J. Biochem. 115, 304-308 (1994)

Vasoactive Intestinal Induces Differentiation and MAP Kinase Activation in PC12h Cells

Nobuaki Okumura, Yoshie Miyatake, Tetsuya Takao, Teruya Tamaru, Katsuya Nagai, Masato Okada, and Hachiro Nakagawa Division of , Institute for Protein Research, Osaka University, Suita, Osaka 565

Received for publication, October 4, 1993

Vasoactive intestinal peptide (VIP), a coupled with adenylate cyclase, was found to induce neurite extension of PC12h cells. Neurites appeared within 1 h after addition of VIP and extended for at least 24 h. The half-maximal concentration for the effect of VIP was 50 nM. In addition to the morphological change, VIP induced expression of VGF protein, a neuron-specific protein associated with neuronal differentiation. Western blotting with anti-phosphotyrosine antibody showed that VIP stimulated tyrosine phos phorylation of two proteins of 42 and 44 kDa, which may be two isoforms of MAP kinase, erkl and erk2. Activation of MAP kinases was confirmed by ion-exchange chromatography on a Mono Q column, from which VIP-induced kinase activity was co-eluted with MAP kinase-immunoreactivity. Tyrosine-phosphorylation of MAP kinases was also stimulated by forskolin or dibutyryl cAMP, indicating that activation of MAP kinases by VIP might be mediated by cAMP. These results suggest that VIP-induced differentiation of PC12 cells is associated with cAMP-dependent activation of MAP kinases.

Key words: cAMP, MAP kinase, neuronal differentiation, PC12, Vasoactive intestinal peptide, VGF.

Differentiation and survival of neuronal cells are regulated such as mitogens and (8). Activation of MAP kinase by a variety of neurotrophic factors including nerve growth is induced by phosphorylation of both threonine and tyro factor (NGF), brain-derived neurotrophic factor (BDNF), sine residues of the enzyme as a result of successive neurotrophin 3 (NT3), and fibroblast growth factors stimulation of ras, raf-1, and MAPKK (9-11). However, (FGFs). These factors act through specific cell surface little is known about the mechanisms for cAMP-dependent receptors that have tyrosine kinase activity (1). PC12 cell differentiation of neuronal cells. In this study, we found line, a clonal cell line established from a pheochromo that, in PC12h cells, VIP stimulated extension of neurites cytoma, has been widely used as a model to study the and expression of VGF, a neuron-specific protein associated molecular mechanisms of action of neurotrophic factors, with neuronal differentiation (12-15). Using this system, since the cells differentiate into neuron-like cells in re we examined whether VIP activated MAP kinase in the sponse to NGF (2). cells. In addition to NGF, CAMP and its analogs can also induce neuronal phenotypes in PC12 cells (3, 4). These findings MATERIALSAND METHODS suggest that neuronal differentiation in PC12 cells is also stimulated by a cAMP-dependent pathway. It is not yet Materials-PC12h cells were kindly provided by Profes clear, however, what kinds of physiological factor are real sor H. Hatanaka of our institute. VIP was purchased from inducers of differentiation by the latter pathway in PC12 Peptide Institute (Osaka) and NGF was from Chemicon cells. VIP is known as a neuropeptide that increase the International (U.S.A.). Anti-MAP kinase antibody was cellular cAMP level through activation of G protein. VIP is obtained from Santa Cruz Biotechnology (U.S.A.), anti not only a putative neurotransmitter, but also a putative phosphotyrosine antibody (4G10) was from Upstate Bio neurotrophic factor for several types of neurons (5, 6). It technology (U.S.A.), and a cAMP radioimmunoassay sys has recently been shown that PACAP, a VIP-related tem and ECL blotting system were from Amersham peptide, has neurotrophic activity (7). From these findings, International plc. (England). Rabbit anti-VGF antibody we postulated that VIP is a possible neurotrophic factor was raised against synthetic peptide of VGF450-469 (15). that acts via a CAMP-dependent mechanism. Cell Culture-PC12h cells were maintained in collagen. Recently, much information has been obtained on the coated dishes containing Dulbecco's modified Eagle's signal transduction mechanisms of NGF from studies using medium (DMEM) with 5% horse serum and 5% calf serum PC 12 cells. One of the key enzymes of NGF action is MAP and subcultured every 7-8 d. For each experiment, the cells kinase, which is activated in response to a variety of stimuli were plated at a density of 1 x 104 cells/cm' and cultured for 2 d. The medium was then replaced by DMEM without Abbreviations: NGF, ; VIP, vasoactive intestinal serum and incubation was continued for 16 h. peptide. Western Blotting-PC12h cells were washed three times

