- Email: [email protected]
Dynamic palmitoylation of neuromodulin (GAP-43) in cultured rat cerebellar neurons and mouse N1E-115 cells
Neuroscience Letters 234 (1997) 156–160
Dynamic palmitoylation of neuromodulin (GAP-43) in cultured rat cerebellar neurons and mouse N1E-115 cells Lauren P. Baker, Daniel R. Storm* Department of Pharmacology, University of Washington, Box 357280, Seattle, WA 98195-7280, USA Received 17 July 1997; received in revised form 19 August 1997; accepted 26 August 1997
Abstract We conducted pulse-chase and metabolic labeling experiments to determine directly whether palmitoylation of neuromodulin in neurons is dynamic, and if acylation is regulated. The rates of turnover of neuromodulin protein and associated palmitoyl groups were quantified using cultured cerebellar granule neurons and the neuronal cell line N1E-115. The half-life of [3H]palmitate bound to neuromodulin was approximately 5 h, whereas the half-life of the [35S]methionine-labeled neuromodulin was greater than 50 h. Metabolic and pulse-chase labeling experiments were carried out in the presence of various activators of cellular signaling pathways. Our data indicate that dynamic acylation and deacylation of neuromodulin in neurons are constitutive and are not regulated by G protein activation or other signals that control growth cone dynamics. 1997 Elsevier Science Ireland Ltd.
Keywords: Neuromodulin; GAP-43; Palmitoylation; Neuron; N1E-115; Growth cone
Neuromodulin is a neural-specific calmodulin-binding protein and PKC substrate that is highly expressed in growth cones and appears to play an important role in process outgrowth and path-finding [1,17,20]. Neuromodulin is palmitoylated in vivo on cysteine residues 3 and 4, and this is necessary and sufficient for anchoring of the protein to the plasma membrane [18]. Many palmitoylated proteins undergo dynamic acylation-deacylation cycles and in some cases, palmitoylation is tightly coupled to and regulated by activation of cognate signal transduction pathways [14]. With respect to neuromodulin, the kinetics for palmitoylation and deacylation in neurons have not been elucidated. One hypothesized function of neuromodulin is the amplification of signal transduction pathways activated by growth cone stimuli through Go or Gi [19,21,22]. In this study, we carried out experiments to determine whether neuromodulin acylation is dynamic and whether palmitoylation is regulated by activation of G proteins or by other signal transduction pathways.
* Corresponding author. Tel.: +1 206 5437028; fax: +1 206 6853822; e-mail: [email protected]
Cerebellar granule neurons were cultured from postnatal day 6–7 rat pups according to Lasher and Zagan [12] and Trenkner [24], in neurobasal medium plus 1 × B-27 (Gibco BRL), 1% FBS, 0.5 mM glutamine, 1% penicillin/streptomycin, 30 mM KCl and 5 mM cytosine b-d-arabinofuranoside. Cultures were maintained in a 37°C incubator at 5% CO2 and were used for experiments after 1–15 days in culture. The N1E-115 neuronal cell line was maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS and 1% penicillin/streptomycin. Cells were differentiated by serum starvation and used for experiments after 3–4 days. For metabolic labeling experiments with [3H]palmitic acid, cerebellar granule cells and differentiating N1E-115 cells were labeled with 0.5 mCi/ml [9,10-3H]palmitic acid (DuPont NEN; 56 Ci/mmol) for 3 h in medium containing 1% FBS and 5 mM sodium pyruvate. For pulse-chase experiments, cultures were chased with 100 mM unlabeled palmitic acid. For labeling experiments with [35S]methionine, cells were starved for 30 min with cysteine/methionine-free DMEM containing 1% dialyzed FBS, and labeled with 0.2–0.5 mCi/ml EXPRE35S35S [35S]protein labeling mix (DuPont NEN; 1175 Ci/mmol) for 3 h. For pulse-
0304-3940/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0304- 3940(97) 00667- 8
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
chase experiments, cultures were chased in complete medium containing 1% FBS. Cultures were harvested with 0.1 ml 1% SDS in 20 mM HEPES, pH 7.5, and 0.9 ml modified RIPA buffer (0.1% SDS, 1% NP-40, 1 mM EDTA, 150 mM NaCl, 20 mM Na2HPO4, pH 7.4) at 4°C. Samples were sheared with a 23-G needle and protein concentrations were determined using a BCA protein assay kit (Pierce). Identical protein amounts of whole-cell lysate were immunoprecipitated at 4°C overnight using 0.5–1.0 mg monoclonal anti-GAP-43 antibody 91E12 (Boehringer Mannheim) and 20 ml protein G PLUS-agarose (Oncogene Science). The samples were heated at 95°C for 1 min in SDS sample buffer (without DTT), and subjected to PAGE analysis according to Laemmli [11]. Gels were fixed with 10% methanol and 5% acetic acid and for treatment with neutral hydroxylamine, were incubated in 1 M NH2OH (pH 7.0) or 1 M Tris (pH 7.0) for 12–16 h at room temperature. Gels were then treated with amplify (Amersham), dried, and subjected to fluorography for 1–8 weeks at −80°C, and the data was analyzed by densitometry. Following fluorography, corresponding neuromodulin bands were cut out, solubilized in solvable (DuPont NEN) and radioactivity was quantified by scintillation counting. The goals of this study were to determine whether neuromodulin acylation occurs as part of a dynamic acylationdeacylation cycle in neurons, and whether such a cycle is regulated by signal transduction pathways activated in the growth cone. We used cerebellar granule cells as a source of primary neurons that expresses high levels of neuromodulin, and