Epidermal growth factor, but not nerve growth factor, stimulates tyrosine-specific protein-kinase activity in pheochromocytoma (PC12) plasma membranes

BIOCttlMIE, 1985, 67, 1177-1183

Epidermal growth factor, but not nerve growth factor, stimulates tyrosine-specific protein-kinase activity in pheochromocytoma (PC12) plasma membranes. J o h a n n e s B O O N S T R A * , Paul T. v a n d e r S A A G °, Alie F E I J E N , A r n o l d B I S S C H O P * * a n d Siegfried d e L A A T .

Hubrecht Laboratory, Uppsalalaan 8, 3584 CT Utrecht. Department of Molecular Cell Biology University of Utrecht, Utrecht. Laboratory of Carcinogenesis and Mutagenesis, National Institute of Health, Bilthoven, The Netherlands. (Requ le 6-2-1985. accept~ le 21-3-1985)

R6sum6 B

Les cellules de phdochromocytoma (PCI2) de rat contiennent des rdcepteurs de membrane plasmique spdcifiques et pour le facteur de croissance dpidermique (EFG) et pour le facteur de croissance nerveuse (NGF). Alors que l'addition d'EFG attx cellules PC12 provoque une augmentation persistante de la prolifdration, l'addition de NGF induit une stimulation de croissance transitoire, suivie d'un arrdt de croissance et d'une diffdrenciation neuronale. Malgrd ces diffdrences de rdponse biologique, EGF et NGF partagent un certain nombre de rdponses mddides par les rdcepteurs et qui sont vraisemblablement tides h leur effet sur la prolifdration cellulaire. Dans le prdsent article nous montrons que EGF, mais pas NGF, est capable de stimuler la phosphorylaion des protdines membranaires. En outre, EGF est capable de stimuler la phosphorylation d'un peptide synthdtiques (RR-SRC) par les membranes de PCI2 d'une mani~re ddpendant de la concentration. L'analyse cin(tique de la rdaction de phosphorylation indique que EGF augmente la Vmax de 13 h 70pmoles/mhl/mg protdine, alors que l'on n'observe attcun changement de Kin. De plus, EGF est capable de stimuler la phosphorylation de la tyroshle de l'angiotensine I et I1, au m~me degr~ que le RR-SRC. En revanche, on n'a observd aucun effet du NGF sur la phosphorylation des peptides par les membranes PC12. Des expdriences de pontage ddmontrent la prdsence de rdcepteurs aussi bien pour EGF que pour NGF dans les membranes PC12. Ces effets diffdrents d'EGF et de NGF sur l'activation de l'activitd protdine kinase associde aux membranes ddmontrent que NGF pourrait dtre capable de stimuler transitoirement la croissance sans stimuler I'activitd protdine kinase.

S u m m a r y - - Rat pheochromocytonla (PCI2) cells contain specific plasma nlembrane receptors for both epidermal growth factor (EGF) and nerve growth factor (NGF). Whereas EGF addition to PCI2 cells causes a persistant enhancement of proliferation, NGF addition induces a transient stimulation of growth, followed by growth arrest and neuronal differentiation. Despite these

0 To whom reprint requestsshould be addressed.

1178

J. Boonstra and coll. differences hi biological response, EGF and NGF share a number of earl), receptor-mediated responses, which are likely te be related to their effect on cell proliferation, hi this paper we show that EGF, but not NGF, is able to stinndate the phosphorylation of membrane protehts, hs addition, EGF was able to stimulate phosphorylation of a synthetic peptide (RR-SRC) by PCI2 membranes ht a concentration-dependent manner. Kinetic analysis of the phosphorylation reaction indicated that EGF increased the Vmax front 13 to 70 pmoles/min/mg protehz, while no change was observed in Kin. Furthermore, EGF was able to sthnulate tyrosine phosphorylation of angiotensin I and II, to the same extent as RR-SRC. In contrast no effects of NGF on peptide phosphorylation by PC12 membranes were observed. Cross-linkhtg experiments dentonstrated the presence of receptors for both NGF and EGF ht PC12. membranes. These different effects of NGF and EGF on activation of membrane-associated proteht-kinase activity demonstrate that NGF might be able to stimulate growth transientl), without stimulating protein kinase activity.