304 J. Biochem. VIP and Neuronal Differentiation 305 with homogenizing buffer consisting of 0.25 M sucrose, 1 containing 5% glycerol instead of 50 mM ff-glycerophos mM EDTA, 1 mM phenylmethyl sulfonyl fluoride, 1 mM phate and eluted with a linear gradient of 0-0.5 M NaCl. sodium vanadate, and 10 pM sodium molybdate, then For determination of myelin basic protein (MBP) phospho collected in 150 ,u l of the homogenizing buffer. The cell rylating activity, a 20-p1 sample of each fraction was extracts (10 pg protein) were separated by 8 or 10% SDS incubated with 10 pg of MBP, 5 mM magnesium acetate polyacrylamide gel electrophoresis and electrophoretically and [ y-32P]ATP (2,500 cpm/pmol) in a final volume of 40 blotted to a nitrocellulose membrane (16). For Western pl at 30°C for 15 min, then half of the reaction mixture was blotting with anti-VGF or anti-MAP kinase antibody, the spotted onto a phosphocellulose paper. The paper was membrane was blocked with 1% bovine serum albumin washed extensively with 0.5% orthophosphate and the (BSA) in phosphate buffered saline (PBS), incubated with radioactivity was measured with a liquid scintillation the antiserum diluted 1:1,000 with BSA in PBS for 2 h, and counter. washed three times with PBS containing 0.1% Tween 20. Immunoreactive bands were detected by the peroxidase RESULTS anti-peroxidase method (17). Western blotting with anti phosphotyrosine antibody was carried out as described To examine the effect of VIP on neuronal differentiation, we above, except that Tris-buffered saline was used instead of tested whether VIP induced neurite extension of PC12h phosphate-buffered saline and the primary antibody was cells. As shown in Fig. 1, VIP caused morphological change diluted 1:10,000. For semiquantitative analysis of VGF, of the cells within 1 h. At 1 h after addition of VIP, many immunoreactive bands were detected by ECL blotting short processes extended from the cell body. At 6 h, the system and analyzed by scanning densitometry. neurites became larger but fewer than at 1 h. Forskolin and Protein was determined by the method of Lowry with dbcAMP also caused similar morphological change, while BSA as a standard. Molecular weight markers were myosin NGF did not induce neurites until 12-16 h after its addition (200 kDa), f3-galactosidase (116 kDa), phosphorylase b (97 (data not shown). At 24 h, the neurites induced by VIP, kDa), BSA (66 kDa), and ovalbumin (45 kDa) from Bio forskolin, or dbcAMP were further extended and became Rad. indistinguishable in shape from those induced by NGF. Determination of cAMP-To cell extracts prepared as Figure 2A shows a dose-response curve for the effect of described above, ice-cold ethanol was added to give a final VIP on neurite extension in PC12h cells. PC12h cells were concentration of 65%. After centrifugation at 10,000 rpm incubated with VIP for 24 h, and then the numbers of cells for 5 min, the supernatants were dried under vacuum and whose neurites were longer than the cell body were count the cAMP levels were determined with a radioimmunoas ed. The half-maximal concentration of VIP was about 50 say kit according to manufacturer's protocol. nM, and the effect of VIP reached maximum at a concentra Partial Purification of MAP Kinase-PC12 cells col tion of about 100 nM. At this concentration, more than 60% lected from 4 dishes (100 mm) were homogenized in 5 ml of of the cells had neurites. To examine whether neurite Buffer A, consisting of 20 mM Tris-HCI pH 7.5, 50 mM extension is correlated with elevation of cAMP, we mea f3-glycerophosphate, 5 mM EGTA, 1 mM PMSF, 12 mM sured its level in cells stimulated with VIP. As shown in Fig. f3-mercaptoethanol, 1 mM Na3VO4, and 0.1 mM Na2MoO4, 2B, VIP increased the cAMP level more than 10-fold at 20 and centrifuged at 100,000 X g for 30 min. The supernatant nM and more than 100-fold at 100 nM. At concentrations was applied on a Mono Q column equilibrated with Buffer A over 100 nM, VIP caused maximum effect on cAMP forma

Fig. 1. Effect of VIP on neurite extension of PC12h cells. PC12h cells were serum-starved for 16 h, then incubated in serum-free medium with 200 nM VIP for 0 h (a), 1 h (b), 24 h (c), or without VIP for 24 h (d).