differentiating neuronal N1E-115 cells that express high levels of neuromodulin as well as receptors for factors that induce growth cone collapse, including thrombin and LPA [8,9,23]. In immunoprecipitates (Fig. 1A, lanes 3 and 6), and by direct lysate loading (lanes 2 and 5), a single [3H]palmitic acid-labeled protein band migrated at approximately 50 kDa on 10% polyacrylamide gels, and co-migrated with neuromodulin identified by Western blot (Fig. 1A, lanes 1 and 4) using an antibody generated against the ‘IQ’ domain of neuromodulin and neurogranin [4]. The [3H]palmitic acid incorporated into neuromodulin in N1E-115 cells (Fig. 1B) and cerebellar neuron cultures (data not shown) was sensitive to treatment with 1 M neutral hydroxylamine (pH 7.0), indicating that palmitoylation of neuromodulin occurs via thioester linkage. In addition, treatment of immunoprecipitates with 50 mM DTT significantly reduced the amount of [3H]palmitate associated with neuromodulin on gels (Fig. 1B). If a significant turnover of the palmitoyl groups occurs at the plasma membrane, these cells should incorporate [3H]palmitic acid into neuromodulin in the absence of protein synthesis. Preincubation of differentiating N1E-115 cells with 50 mg/ml cycloheximide for 60 min before labeling with [35S]methionine for 60 min completely inhibited synthesis of neuromodulin (Fig. 1C), while the cells still incorporated [3H]palmitic acid (Fig. 1D), indicating that
157
previously synthesized neuromodulin undergoes a dynamic acylation-deacylation cycle. To compare directly the rates of turnover for the neuromodulin protein and associated palmitate, pulse-chase experiments were carried out using cerebellar granule cells. The half-life for turnover of palmitate attached to neuromodulin was approximately 5 h (Fig. 2). In contrast, the turnover of the protein itself was surprisingly slow, with a half-life greater than 48 h (Fig. 2). These turnover rates were consistently observed in neurons ranging from 1–15 days in culture. Neuronal growth cones are sensitive to several types of extracellular stimuli. Therefore, we examined the rates of deacylation and acylation of neuromodulin in cerebellar neurons in the presence of agents known to affect growth cone dynamics or specific signal transduction systems. In
Fig. 1. (A) Immunoprecipitation of [3H]palmitate labeled neuromodulin. Cerebellar granule cells (lanes 1–3). NIE-115 cells (lanes 4–6). Neuromodulin is recognized by anti ‘IQ’ antibody in Western blot (lanes 1 and 4). Note the presence of another IQ-domain containing protein, p68, in the N1E-115 lysate (lane 4). Direct loading of cell lysate to gels (lanes 2 and 5) shows that neuromodulin is the major [3H]labeled protein. Immunoprecipitation of neuromodulin (lanes 3 and 6) shows a single [3H]palmitate labeled band. (B) [3H]Palmitate labeling of neuromodulin is sensitive to neutral hydroxylamine and to the presence of DTT. Differentiating N1E-115 cells were labeled with 0.5 mCi/ml [3H]palmitic acid for 3 h, cells were lysed, neuromodulin was immunoprecipitated, and samples were run on 10% polyacrylamide gels. Gels were incubated in the presence of 1 M Tris (pH 7.0) or 1 M NH2OH (pH 7.0) for 12–16 h. Separate immunoprecipitates were treated with SDS sample buffer in the absence (−) or presence (+) of 50 mM DTT. (C) N1E-115 cells incorporate [3H]palmitic acid into neuromodulin in the absence of protein synthesis. Differentiating N1E-115 cells were pretreated with the indicated concentrations of cycloheximide for 1 h before labeling with 0.5 mCi/ml of [35S]methionine for 1 h. Protein synthesis was completely inhibited by 50 mg/ml pretreatment with cycloheximide for 1 h. (D) N1E-115 cells pretreated with 50 mg/ml cycloheximide for 1 h still incorporated [3H]palmitic acid into neuromodulin.
158
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
compound that inhibits palmitoylation of neuromodulin as well as other proteins [15]. Under the conditions used (absence of serum and pretreatment with 50 mg/ml cycloheximide), neuromodulin synthesis and early/posttranslational palmitoylation is inhibited without affecting dynamic palmitoylation at the plasma membrane [10,16]. Tunicamycin (50 mg/ml) inhibited the rate of incorporation of [3H]palmitic acid into neuromodulin by approximately 50% (Fig. 4A). Carbachol, isoproterenol, forskolin, PDBU, and KCl did not affect the rates of palmitoylation. In addition, mastoparan, a peptide that binds to and activates Go and Gi [7] had no effect (Fig. 4B–D). Analogous experiments were carried out using the N1E-115 neuronal cell line that expresses receptors for LPA and thrombin. Incorporation of [3H]palmitic acid into neuromodulin was not affected by mastoparan, thrombin, LPA, carbachol, forskolin, PDBU, or 60 mM KCl (Fig. 5A,B). The objectives of this study were to determine if neuro-
Fig. 2. Differential turnover of [3H]palmitate versus [35S]methioninelabeled neuromodulin protein in cerebellar granule cells. Cells were grown in culture for 15 days and then pulse-chased with 0.5 mCi/ml [3H]palmitic acid or 0.2 mCi/ml EXPRE35S35S (see text). (A) Controls for immunoprecipitation (c) were carried out in the absence of antiGAP-43 antibody. The percent label remaining was analyzed over time using densitometry (B). The half-life of [3H]palmitic acid labeledneuromodulin was approximately 5 h, whereas the half-life of [35S]methionine-labeled neuromodulin is greater than the 48 h course of this experiment. These results are indicative of seven independent experiments.