Introduction

In the presence of nerve growth factor (NGF), rat pheochromocytoma cells (clone PCI2) acquire within a few days a phenotype which resembles sympathetic neurons [1, 2]. The first interaction of NGF with PCI2 cells occurs through binding to specific high affinity receptor sites at the cell surface [3-6]. In addition to NGF-receptors, PCI2 cells contain distinct cell surface receptors for epidermal growth factor (EGF) [7-9]. NGF and EGF display, however, opposite biological effects : NGF causes initially a transient stimulation of growth followed by an arrest of cell proliferation and induction of differentiation whereas, EGF enhances the cell proliferation rate persistently [10]. Despite these different biological effects, these factors have a number of characteristics in common, including the induction of ornithine decarboxylase [9, I1], the enhancement of cellular adhesion [8], changes in surface morphology and coated pit formation [12], changes in phospho-inositide metabolism and increases in cation and nutrient transport [9, 10, 13]. These rapid responses of NGF in PCI2 cells can be correlated with the enhancement of proliferation and are dissociable with the differentiation inducing effect of NGF. Recently it has been demonstrated that a number of growth factors, such as insulin [14, 15], insulin-like growth factor [16, 17], EGF [18-20] and platelet-derived growth factor (PDGF) [21] cause a rapid increase in the phosphorylation state of cellular proteins, including the receptor molecule itself. The kinase activity was associated with the receptor molecule, the phosphorylation being specifically located on tyrosine-residues.

Furthermore, NGF has also been demonstrated to change the phosphorylation pattern of a variety of cytoplasmic proteins [22-24]. EGF, PDGF and transforming growth factors have also been demonstrated to stimulate the phosphorylation of a synthetic, tyrosine-containing peptide by A431 cell membranes [25,26]. Thus phosphorylation of the synthetic peptide can be used as a simple method to demonstrate the presence of tyrosine phosphorylating systems in cellular membranes. This was used to demonstrate tyrosine phosphorylation in rat spleen and other tissues [27], normal and leukaemic human haematopoietic cells [28], lymphoma cells [29] and during sea urchin embryogenesis [30]. In the course of our studies on the mechanism of action of both EGF and NGF in PCI2 cells, the effects of these factors were studied on plasma membrane proteins and synthetic tyrosine-peptide phosphorylation. It is shown here that EGF, but not NGF, is able to stimulate (tyrosine) protein-kinase activity in PCI2 plasma membranes. The absence of NGF-stirnulated protein and synthetic peptide phosphorylation was not due to the absence of NGF-receptors in PCI2 membranes. This difference might be related to the different biological effects of NGF and EGF on PCI2 cells.

Materials and Methods

~laterials [y-32P]ATP (S.A. 3000 Ci/mmol), ~25I-EGF (S.A. 100 p.Ci//.tg EGF) and Na~:5I (carrier-free) were obtained from New England Nuclear (Boston, Mass).

Effect of EGF and NGF on tyroshle phosphorylation Sources of non-radioactive materials used were as follows : EGF (receptor grade) and NGF (2.5 S) from Collaborative Research Corp. (Waltham, Mass); disuccinimidylsuberate and ethyldimethylisopropylaminocarbodiimide from Pierce Chemical Co. (Rockford, IL); Angiotensin I and II from Boehringer (Mannheim, FRG). The tyrosine-containing peptide was synthesized as described [29]. ~25I-NGF was prepared by the chloramine T method [31] to a specific activity of approximately 150 cpm per pg protein.

Cell culture and preparation of membranes PCI2 cells were grown at 37°C in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 5 % heat-inactivated horse serum, as described previously [1, 2]. Plasma membranes were prepared according to the method described by Thom et al. [32].