Vol. 115, No. 2, 1994 306 N. Okumura et al.

Fig. 3. Induction of VGF protein by VIP, forskolin, and dbcAMP. A: PC12 cells were incubated in serum-free medium (a) containing 100 nM VIP (b), 10,uM forskolin (c), or 1 mM dbcAMP (d) for 24h. Crude extracts from the cells were analyzed by Western Fig. 2. Dose-response curve for effect of VIP on neurite exten blotting with anti-VGF antibody and immunoreactive bands were sion of PC12h cells. A: PC12h cells were serum-starved for 16 h, detected by peroxidase-antiperoxidase reaction. B: PC12 cells in then incubated in serum-free medium containing VIP at the concen cubated with 100 nM VIP for indicated times were analyzed by trations indicated. After incubation for 24 h, the number of cells Western blotting with anti-VGF antiserum. Immunoreactive bands having neurites longer than their cell bodies were counted. B: After were detected by ECL blotting system and quantitated by scanning serum-starvation, PC12h cells were stimulated for 10 min with VIP densitometry. Values are means ± SE from three independent experi at the concentrations indicated and cellular cAMP levels were ments. measured. Values are means±SE from three independent experi ments.

Fig. 4. Tyrosine phosphorylation induced by VIP. A: PC12h stimulated with 200 nM VIP for indicated times and subjected to cells were serum-starved for 16 h and stimulated for 5 min with VIP Western blotting with anti-phosphotyrosine antibody. D: PC12h cells at concentrations indicated. Crude homogenates from the cells were were serum-starved for 0 h (lanes 1 and 2), 1 h (lanes 3 and 4), or 16 analyzed by Western blotting with anti-phosphotyrosine antibody. B: h (lanes 5 and 6), then stimulated with 200 nM VIP (lanes 2, 4, and 6) Crude homogenate from PC12h cells stimulated with 200 nM VIP or vehicle (lanes 1, 3, and 5) for 5 min. Crude homogenates from the (lanes 2 and 4) or unstimulated (lanes 1 and 3) were analyzed by cells were analyzed by Western blotting with anti-phosphotyrosine Western blotting with anti-phosphotyrosine antibody (lanes 1 and 2) antibody. or anti-MAP kinase antibody (lanes 3 and 4). C: PC12h cells were

J. Biochem. VIP and Neuronal Differentiation 307

Fig. 6. Effects of VIP, forskolin, and dbcAMP on tyrosine phosphorylation in PC12h cells. A: PC12h cells were serum starved for 16 h (lane 1), then stimulated with 200 nM VIP (lane 2), 10 kM forskolin (lane 3), and 1 mM dbcAMP (lane 4). Extracts from the cells were analyzed by Western blotting with anti -phosphotyro sine antibody. B: PC12h cells serum-starved for 16 h were stimulated with dbcAMP for indicated times and analyzed by Western blotting with anti-phosphotyrosine antibody.