pulse-chase experiments, the rates of deacylation varied somewhat during these short time courses, due to the rapid equilibration and subsequent release of label from lipid stores. However, neither carbachol nor isoproterenol reproducibly affected the rate of depalmitoylation of neuromodulin. The rate of deacylation was also unaffected by forskolin (Fig. 3A). Neuromodulin interacts with calmodulin in vitro [3] and in vivo [5,6] and is also phosphorylated by PKC [2]. Therefore, we measured the effect of PKC activation and intracellular calcium increase on the rate of deacylation of neuromodulin. TPA, 60 mM KCl, or ionomycin had no effect on the rate of neuromodulin deacylation (Fig. 3B). Similar results were obtained using cortical cultures treated with carbachol, isoproterenol, forskolin, or TPA (data not shown). Metabolic labeling of cultured cerebellar granule cells with [3H]palmitic acid resulted in incorporation of label that reached a maximum by 120 min (Fig. 4A–D, controls), regardless of whether cells had been pretreated with cycloheximide, providing additional evidence for active deacylation of neuromodulin during the labeling period. To verify that this assay was sensitive enough to detect changes in palmitoylation rates, cells were treated with tunicamycin, a
Fig. 3. Pulse-chase labeling of cerebellar granule cells with [3H]palmitic acid. (A,B) Cultures were pulse-chased with 0.5 mCi/ml [3H]palmitic acid (see text) in the absence or presence of the indicated agonists for 0–120 min. The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: (A) control (7); carbachol, 1 mM (2); isoproterenol, 100 mM (3); forskolin, 100 mM (4). (B) TPA, 200 nM (4); KCl, 60 mM (1); ionomycin, 10 mM (2). The data were analyzed by scintillation counting and are presented as average ± SD.
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
modulin undergoes dynamic palmitoylation in neurons, and if so, to determine if palmitoylation is regulated. This is the first study to report the quantification of the half-life of the neuromodulin-associated palmitoyl groups in neurons, as well as the first to directly demonstrate a dynamic cycle of palmitoylation and depalmitoylation of neuromodulin in intact neurons. The rate of turnover of palmitate associated with neuromodulin is at least 10 times faster than the rate of turnover of the protein itself. In addition, neuromodulin incorporated palmitic acid in the absence of protein synthesis, and labeling saturated within 120 min, indicating active deacylation during the labeling period. The half-life of [3H]palmitic acid on neuromodulin of approximately 5 h is relatively long, compared with other palmitoylated proteins that undergo dynamic and/or regulated acylation, suggesting that the level of neuromodulin in neurons is relatively stable. Experiments with activators of various cellular signaling pathways were carried out under conditions that measured the rate of neuromodulin depalmitoylation (pulse-chase) and the rate of palmitoylation + depalmitoylation (metabolic labeling). These experiments indicate that neuromodulin palmitoylation is not regulated by agents that activate Go or Gi or adenylyl cyclase. In addition, agonists that stimulate growth cone collapse were without effect, as were activators of PKC and increased intracellular calcium. The
159
lack of effect of a muscarinic agonist and mastoparan on palmitoylation is particularly interesting since it has been hypothesized that neuromodulin signals through Go or Gi [21] and that the G protein and palmitoylation domains overlap [13,21,25]. The physiological relevance of the interaction between non-palmitoylated neuromodulin and G proteins is unclear and remains to be demonstrated in vivo. In summary, this is the first study to report the quantification of the half-life of the neuromodulin-associated palmitoyl groups in neurons, and to show that there is a dynamic cycle of palmitoylation and depalmitoylation of neuromodulin in intact neurons. Palmitoylation of neuromodulin in cultured rat cerebellar granule cells is not regulated by increases in intracellular calcium, activation of Go or Gi proteins, PKC, or, in N1E-115 cells, by agents that induce growth cone collapse. These data suggest that the neuromodulin acylation-deacylation cycle occurs constitutively and that a constant ratio of cytosolic to membrane-associated neuromodulin is required for the normal function of neuromodulin in neurons. We thank Drs. Thomas Hinds, Yuechueng Liu, and JeanChristophe Deloulme for valuable discussions and reading of the manuscript, and members of the Storm lab for critical reading of the manuscript. This work was supported by National Institutes of Health NRSA postdoctoral fellowship
Fig. 4. Metabolic labeling of cerebellar granule cells with [3H]palmitic acid. Cerebellar cells were labeled with 0.5 mCi/ml [3H]palmitic acid in the absence or presence of the inhibitor, tunicamycin, for 0–120 min, or agonists for 0–180 min. The data were analyzed by scintillation counting, and single representative experiments are shown. (A) Cells were labeled after pretreatment with 50 mg/ml cycloheximide for 1 h in the absence of FBS, and in the absence or presence of 50 mg/ml tunicamycin. Tunicamycin (50 mg/ml) inhibited incorporation of [3H]palmitic acid by approximately 50% after 120 min. These results are representative of three independent experiments. (B–D) The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: control (6); carbachol, 1 mM (4); mastoparan, 10 mM (5); isoproterenol, 100 mM (2); forskolin, 100 mM (3); PDBU, 200 nM (2); KCl, 60 mM (2).
160
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Fig. 5. Metabolic labeling of N1E-115 cells with [3H]palmitic acid. (A,B) Differentiating N1E-115 cells were labeled with 0.5 mCi/ml [3H]palmitic acid in the absence or presence of the indicated agonists for 0–120 min. The data were analyzed by scintillation counting, and single representative experiments are shown. The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: (A) control (5); mastoparan, 10 mM (2); thrombin, 50 ng/ml (2); LPA, 1 mM (2). (B) carbachol, 1 mM (2); forskolin, 100 mM (2); PDBU, 200 nM (2); KCl, 60 mM (2).