Phosphorylation assays Membrane phosphorylation. The phosphorylation assay was performed in a total volume of 601al. The reaction mixture contained 20 mM Hepes (pH 7.4), 2 mM MnCI, 50 laM Na3VO4, 10 mM p-nitrophenylphosphate, 0.125 mg/ml bovine serum albumin and 10 laM [y.3.,P]ATP (final specific activity : 5000-10,000 cpm/pmol). The phosphorylation reaction was started by the addition of membranes (approximately 50 lag protein). The reaction was stopped after 10 mn at 0°C by the addition of 2% SDS in 50mM Tris-HCl (pH 6.8) and subsequently boiled for 5 mn. These samples were subjected to gel electrophoresis and autoradiography, as described below. Peptide phosphorylation. The peptide phosphorylation assay was performed essentially as described [25, 26] in a total volume of 30 lal. The reaction mixture contained 20 mM Hepes (pH 7.4), 2 mM MnCI2, 0.2 % Nonidet P40, 10 mM p-nitrophenylphosphate, 50 laM Na3VO4 and 10 pM [y?"P]ATP (final specific activity : 5000-I0,000 cpm/pmol). The peptide substrate was added to a final concentration of 2 mM. The phosphorylation reaction was started by the addition of membranes (15-20 lag protein). The reaction was stopped after 3 mn of incubation at 30°C, by addition of 50 lal trichloroacetic acid (5 %) and 20 I11 bovine serum albumin (25 mg/ml). The precipitated proteins were pelleted by centrifugation (4mn at 10,000xg) and a 40 I.tl aliquot of the supernatant was spotted on a 4 cm 2 phosphocellulose paper (Whatman P81). The papers were washed 6 times for 2 mn with 75 mM phosphoric acid, and dried. The dried papers were placed in a scintillation vial containing 3.5 ml Dynagel and the vials were counted for 32p in a Packard 2420 liquid scintillation counter. Non-specific phosphorylation was corrected by subtracting the radioactivity bound to phosphocellulose paper in assays performed in the absence of peptide.

1179

Cross-linkhlg e.rperiments Cross-linking of ~2~I-EGF to PCI2 cell membranes has been performed as described [33]. Plasma membranes (20 lag membrane protein) were incubated in a total volume of 100 lal, containing I0 mM Hepes (pH 7.4), i % bovine serum albumin and 0.5 laCi t25I-EGF, for 2 h at room temperature. Subsequently I I.tl of 0.1 M disuccinimidylsuberate (freshly made in DMSO) was added and the mixture was incubated for 15 mn at room temperature. Unreacted reagent was quenched by addition of 0.9 ml Tris-HCI and 1 mM EDTA (pH 7.4). Cross-linking of ~25I-NGF to PCI2 membranes has been performed as described [5]. Plasma membranes were incubated in the presence of ~25I-NGF (approximately 50 ng/ml) in 10 mM Hepes (pH 7.4) and 1% bovine serum albumin for 2 h at room temperature. Subsequently. ethyldimethylisopropylaminocarbodiimide was added to a final concentration of 2 mM for 15 mn at room temperature. Unreacted reagent was quenched by the addition of 0.9 ml 20 mM Tris-HCI (pH 7.4).

SDS-polyacrylamide autoradiography

gel

electrophoresis

and

The affinity labeled membranes were pelleted by centrifugation and the pellets solved in 50 tli 2 % SDS in 50mM Tris-HCI (pH 6.8), and boiled for 5 mn. These samples were analyzed by one-dimensional SDS-polyacrylamide gel elefitrophoresis, according to the discontinuous system, of Laemmli [34], using 7-15 % polyacrylamide gradient gels. The gels were stained and dried on Whatman 3 MM paper, and subjected to autoradiography on Kodak X-Omat AR film with X-Omatic TM regular intensifying screens.