min after VIP addition (Fig. 4C) and took place both in the presence and absence of serum (Fig. 4D). To confirm that MAP kinases are activated in response to Fig. 5. Mono Q column chromatography of PC12h cells. PC12h cells were serum-starved for 16 h and stimulated with 5 nM NGF VIP, MAP kinases were partially purified from stimulated (panel A) or 200 nM VIP (panel B) for 5 min. Soluble fractions from and unstimulated cells by Mono Q column chromatography. control cells (0) or stimulated cells (•) were separated by ion As shown in Fig. 5, VIP as well as NGF activated a kinase exchange chromatography on a Mono Q column and kinase activity activity that was eluted from the column with 0.2-0.3 M was measured with myelin basic protein as a substrate. The inset NaCl. The elution profile of the kinase was similar to that of shows Western blotting of each fraction with anti-MAP kinase MAP kinases, as detected by Western blotting with anti antibody. MAP kinase antibody. Since VIP increases the cAMP level in PC12h cells, it is plausible that cAMP mediates the effect of VIP on activa tion as well as neurite extension. tion of MAP kinases. If this is the case, MAP kinase might Western blotting with anti-VGF antibody showed that also be activated when the cAMP-dependent pathway is VIP induced VGF proteins, doublet proteins with molecular stimulated directly. To test this possibility, we examined weight of 75 and 90 kDa (Fig. 3A). Since VIP increases the effects of forskolin and dibutyryl cAMP (dbcAMP) on cAMP levels in PC12h cells, it is plausible that induction of tyrosine phosphorylation of cellular proteins. As shown in VGF by VIP is mediated by cAMP. Confirming this possi Fig. 6, both forskolin and dbcAMP, as well as VIP, stimu bility, forskolin and dbcAMP also stimulated induction of lated tyrosine phosphorylation of MAP kinases, suggesting VGF protein. The induction of VGF began at 2 h after VIP that they are activated by the cAMP-dependent pathway. addition and reached the maximum level at about 6 h (Fig. 3B). DISCUSSION Figure 4 shows VIP-induced phosphorylation of MAP kinases. Western blotting with anti-phosphotyrosine anti Differentiation of neuronal cells can be induced not only by body showed that VIP induced tyrosine phosphorylation of stimulation of tyrosine kinase receptors, but also by a 42-kDa protein, which was probably an isoform of MAP activation of a cAMP-dependent pathway (3, 4). In the kinase, erk2 (Fig. 4A). Erk 1, another isoform with Mr of present study, we showed that VIP, a neuropeptide stimu 44,000, also seemed to be tyrosine-phosphorylated, but the lating adenylate cyclase, induced neurite extension and change was not very clear because of strong interference by VGF expression in PC12h cells, suggesting that VIP may an overlapping band. Western blotting with anti-MAP stimulate differentiation of PC12h cells by CAMP-depend kinase antibody showed that each immunoreactive band of ent mechanisms. Using this system, we showed that VIP, erkl and erk2 split into two bands after stimulation with forskolin and dbcAMP activated MAP kinases. VIP, confirming that MAP kinases were phosphorylated in VIP is present in endocrine cells of the gastrointestinal response to VIP (Fig. 4B). The concentration of VIP tract and in a variety of central and peripheral neurons. In required for MAP kinase phosphorylation was at least 10 the peripheral nervous system, VIP is present in pre and nM, which elevated the cAMP level about twofold (Fig. post-synaptic fibers of the sympathetic nervous system 2B). The tyrosine-phosphorylation was induced within 5 including the fibers innervating adrenal chromaffin cells

Vol. 115, No. 2, 1994 308 N . Okumura et al.