# NS09528 (L.P.B.) and by National Institutes of Health grants NS20498 and NS31496. [1] Aigner, L., Arber, S., Kapfhammer, J.P., Laux, T., Schneider, C., Botteri, F., Brenner, H.R. and Caroni, P., Overexpression of the neural growth-associated protein GAP-43 induces sprouting in the adult nervous system of transgenic mice, Cell, 83 (1995) 269–278. [2] Alexander, K.A., Cimler, B.M., Meier, K.E. and Storm, D.R., Regulation of calmodulin binding to P-57. A neurospecific calmodulin binding protein, J. Biol. Chem., 262 (1987) 6108– 6113. [3] Andreasen, T.J., Luetje, C.W., Heideman, W. and Storm, D.R., Purification of a novel calmodulin binding protein from bovine cerebral cortex membranes, Biochemistry, 22 (1983) 4615– 4618. [4] Buelt, M.K., Glidden, B.J. and Storm, D.R., Regulation of p68 RNA helicase by calmodulin and protein kinase C, J. Biol. Chem., 269 (1994) 29367–29370. [5] Chao, S., Benowitz, L.I., Krainc, D. and Irwin, N., Use of a twohybrid system to investigate molecular interactions of GAP-43, Mol. Brain Res., 40 (1996) 195–202. [6] Gamby, C., Waage, M.C., Allen, R.G. and Baizer, L., Analysis of the role of calmodulin binding and sequestration in neuromo-
[15]
[16]
[17] [18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
dulin (GAP-43) function, J. Biol. Chem., 271 (1996) 26698– 26705. Higashijima, T., Uzu, S., Nakajima, T. and Ross, E.M., Mastoparan, a peptide toxin from wasp venom, mimics receptors by activating GTP-binding regulatory proteins (G proteins), J. Biol. Chem., 263 (1988) 6491–6494. Jalink, K., Eichholtz, T., Postma, F.R., van Corven, E.J. and Moolenaar, W.H., Lysophosphatidic acid induces neuronal shape changes via a novel, receptor-mediated signaling pathway: similarity to thrombin action, Cell Growth Diff., 4 (1993) 247–255. Jalink, K. and Moolenaar, W.H., Thrombin receptor activation causes rapid neural cell rounding and neurite retraction independent of classic second messengers, J. Cell Biol., 118 (1992) 411–419. James, G. and Olson, E., Experimental approaches to the study of reversible protein fatty acylation in mammalian cells, Methods, 1 (1990) 270–275. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 (1970) 680–685. Lasher, R.S. and Zagan, I.S., The effect of potassium on neuronal differentiation in cultures of dissociated newborn rat cerebellum, Brain Res., 41 (1972) 482–488. Liu, Y.-C., Chapman, E.R. and Storm, D.R., Targeting of neuromodulin (GAP-43) fusion proteins to growth cones in cultured rat embryonic neurons., Neuron, 6 (1991) 411–420. Milligan, G., Parenti, M. and Magee, A.I., The dynamic role of palmitoylation in signal transduction, Trends Biochem. Sci., 20 (1995) 181–186. Patterson, S.I. and Skene, J.H.P., Novel inhibitory action of tunicamycin homologues suggests a role of dynamic protein fatty acylation in growth cone-mediated neurite extension, J. Cell Biol., 124 (1994) 521–536. Patterson, S.I. and Skene, J.H.P., Inhibition of dynamic protein palmitoylation in intact cells with tunicamycin, Methods Enzymol., 250 (1995) 284–300. Skene, J.H.P., Axonal growth-associated proteins, Annu. Rev. Neurosci., 12 (1989) 127–156. Skene, J.H.P. and Virag, I., Posttranslational membrane attachment and dynamic fatty acylation of a neuronal growth cone protein, J. Cell. Biol., 108 (1989) 613–624. Strittmatter, S.M., Cannon, S.C., Ross, E.M., Higashijima, T. and Fishman, M.C., GAP-43 augments G-protein-coupled receptor transduction in Xenopus laevis oocytes, Proc. Natl. Acad. Sci. USA, 90 (1993) 5327–5331. Strittmatter, S.M., Fankhauser, C., Huang, P.L., Mashimo, H. and Fishman, M.C., Neuronal pathfinding is abnormal in mice lacking the neuronal growth cone protein GAP-43, Cell, 80 (1995) 445–452. Strittmatter, S.M., Valenzuela, D., Kennedy, T.E., Neer, E.J. and Fishman, M.C., Go is a major growth cone protein subject to regulation by GAP-43, Nature, 344 (1990) 836–841. Strittmatter, S.M., Valenzuela, D., Sudo, Y., Linder, M.E. and Fishman, M.C., An intracellular guanine nucleotide release protein for Go; GAP-43 stimulates isolated a subunits by a novel mechanism, J. Biol. Chem., 266 (1991) 22465–22471. Suidan, H.S., Stone, S.R., Hemmings, B.A. and Monard, D., Thrombin causes neurite retraction in neuronal cells through activation of cell surface receptors, Neuron, 8 (1992) 363–375. Trenkner, E., Cerebellar cells in culture. In G. Banker and K. Goslin (Eds.), Culturing Nerve Cells, MIT Press, Cambridge, MA, 1991, pp. 283–307. Zuber, M.X., Strittmatter, S.M. and Fishman, M.C., A membrane-targeting signal in the amino terminus of the neuronal protein GAP-43, Nature, 341 (1989) 345–348.