Results and discussion P C I 2 cells express functional receptor sites at their cell surface for N G F as well as E G F and show diverse biological responses to these polypeptide growth factors. This is why P C I 2 cells. provide a suitable system for analysis o f the nature and relevance o f the early cellular responses evoked by the respective ligand-receptor interactions. Here we c o m p a r e the ability o f N G F and E G F to stimulate protein kinase activity in general, and tyrosine-protein-kinase activity in particular, in P C I 2 plasma membranes. The first indications for differential effects o f N G F and E G F on protein kinase-activation were obtained by monitoring the p h o s p h o r y l a t i o n o f P C I 2 plasma m e m b r a n e proteins in the presence or absence o f E G F (1.6 lag/ml) or N G F (1.6 I.tg/ml) after a 10 mn preincubation at 0°C.

1180

J. Boonstra and coll.

As shown in Figure 1, EGF stimulated phosphorylation of the E G F receptor (p155), whereas no stimulation of protein phosphorylation was observed upon addition of NGF. Since the absence of NGF-stimulated protein phosphorylation in PCI2 plasma membranes (Fig. 1) might be due to the absence of physiological substrates under the experimental conditions used, the phosphorylating capacity of N G F and E G F was also determined by the use o f a synthetic peptide, called RR-SRC. The sequence o f this peptide, Arg-Arg-LeuIle-Glu-Asp-Ala-Glu-Tyr-Ala-Ala-Arg-Gly is related to the sequence of the site o f autophosphorylation in pp60 ~rc,containing a single tyrosine residue corresponding to position 416 of the transforming protein [35,36]. RR-SRC has been demonstrated to serve as a substrate for EGFstimulated, tyrosine-directed protein kinase in cell membranes [25,30], and, as such, provides a simple probe for tyrosine phosphorylation. As shown in Table I, E G F stimulated the phosphorylation of RR-SRC from 11.3 to 30.9 pmoles/

~150

min/mg protein, in contrast to NGF. Furthermore it has been demonstrated recently that the EGF-receptor kinase of A431 cells was also able to phosphorylate angiotensin I and II at tyrosine residues [37]. As shown in Table I, EGF also stimulated phosphorylation of these peptides by PCI2 membranes, comparable to RR-SRC, whereas no stimulation o f phosphorylation of angiotensin I and II by N G F was observed. Finally, the EGF stimulated protein-kinase activity was specifically directed to tyrosine residues, since no stimulation of phosphorylation of a synthetic serine-containing peptide by E G F was observed (unpublished observations). It might be argued that the absence of any effect of N G F on the phosphorylation of membrane proteins and synthetic tyrosine-containing peptides could be due to preferential loss of N G F receptors during the plasma membrane isolation procedure. This is why the presence o f E G F and N G F receptors was checked by chemical crosslinking experiments. PCI2 membranes were incubated with ~25I-EGF and ~25I-NGF, respectively, for 2 hours at room temperature. The bound E G F and N G F were then cross-linked to their receptors by the chemical cross-linkers disuccinimidylsuberate and ethyldimethylisopropylaminocarbodiimide, respectively. Following SDS-polyacrylamide gel electrophoresis and autoradiography it was shown that both E G F and N G F were linked to their receptors (Mr 165 kd and 95 kd, respectively) in a concentration dependent manner

TABLE |

,.o6 t 1

2

D---2o

Effect of EGF and NGF on synthetic peptide phosphorylation b)' PCI2 plasma membranes. Peptide phosphorylation was performed as described under Materials and Methods. The final concentrations used were follows : EGF, 1.6 I.tg/ml; NGF, 1.6 pg/ml; RR-SRC, angiotensin I and angiotensin II, 2 mM. The points represent the mean of a triplicate measurement from a representative experiment. The numbers between brackets indicate the stimulation factor as compared to control values.

RR-SRC

3

FIG. 1. - - Effect of EGF and NGF on protein phosphor)'lation b)" PCI2 membranes. Membrane phosphorylation was performed as described under Materials and Methods in the absence (lane I} or in the presence of 1.6 p g / m l N G F (lane2) or 1.61ag/ml EGF (lane 3).