(18). It is known that VIP activates tyrosine hydroxylase kinase receptors of growth factors is downstream of ras, activity in primary cultured adrenal chromaffin cells (19). raf-1, and MAPKK (9-11). Recently, it was shown that In addition, VIP induces morphological differentiation of activation of ras is mediated by interaction of activated primary cultured neurons from the superior cervical gan receptors with SOS and GRB2 (24). However, VIP-induced glion (5). It, thus, seems likely that the effects of VIP on activation of MAP kinases might be mediated by a some PC12h cells reflect the neurotrophic activity of VIP. It was what different pathway, since VIP receptors do not have reported, however, that VIP did not stimulate neurite tyrosine kinase activity (25). extension of PC 12 cells even at 1,u M (7). This finding is not consistent with our present results, but the discrepancy can We thank Dr. S. Aimoto, Research Center for Protein Engineering of be explained on the basis of the difference of sensitivity of our institute, for peptide synthesis. the subclones of the cell line used, because the cells in the earlier study failed to produce cAMP on stimulation by 1 REFERENCES ,uMVIP(7). 1. Chao, M.V. (1992) Neuron 9, 583-593 VGF is a neuron-specific protein induced by neurotrophic 2. Greene, L.A. & Tischler, A.S. (1983) Adv. Cell. Neurobiol. 3, factors (13-16). Its function is unknown, but it appears to 373-414 be essential for neuronal differentiation. In the present 3. Schubert, D., LaCorbiere, M., Whitlock, C., & Stallcup, W. study, we found that, in addition to other neurotrophic (1978) Nature 273, 718-723 factors, VIP induced VGF protein in PC 12h cells, indicating 4. Gunning,P.W., Landreth, G.E., Bothwell,M.A., &Shooter, E.M. that VIP induced neuronal differentiation of the cells. In (1981) J. Cell Biol. 89, 240-245 5. Pincus, D.W., DiCicco-Bloom,E.M., & Black LB. (1990) Nature addition, this finding suggests that the signal transduction 343,564-567 pathways of VIP and NGF are convergent, even though 6. Pence, J.C. & Shorter, N.A. (1990) Cancer Res. 50, 5177-5183 they stimulate different intracellular signals. It remains 7. Deutsch, P.J. & Sun, Y. (1992) J. Biol. Chem. 267, 5108-5113 undetermined, however, what mechanisms are responsible 8. Sturgill, T.W. & Wu, J. (1991) Biochim. Biophys. Acta 1092, for the induction of VGF protein. 350-357 9. Thomas, S.M., DeMarco, M., D'Arcangelo, G., Halegoua, S., & In PC12 cells, NGF binding to its receptor leads to Brugge, J.S. (1992) Cell 68, 1031-1040 activation of several different pathways. Of these, some are 10. Wood, K.W., Sarnecki, C., Roberts, T.M., & Blenis, J. (1992) common to cell proliferation, while others are unique to Cell 68, 1041-1050 neuronal differentiation (9, 10, 20). MAP kinases are 11. Kosako, H., Nishida, E., & Gotoh, Y. (1993) EMBO J. 12, 787 serine/threonine kinases activated by NGF as well as 794 mitogens (8). Our present results that MAP kinases were 12. Levi, A., Eldridge, J.D., & Paterson, B.M. (1985) Science 229, activated by VIP indicate the involvement of the kinases in 393-395 13. Possenti, R., Eldridge, J.D., Paterson, B.M., Grasso, A., & Levi, the signal transduction pathway of VIP. The VIP concentra A. (1989) EMBOJ. 8, 2217-2223 tion required for MAP kinase activation was, however, 14. Salton, S.R.J., Fischberg, D.J., & Dong,K. -W. (1991) Mol. Cell. lower than those for neurite extension. One possible inter Biol. 11, 2335-2349 pretation is that neuronal differentiation is induced only 15. Miyatake, Y., Okumura, N., Nagai, K., & Nakagawa, H. (1993) when prolonged activation of MAP kinases is achieved by Biochem.Mol. Biol. Int. 30, 231-236 high doses of VIP. Alternatively, activation of MAP kinases 16. Towbin,H., Staehelin, T., & Gordon,J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354 may not be enough to induce neuronal differentiation and 17. Harlow, E. & Lane, D. (1988) Antibodies, pp. 471-510, Cold another pathway stimulated by high doses of VIP may also Spring Harbor Laboratory, New York be essential to the process. 18. Yoshikawa, M., Saito, H., Sano T., Ohuchi, T., Ishimura, Y., It was reported that MAP kinases are activated not only Morita, K., Saito, S., & Oka, M. (1990) Neurosci. Lett. 111, 75 by simulation of tyrosine kinase receptors, but also by G 79 19. Tischler, A.S., Perlman, R.L., Costopoulos,D., & Horwitz, J. protein-coupled receptors such as (21). Since (1985) Neurosci. Lett. 61, 141-146 VIP increased the cAMP level in PC12h cells, it is possible 20. Rabin, S.J., Cleghon,V., & Kaplan, D.R. (1993) Mob Cell. Biol. that VIP-stimulated activation of MAP kinase is mediated 13, 2203-2213 by cAMP. Although previous reports showed that cAMP 21. Ahn, N.G., Robbins,D.J., Haycock,J.W., Seger, R., Cobb,M.H., did not have an effect on MAP kinase activity (22, 23), we & Krebs, E.G. (1992) J. Neurochem.59, 147-156 provided evidence that MAP kinases could be activated by 22. Sano, M. (1992) J. Neurochem.59, 1263-1272 forskolin and dbcAMP. The discrepancy might be due to a 23. Tsao, H., Aletta, J.M., &Greene, L.A. (1990) J. Biol. Chem.265, difference in the method of detection of MAP kinase 15471-15480 24. McCormick,F. (1993) Nature 363, 15-16 activation or in the subclone used. 25. Ishihara, T., Shigemoto, R., Mori, K., Takahashi, K., & Nagata, It is not known what mechanisms operate for activation S. (1992) Neuron 8, 811-819 of MAP kinases by VIP. MAP kinase activation by tyrosine

J. Biochem.