Dynamic palmitoylation of neuromodulin (GAP-43) in cultured rat cerebellar neurons and mouse N1E-115 cells Lauren P. Baker, Daniel R. Storm* Department of Pharmacology, University of Washington, Box 357280, Seattle, WA 98195-7280, USA Received 17 July 1997; received in revised form 19 August 1997; accepted 26 August 1997
Abstract We conducted pulse-chase and metabolic labeling experiments to determine directly whether palmitoylation of neuromodulin in neurons is dynamic, and if acylation is regulated. The rates of turnover of neuromodulin protein and associated palmitoyl groups were quantified using cultured cerebellar granule neurons and the neuronal cell line N1E-115. The half-life of [3H]palmitate bound to neuromodulin was approximately 5 h, whereas the half-life of the [35S]methionine-labeled neuromodulin was greater than 50 h. Metabolic and pulse-chase labeling experiments were carried out in the presence of various activators of cellular signaling pathways. Our data indicate that dynamic acylation and deacylation of neuromodulin in neurons are constitutive and are not regulated by G protein activation or other signals that control growth cone dynamics. 1997 Elsevier Science Ireland Ltd.
Keywords: Neuromodulin; GAP-43; Palmitoylation; Neuron; N1E-115; Growth cone
Neuromodulin is a neural-specific calmodulin-binding protein and PKC substrate that is highly expressed in growth cones and appears to play an important role in process outgrowth and path-finding [1,17,20]. Neuromodulin is palmitoylated in vivo on cysteine residues 3 and 4, and this is necessary and sufficient for anchoring of the protein to the plasma membrane [18]. Many palmitoylated proteins undergo dynamic acylation-deacylation cycles and in some cases, palmitoylation is tightly coupled to and regulated by activation of cognate signal transduction pathways [14]. With respect to neuromodulin, the kinetics for palmitoylation and deacylation in neurons have not been elucidated. One hypothesized function of neuromodulin is the amplification of signal transduction pathways activated by growth cone stimuli through Go or Gi [19,21,22]. In this study, we carried out experiments to determine whether neuromodulin acylation is dynamic and whether palmitoylation is regulated by activation of G proteins or by other signal transduction pathways.
* Corresponding author. Tel.: +1 206 5437028; fax: +1 206 6853822; e-mail: [email protected]
Cerebellar granule neurons were cultured from postnatal day 6–7 rat pups according to Lasher and Zagan [12] and Trenkner [24], in neurobasal medium plus 1 × B-27 (Gibco BRL), 1% FBS, 0.5 mM glutamine, 1% penicillin/streptomycin, 30 mM KCl and 5 mM cytosine b-d-arabinofuranoside. Cultures were maintained in a 37°C incubator at 5% CO2 and were used for experiments after 1–15 days in culture. The N1E-115 neuronal cell line was maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS and 1% penicillin/streptomycin. Cells were differentiated by serum starvation and used for experiments after 3–4 days. For metabolic labeling experiments with [3H]palmitic acid, cerebellar granule cells and differentiating N1E-115 cells were labeled with 0.5 mCi/ml [9,10-3H]palmitic acid (DuPont NEN; 56 Ci/mmol) for 3 h in medium containing 1% FBS and 5 mM sodium pyruvate. For pulse-chase experiments, cultures were chased with 100 mM unlabeled palmitic acid. For labeling experiments with [35S]methionine, cells were starved for 30 min with cysteine/methionine-free DMEM containing 1% dialyzed FBS, and labeled with 0.2–0.5 mCi/ml EXPRE35S35S [35S]protein labeling mix (DuPont NEN; 1175 Ci/mmol) for 3 h. For pulse-
0304-3940/97/$17.00 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0304- 3940(97) 00667- 8
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
chase experiments, cultures were chased in complete medium containing 1% FBS. Cultures were harvested with 0.1 ml 1% SDS in 20 mM HEPES, pH 7.5, and 0.9 ml modified RIPA buffer (0.1% SDS, 1% NP-40, 1 mM EDTA, 150 mM NaCl, 20 mM Na2HPO4, pH 7.4) at 4°C. Samples were sheared with a 23-G needle and protein concentrations were determined using a BCA protein assay kit (Pierce). Identical protein amounts of whole-cell lysate were immunoprecipitated at 4°C overnight using 0.5–1.0 mg monoclonal anti-GAP-43 antibody 91E12 (Boehringer Mannheim) and 20 ml protein G PLUS-agarose (Oncogene Science). The samples were heated at 95°C for 1 min in SDS sample buffer (without DTT), and subjected to PAGE analysis according to Laemmli [11]. Gels were fixed with 10% methanol and 5% acetic acid and for treatment with neutral hydroxylamine, were incubated in 1 M NH2OH (pH 7.0) or 1 M Tris (pH 7.0) for 12–16 h at room temperature. Gels were then treated with amplify (Amersham), dried, and subjected to fluorography for 1–8 weeks at −80°C, and the data was analyzed by densitometry. Following fluorography, corresponding neuromodulin bands were cut out, solubilized in solvable (DuPont NEN) and radioactivity was quantified by scintillation counting. The goals of this study were to determine whether neuromodulin acylation occurs as part of a dynamic acylationdeacylation cycle in neurons, and whether such a cycle is regulated by signal transduction pathways activated in the growth cone. We used cerebellar granule cells as a source of primary neurons that expresses high levels of neuromodulin, and differentiating neuronal N1E-115 cells that express high levels of neuromodulin as