Control EGF NGF

11.3 30.9 (2.7) 10.7 (0.9)

Angiotensin I Angiotensin II (pmol/min/mg protein) 14.1 34.9 (2.5) 13.6 (i)

6.8 20.2 (3.0) 4.9 (0.7)

Effect o f E G F and N G F on O'roshle phosphorylation

(Fig. 2). The apparent molecular weight of the NGF-receptor complex of 95 kd is in perfect agreement with recent reports on the NGF-receptor complex from PCI2 cell membranes [5] and from human melanoma (A875) cells [38]. It should be realized, however, that PCI2 cells were shown to contain two classes of NGF-binding sites, and in this respect the biologically-active class might possibly be lost during membrane preparation. Characterization of E G F stimulated RR-SRC phosphorylation demonstrated that the maximal phosphorylation rate was maintained for 6 min, after which a decrease was observed. Furthermore, the RR-SRC phosphorylation was dependent upon the EGF concentration with maximal rates of 3 l.tg/ml or higher and with half-maximal effects at approximately 0.5 p g / m l EGF. The phosphorylation rate increased also linearly with the protein concentration until approximately 1 mg/ml, after which no further increase was observed in a 3 mn assay (data not shown).

1181

The kinetic constants for the peptide phosphorylation by PCI2 membranes in the presence and absence of E G F were determined, and the Lineweaver-Burk plots are shown in Figure 3. In the absence of EGF, the reaction exhibited a Km of 1.4 mM with a Vm,x of 14 p m o l e s / m i n / m g protein. In the presence of E G F at a concentration of 0.3 and 1.6 pg/ml, the Vm~ was increased to 32 and 66 p m o l e s / m i n / m g protein, respectively, whereas no change in Km was observed (Fig. 3). A I0 mn preincubation of PCi2 membranes with

•

°

c o~

~ o2

01

,

-,

z:*~

FIG.

3.

!

, --

i

2

~

~

~

Lineweaver-Burk plotforpeptidephosphorl'lation

by PC12 plasma membranes. 155--,'-

:~

1

2 EGF

3

"--95

4

NGF

F1G. 2. -- Affinity-labeling of EGF and NGF receptors in PCI2 plasma membranes.

~"SI-NGFand ~"SI-EGFwere cross-linked to PCI2 plasma membranes, as described under Materials and Methods. ~:"I-EGF labeling in the absence (lane 2) or presence (lane 1) of unlabeled EGF (5 ~g/ml). J"SI-NGFlabeling in the absence (lane 4) or presence (lane 3) of unlabeled NGF (2 !.tg/ml).

Peptide phosphorylation was performed as described under Materials and Methods in the absence (e) or presence of 0.3 (o) or 1.6 (a) pg/ml EGF. • represent points in which the membranes were preincubated with 1.6 lag/ml EGF for 10 mn at 3°C.

1.6 pg/ml EGF resulted in a slightly higher Vm,x of 74 p m o l e s / m i n / m g protein, whereas again no change in Km was observed (Fig. 3). The Km values were of the same order of magnitude as those reported for A431 plasma membranes [25,26], but the Vm,x values were considerably lower. This difference is likely to be due to the large difference in number of EGF receptors, and hence kinase molecules, per cell, 2 x 10 6 receptors/cell in A431 cells compared to 0.6x l0 s receptors/cell in PCI2 cells [11]. Other cell types with similar numbers of receptors per cell, such as 3T3 cells, were reported to have much lower phosphorylation rates than those shown here. Compared to those cells however, PCI2 cells are rather small and thus the number of receptors per unit of cell surface area might be higher, thereby explaining the difference.

