Matus A. J. Cell Biol. 1991; 114: 735-743Crossref PubMed Scopus (177) Google Scholar). Nevertheless, because of its ability to act as a common substrate for both CaM kinase II and PKA, as well as other kinases/phosphatases, MAP-2 may provide a point of signal convergence for the integrated control of insulin secretion. A considerable body of evidence generated from the use of microtubule disrupting drugs support a role for the dynamic assembly/disassembly of microtubules in the mechanism of insulin secretion (20Somers G. Blondel B. Orci L. Malaisse W.J. Endocrinology. 1979; 104: 255-264Crossref PubMed Scopus (54) Google Scholar, 21Lacy P.E. Walker M.M. Fink C.J. Diabetes. 1972; 21: 987-998Crossref PubMed Scopus (284) Google Scholar, 22Lacy P.E. Finke E.H. Codilla R.C. Lab. Invest. 1975; 33: 570-576PubMed Google Scholar, 53Farshori P.Q. Goode D. J. Submicrosc. Cytol. Pathol. 1997; 26: 137-146Google Scholar). Dark-field microscopic studies have convincingly demonstrated that secretory granules derived from pancreatic ॆ-cells physically associate with stabilized microtubules through visible link structures, which were suggested to be MAPs, although not identified (54Suprenant K.A. Dentler W.L. J. Cell Biol. 1982; 93: 164-174Crossref PubMed Scopus (77) Google Scholar). The phosphorylation of MAP-2 by CaM kinase II and PKA leads, at leastin vitro, to the increased disassembly of microtubules (19Yamamoto H. Fukunaga K. Goto S. Tanaka E. Miyamoto E. J. Neurochem. 1985; 44: 759-768Crossref PubMed Scopus (157) Google Scholar) possibly through microtubule domain 舠stiffening舡 as shown for the low molecular weight MAP, tau (55Lichtenberg B. Mandelkow E.M. Hagestedt T. Mandelkow E. Nature. 1988; 334: 359-362Crossref PubMed Scopus (50) Google Scholar). The site-specific phosphorylation of MAP-2 by CaM kinase II could, therefore, regulate the association of secretory granules with microtubules in the ॆ-cell and/or regulate their translocation toward the exocytotic site as a result of changes in microtubule dynamics. Indeed such a role for Ca2+-dependent kinases in granule translocation has recently been obtained from video microscopy experiments in living ॆ-cells (56Hisatomi M. Hidaka H. Niki I. Endocrinology. 1996; 137: 4644-4649Crossref PubMed Scopus (62) Google Scholar) and is consistent with recent evidence that this enzyme acts at a site proximal to granule exocytosis (3Easom R.A. Filler N.R. Ings E.I. Tarpley J. Landt M. Endocrinology. 1997; 138: 2359-2364Crossref PubMed Scopus (30) Google Scholar). These pieces of evidence, combined with recent demonstrations that CaM kinase II is present in highly purified secretory granule membranes of ॆ-cell insulinoma tissue, 2R. A. Easom and C. J. Rhodes, unpublished observations. suggest that this enzyme may be perfectly poised to regulate insulin secretion via the regulation of microtubule function and its association with secretory granules. We thank Jill Meisenhelder and Tony Hunter of the Salk Institute, La Jolla, CA for technical assistance with the two-dimensional phosphopeptide maps.