well as receptors for factors that induce growth cone collapse, including thrombin and LPA [8,9,23]. In immunoprecipitates (Fig. 1A, lanes 3 and 6), and by direct lysate loading (lanes 2 and 5), a single [3H]palmitic acid-labeled protein band migrated at approximately 50 kDa on 10% polyacrylamide gels, and co-migrated with neuromodulin identified by Western blot (Fig. 1A, lanes 1 and 4) using an antibody generated against the ‘IQ’ domain of neuromodulin and neurogranin [4]. The [3H]palmitic acid incorporated into neuromodulin in N1E-115 cells (Fig. 1B) and cerebellar neuron cultures (data not shown) was sensitive to treatment with 1 M neutral hydroxylamine (pH 7.0), indicating that palmitoylation of neuromodulin occurs via thioester linkage. In addition, treatment of immunoprecipitates with 50 mM DTT significantly reduced the amount of [3H]palmitate associated with neuromodulin on gels (Fig. 1B). If a significant turnover of the palmitoyl groups occurs at the plasma membrane, these cells should incorporate [3H]palmitic acid into neuromodulin in the absence of protein synthesis. Preincubation of differentiating N1E-115 cells with 50 mg/ml cycloheximide for 60 min before labeling with [35S]methionine for 60 min completely inhibited synthesis of neuromodulin (Fig. 1C), while the cells still incorporated [3H]palmitic acid (Fig. 1D), indicating that
157
previously synthesized neuromodulin undergoes a dynamic acylation-deacylation cycle. To compare directly the rates of turnover for the neuromodulin protein and associated palmitate, pulse-chase experiments were carried out using cerebellar granule cells. The half-life for turnover of palmitate attached to neuromodulin was approximately 5 h (Fig. 2). In contrast, the turnover of the protein itself was surprisingly slow, with a half-life greater than 48 h (Fig. 2). These turnover rates were consistently observed in neurons ranging from 1–15 days in culture. Neuronal growth cones are sensitive to several types of extracellular stimuli. Therefore, we examined the rates of deacylation and acylation of neuromodulin in cerebellar neurons in the presence of agents known to affect growth cone dynamics or specific signal transduction systems. In
Fig. 1. (A) Immunoprecipitation of [3H]palmitate labeled neuromodulin. Cerebellar granule cells (lanes 1–3). NIE-115 cells (lanes 4–6). Neuromodulin is recognized by anti ‘IQ’ antibody in Western blot (lanes 1 and 4). Note the presence of another IQ-domain containing protein, p68, in the N1E-115 lysate (lane 4). Direct loading of cell lysate to gels (lanes 2 and 5) shows that neuromodulin is the major [3H]labeled protein. Immunoprecipitation of neuromodulin (lanes 3 and 6) shows a single [3H]palmitate labeled band. (B) [3H]Palmitate labeling of neuromodulin is sensitive to neutral hydroxylamine and to the presence of DTT. Differentiating N1E-115 cells were labeled with 0.5 mCi/ml [3H]palmitic acid for 3 h, cells were lysed, neuromodulin was immunoprecipitated, and samples were run on 10% polyacrylamide gels. Gels were incubated in the presence of 1 M Tris (pH 7.0) or 1 M NH2OH (pH 7.0) for 12–16 h. Separate immunoprecipitates were treated with SDS sample buffer in the absence (−) or presence (+) of 50 mM DTT. (C) N1E-115 cells incorporate [3H]palmitic acid into neuromodulin in the absence of protein synthesis. Differentiating N1E-115 cells were pretreated with the indicated concentrations of cycloheximide for 1 h before labeling with 0.5 mCi/ml of [35S]methionine for 1 h. Protein synthesis was completely inhibited by 50 mg/ml pretreatment with cycloheximide for 1 h. (D) N1E-115 cells pretreated with 50 mg/ml cycloheximide for 1 h still incorporated [3H]palmitic acid into neuromodulin.
158
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
compound that inhibits palmitoylation of neuromodulin as well as other proteins [15]. Under the conditions used (absence of serum and pretreatment with 50 mg/ml cycloheximide), neuromodulin synthesis and early/posttranslational palmitoylation is inhibited without affecting dynamic palmitoylation at the plasma membrane [10,16]. Tunicamycin (50 mg/ml) inhibited the rate of incorporation of [3H]palmitic acid into neuromodulin by approximately 50% (Fig. 4A). Carbachol, isoproterenol, forskolin, PDBU, and KCl did not affect the rates of palmitoylation. In addition, mastoparan, a peptide that binds to and activates Go and Gi [7] had no effect (Fig. 4B–D). Analogous experiments were carried out using the N1E-115 neuronal cell line that expresses receptors for LPA and thrombin. Incorporation of [3H]palmitic acid into neuromodulin was not affected by mastoparan, thrombin, LPA, carbachol, forskolin, PDBU, or 60 mM KCl (Fig. 5A,B). The objectives of this study were to determine if neuro-
Fig. 2. Differential turnover of [3H]palmitate versus [35S]methioninelabeled neuromodulin protein in cerebellar granule cells. Cells were grown in culture for 15 days and then pulse-chased with 0.5 mCi/ml [3H]palmitic acid or 0.2 mCi/ml EXPRE35S35S (see text). (A) Controls for immunoprecipitation (c) were carried out in the absence of antiGAP-43 antibody. The percent label remaining was analyzed over time using densitometry (B). The half-life of [3H]palmitic acid labeledneuromodulin was approximately 5 h, whereas the half-life of [35S]methionine-labeled neuromodulin is greater than the 48 h course of this experiment. These results are indicative of seven independent experiments.