1182

J. Boonstra and coll.

The absence of NGF-stimulated receptor-associated tyrosine protein-kinase activity distinguishes the NGF-receptor complex from the other polypeptide growth factor-receptor complexes, such as EGF, PDGF, insulin and insulin-like growth factor [14-21]. This finding is all the more striking since N G F and EGF show a great similarity in a variety of other early cellular responses in PCI2 cells. These include the induction o f ornithine decarboxylase [9, 11], the enhancement of cellular adhesion [8] and the stimulation of amiloride-sensitive electroneutral N a + , H ÷ exchange [10]. Our present results thus indicate that tyrosine protein-kinase activation is not necessarily coupled to other early responses induced by the growth factor binding. Independently, evidence was provided that the activation of N a + , H + exchange results from an ailosteric modification of the cytoplasmic H ÷ binding site, and it was suggested that this is linked to the activation of Ca 2+- and lipid-dependent protein kinase C [39, 40]. Apparently, receptor activation results in a multitude of initially independent cellular responses. Recently, we have demonstrated that NGFinduced differentiation of PCI2 cells is preceded by a transient increase in the rate of proliferation [10]. The effects of N G F on growth and differentiation were dissociable by inhibition of the early ionic responses, growth being sensitive to amiloride whereas differentiation was unaffected [10]. Others have reported stimulting effects of N G F on cell proliferation of particular PCi2 subclones [41, 42] and of neuroblastoma cells [43]. The inability of N G F to stimulate tyrosine protein-kinase activity by PCI2 membranes thus questions the absolute requirement of growth stimulation on tyrosine phosphorylating activity. However, it could well be that the differential responses of E G F and N G F on tyrosine phosphorylation mediate their different effects on cell proliferation, i.e. persistent versus transient. PCI2 cells, by their capacity to respond to E G F and NGF, provide a suitable system to elucidate the relevance of early responses to growth factor binding, and further studies along these lines are currently in progress.

Acknowledgment lVe would like to thank Dr. A.J. Slotboom, Dept. o f Biochemistrk; State University of Utrecht, for synthesizing the RR-SRC peptide, and Mr. P. Meyer for s k i e d technical assistance.

REFERENCES I. Tischler, A.S. & Greene, L.A. (1975) Nature (Lond.), 258, 341-342. 2. Dichter, M.A., Tischler, A.S. & Greene, L.A. (1977) Nature (Lond.), 268, 501-504. 3. Herrup, K. & Thoenen, H. (1979) Exp. Cell. Res., 121, 71-78. 4. Landreth, G.E. & Shooter, E.M. (1980) Proc. Natl. Acad. Sci. USA, 77, 4751.4755. 5. Grob, P.M., Berlot, C.H. & Bothwell, M.A. (1983) Proc. Natl. Acad. Sci. USA, 80, 6819-6823. 6. Cattaneo, A., Biocca, S., Nasi, S. & Calissano, P. (1983) Fur. J. Biochem., 135, 285-290. 7. Huff, K.R. & Guroff, G. (1979) Biochem. Biophys. Res. Comm., 89, 175-180. 8. Chandler, C.E. & Herschman, H.R. (1980) Cell. Ph)'siol., 105, 275-285. 9. Huff, K.R., End, D. & Guroff, G. (1981) .L Cell. Biol., 88, 189-198. I0. Boonstra, J., Moolenaar, W.H., Harrison, P.H., Moed, P., van der Saag, P.T. & de Laat, S.W. (1983) J. Cell BioL, 97, 92-98. I1. Guroff, G., Dickens, G. & End, D. (1981) J'. Neurochem., 38, 342-349. 12. Connolly, J.L., Green, S.A. & Greene, L.A. (1984) J. Cell Biol., 98, 457-465. 13. McGuire, J.C. & Greene, L.A. (1979) .L Biol. Chem., 254, 3362-3367. 14. Kasuga, M., Zick, Y., Blithe, D.L., Crettaz, M. & Kahn, C.R. (1982) Nature (Lond.). 298, 667-669. 15. Kasuga, M., Fujita-Yamaguchi, Y., Blithe, D.L. & Kahn, C.R. (1983) Proc. Natl. Acad. ScL USA, 80, 2137-2141. 16. Rubin, J.B., Shia, M.A. & Pilch, P.F. (1983) Nature (Lond.), 305, 438-440. 17. Jacobs, S., Kull, F.C., Earp, H.S., Svoboda, M.E., Van Wijk, J.J. & Cuatrecasas, P. (1983) J. Biol. Chem.; 258, 9581-9584. 18. Cohen, S., Carpenter, G. & King, L. (1980) J. Biol. Chem.. 255, 4834-4842. 19. Cassel, D., Pike, L.J., Grant, G.A., Krebs, E.G. & Glaser, L. (1983) .L Biol. Chem., 258, 2945-2950. 20. Erneux, C., Cohen, S. & Garbers, D.L. (1983) J'. BioL Chem.. 258, 4137-4142. 21. Pike, L.J., Bowen-Pope, D.F., Ross, R. & Krebs, E.G. (1983) J. Biol. Chem., 258, 9383-9390. 22. Halegoua, S. & Patrick, J. (1980) Cell, 22, 571-581. 23. End, D., Hanson, M., Hashimoto, S. & Guroff, G. (1982) J. Biol. Chem.. 257, 9223-9225. 24. End, D., Tolson, N., Hashimoto, S. & Guroff, G. (1983) J'. Biol. Chem., 258, 6549-6556. 25. Pike, L.J., Gallis, B., Casnellie, J.E., Bornstein, P. & Krebs, E.G. (1982) Proc. Natl. Acad. Sci. USA, 79, 1443-1447. 26. Pike, L.J., Marquardt, H., Todaro, G.J., Gallis, B., Casnellie, J.E., Bornstein, P. & Krebs, E.~3. (1982) .I. Biol. Chem., 257, 14628-14631. 27. Swarup, G., Dasgupta, J.D. & Garbers, D.L. (1983) J. Biol. Chem.. 258, 10341-10347.