pulse-chase experiments, the rates of deacylation varied somewhat during these short time courses, due to the rapid equilibration and subsequent release of label from lipid stores. However, neither carbachol nor isoproterenol reproducibly affected the rate of depalmitoylation of neuromodulin. The rate of deacylation was also unaffected by forskolin (Fig. 3A). Neuromodulin interacts with calmodulin in vitro [3] and in vivo [5,6] and is also phosphorylated by PKC [2]. Therefore, we measured the effect of PKC activation and intracellular calcium increase on the rate of deacylation of neuromodulin. TPA, 60 mM KCl, or ionomycin had no effect on the rate of neuromodulin deacylation (Fig. 3B). Similar results were obtained using cortical cultures treated with carbachol, isoproterenol, forskolin, or TPA (data not shown). Metabolic labeling of cultured cerebellar granule cells with [3H]palmitic acid resulted in incorporation of label that reached a maximum by 120 min (Fig. 4A–D, controls), regardless of whether cells had been pretreated with cycloheximide, providing additional evidence for active deacylation of neuromodulin during the labeling period. To verify that this assay was sensitive enough to detect changes in palmitoylation rates, cells were treated with tunicamycin, a
Fig. 3. Pulse-chase labeling of cerebellar granule cells with [3H]palmitic acid. (A,B) Cultures were pulse-chased with 0.5 mCi/ml [3H]palmitic acid (see text) in the absence or presence of the indicated agonists for 0–120 min. The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: (A) control (7); carbachol, 1 mM (2); isoproterenol, 100 mM (3); forskolin, 100 mM (4). (B) TPA, 200 nM (4); KCl, 60 mM (1); ionomycin, 10 mM (2). The data were analyzed by scintillation counting and are presented as average ± SD.
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
modulin undergoes dynamic palmitoylation in neurons, and if so, to determine if palmitoylation is regulated. This is the first study to report the quantification of the half-life of the neuromodulin-associated palmitoyl groups in neurons, as well as the first to directly demonstrate a dynamic cycle of palmitoylation and depalmitoylation of neuromodulin in intact neurons. The rate of turnover of palmitate associated with neuromodulin is at least 10 times faster than the rate of turnover of the protein itself. In addition, neuromodulin incorporated palmitic acid in the absence of protein synthesis, and labeling saturated within 120 min, indicating active deacylation during the labeling period. The half-life of [3H]palmitic acid on neuromodulin of approximately 5 h is relatively long, compared with other palmitoylated proteins that undergo dynamic and/or regulated acylation, suggesting that the level of neuromodulin in neurons is relatively stable. Experiments with activators of various cellular signaling pathways were carried out under conditions that measured the rate of neuromodulin depalmitoylation (pulse-chase) and the rate of palmitoylation + depalmitoylation (metabolic labeling). These experiments indicate that neuromodulin palmitoylation is not regulated by agents that activate Go or Gi or adenylyl cyclase. In addition, agonists that stimulate growth cone collapse were without effect, as were activators of PKC and increased intracellular calcium. The
159
lack of effect of a muscarinic agonist and mastoparan on palmitoylation is particularly interesting since it has been hypothesized that neuromodulin signals through Go or Gi [21] and that the G protein and palmitoylation domains overlap [13,21,25]. The physiological relevance of the interaction between non-palmitoylated neuromodulin and G proteins is unclear and remains to be demonstrated in vivo. In summary, this is the first study to report the quantification of the half-life of the neuromodulin-associated palmitoyl groups in neurons, and to show that there is a dynamic cycle of palmitoylation and depalmitoylation of neuromodulin in intact neurons. Palmitoylation of neuromodulin in cultured rat cerebellar granule cells is not regulated by increases in intracellular calcium, activation of Go or Gi proteins, PKC, or, in N1E-115 cells, by agents that induce growth cone collapse. These data suggest that the neuromodulin acylation-deacylation cycle occurs constitutively and that a constant ratio of cytosolic to membrane-associated neuromodulin is required for the normal function of neuromodulin in neurons. We thank Drs. Thomas Hinds, Yuechueng Liu, and JeanChristophe Deloulme for valuable discussions and reading of the manuscript, and members of the Storm lab for critical reading of the manuscript. This work was supported by National Institutes of Health NRSA postdoctoral fellowship
Fig. 4. Metabolic labeling of cerebellar granule cells with [3H]palmitic acid. Cerebellar cells were labeled with 0.5 mCi/ml [3H]palmitic acid in the absence or presence of the inhibitor, tunicamycin, for 0–120 min, or agonists for 0–180 min. The data were analyzed by scintillation counting, and single representative experiments are shown. (A) Cells were labeled after pretreatment with 50 mg/ml cycloheximide for 1 h in the absence of FBS, and in the absence or presence of 50 mg/ml tunicamycin. Tunicamycin (50 mg/ml) inhibited incorporation of [3H]palmitic acid by approximately 50% after 120 min. These results are representative of three independent experiments. (B–D) The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: control (6); carbachol, 1 mM (4); mastoparan, 10 mM (5); isoproterenol, 100 mM (2); forskolin, 100 mM (3); PDBU, 200 nM (2); KCl, 60 mM (2).
160
L.P. Baker, D.R. Storm / Neuroscience Letters 234 (1997) 156–160
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Fig. 5. Metabolic labeling of N1E-115 cells with [3H]palmitic acid. (A,B) Differentiating N1E-115 cells were labeled with 0.5 mCi/ml [3H]palmitic acid in the absence or presence of the indicated agonists for 0–120 min. The data were analyzed by scintillation counting, and single representative experiments are shown. The concentrations of agonists are as follows, with the number of experiments carried out for each treatment indicated in parentheses: (A) control (5); mastoparan, 10 mM (2); thrombin, 50 ng/ml (2); LPA, 1 mM (2). (B) carbachol, 1 mM (2); forskolin, 100 mM (2); PDBU, 200 nM (2); KCl, 60 mM (2).
# NS09528 (L.P.B.) and by National Institutes of Health grants NS20498 and NS31496. [1] Aigner, L., Arber, S., Kapfhammer, J.P., Laux, T., Schneider, C., Botteri, F., Brenner, H.R. and Caroni, P., Overexpression of the neural growth-associated protein GAP-43 induces sprouting in the adult nervous system of transgenic mice, Cell, 83 (1995) 269–278. [2] Alexander, K.A., Cimler, B.M., Meier, K.E. and Storm, D.R., Regulation of calmodulin binding to P-57. A neurospecific calmodulin binding protein, J. Biol. Chem., 262 (1987) 6108– 6113. [3] Andreasen, T.J., Luetje, C.W., Heideman, W. and Storm, D.R., Purification of a novel calmodulin binding protein from bovine cerebral cortex membranes, Biochemistry, 22 (1983) 4615– 4618. [4] Buelt, M.K., Glidden, B.J. and Storm, D.R., Regulation of p68 RNA helicase by calmodulin and protein kinase C, J. Biol. Chem., 269 (1994) 29367–29370. [5] Chao, S., Benowitz, L.I., Krainc, D. and Irwin, N., Use of a twohybrid system to investigate molecular interactions of GAP-43, Mol. Brain Res., 40 (1996) 195–202. [6] Gamby, C., Waage, M.C., Allen, R.G. and Baizer, L., Analysis of the role of calmodulin binding and sequestration in neuromo-
[15]
[16]
[17] [18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
dulin (GAP-43) function, J. Biol. Chem., 271 (1996) 26698– 26705. Higashijima, T., Uzu, S., Nakajima, T. and Ross, E.M., Mastoparan, a peptide toxin from wasp venom, mimics receptors by activating GTP-binding regulatory proteins (G proteins), J. Biol. Chem., 263 (1988) 6491–6494. Jalink, K., Eichholtz, T., Postma, F.R., van Corven, E.J. and Moolenaar, W.H., Lysophosphatidic acid induces neuronal shape changes via a novel, receptor-mediated signaling pathway: similarity to thrombin action, Cell Growth Diff., 4 (1993) 247–255. Jalink, K. and Moolenaar, W.H., Thrombin receptor activation causes rapid neural cell rounding and neurite retraction independent of classic second messengers, J. Cell Biol., 118 (1992) 411–419. James, G. and Olson, E., Experimental approaches to the study of reversible protein fatty acylation in mammalian cells, Methods, 1 (1990) 270–275. Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 (1970) 680–685. Lasher, R.S. and Zagan, I.S., The effect of potassium on neuronal differentiation in cultures of dissociated newborn rat cerebellum, Brain Res., 41 (1972) 482–488. Liu, Y.-C., Chapman, E.R. and Storm, D.R., Targeting of neuromodulin (GAP-43) fusion proteins to growth cones in cultured rat embryonic neurons., Neuron, 6 (1991) 411–420. Milligan, G., Parenti, M. and Magee, A.I., The dynamic role of palmitoylation in signal transduction, Trends Biochem. Sci., 20 (1995) 181–186. Patterson, S.I. and Skene, J.H.P., Novel inhibitory action of tunicamycin homologues suggests a role of dynamic protein fatty acylation in growth cone-mediated neurite extension, J. Cell Biol., 124 (1994) 521–536. Patterson, S.I. and Skene, J.H.P., Inhibition of dynamic protein palmitoylation in intact cells with tunicamycin, Methods Enzymol., 250 (1995) 284–300. Skene, J.H.P., Axonal growth-associated proteins, Annu. Rev. Neurosci., 12 (1989) 127–156. Skene, J.H.P. and Virag, I., Posttranslational membrane attachment and dynamic fatty acylation of a neuronal growth cone protein, J. Cell. Biol., 108 (1989) 613–624. Strittmatter, S.M., Cannon, S.C., Ross, E.M., Higashijima, T. and Fishman, M.C., GAP-43 augments G-protein-coupled receptor transduction in Xenopus laevis oocytes, Proc. Natl. Acad. Sci. USA, 90 (1993) 5327–5331. Strittmatter, S.M., Fankhauser, C., Huang, P.L., Mashimo, H. and Fishman, M.C., Neuronal pathfinding is abnormal in mice lacking the neuronal growth cone protein GAP-43, Cell, 80 (1995) 445–452. Strittmatter, S.M., Valenzuela, D., Kennedy, T.E., Neer, E.J. and Fishman, M.C., Go is a major growth cone protein subject to regulation by GAP-43, Nature, 344 (1990) 836–841. Strittmatter, S.M., Valenzuela, D., Sudo, Y., Linder, M.E. and Fishman, M.C., An intracellular guanine nucleotide release protein for Go; GAP-43 stimulates isolated a subunits by a novel mechanism, J. Biol. Chem., 266 (1991) 22465–22471. Suidan, H.S., Stone, S.R., Hemmings, B.A. and Monard, D., Thrombin causes neurite retraction in neuronal cells through activation of cell surface receptors, Neuron, 8 (1992) 363–375. Trenkner, E., Cerebellar cells in culture. In G. Banker and K. Goslin (Eds.), Culturing Nerve Cells, MIT Press, Cambridge, MA, 1991, pp. 283–307. Zuber, M.X., Strittmatter, S.M. and Fishman, M.C., A membrane-targeting signal in the amino terminus of the neuronal protein GAP-43, Nature, 341 (1989) 345–348.