Effect o f E G F and N G F on tyroshze phosphorylation 28. Tuy, F.P.D., Henry, ,i., Rosenfeld, C. & Kahn, A. (1983) Nature (Lond.), 305, 435-438. 29. Casnellie, .I.E., Harrison, M.L., Pike, L.J., Hellstrom, K.E. & Krebs, E.G. (1982) Proc. Natl. Acad. ScL USA, 79, 282-286. 30. Dasgupta, .I.D. & Garbers, D.L. (1983) .L Biol. Chem.. 258, 6174-6178. 31. Hunter, W.M. & Greenwood, F.C. (1962) Nature (Lond.), 194, 495-497. 32. Thom, D., Poweli, A.J., Lloyd, C.W. & Rees, D.A. (1977) Biochem. J., 168, 187-194. 33. Pilch, P.F. & Czech, M.P. (1979)J. Biol. Chem.. 254, 3375-3381. 34. Laemmli, U.K. (1970) Nature (Lond.), 227, 680-685. 35. Patschinsky, T., Hunter, T., Esch, F.S., Cooper, ,I.A. & Sefton, B.M. (1982) Proc. Natl. Acad. Sci. USA, 79, 973-977. 36. Czernilofsky, A.P., Levinson, A.D., Varmus, H.E., Bishop, ,I.M., Tischer, E. & Goodman, H. (1983) Nature (Lond.), 301, 736-738.

1183

37. Wong, T.W. & Goldberg, A.R. (1983) .L Biol. Chem., 258, I022-1025. 38. Puma, P., Buxser, S.E., Watson, L., Kelleher, D.,I. & Johnson, G.L. (1983) J. Biol. Chem.. 258, 3370-3375. 39. Moolenaar, W.H., Tsien, R.Y., van der Saag, P.T. & de Laat, S.W. (1983) Nature (Lond.), 304, 645-648. 40. Moolenaar, W.H. & de Laat, S.W. (1984) fll : Mechanisms of Receptor Regulation, (S.T. Crooke & G. Poste, eds.) Plenum Press, New-York, in press. 41. Goodman, R. & Herschman, H.R. (1982) J. Neurosci. Res., 7, 453-459. 42. Burstein, D.E. & Greene, L.A. (1982) Develop. Biol., 94, 477-482. 43. Revoltella, R.P. & Butler, R.H. (1980) J. Cell. Physiol.° 104, 27-33.