Upsala J Med Sci 82: 1-5, 1977 Cyclic 3’,5’-GMP Independent Protein Kinase at the Outer Surface of Intact Ehrlich Cells GUNNAR RONQUIST, GUNNAK .‘&REN, STEN EKLUND and CHRISTER WERNSTEDT From the Institute of Medical and P h y s i o l * v i c a l Chemistry, U p p s a l a , S w e d e n ABSTRACT The phosphoryl group transfer from ( Y ~ ~ P ) A T P into ac- ceptor proteins by an endogenous protein kinase a t the surface of Ehrlich cells has been further studied as regards possible stimulation by different concentrations of dibuturyl cyclic guanosine monophosphate (3’,5’-GMP). Using the endogenous acceptor protein of the surface of Ehrlich cells the cyclic nucleotide had no stimulatory effect on the protein kinase of the plasma membrane. The lack of stimulatory action of the cyclic nucleotide was also observed when an exogenous acceptor protein was present. Instead, a slight inhibitory effect was usually seen in both types of experi- ments. Labeled phosphorylserine was always in exces of labeled phosphoryl threonine. Both were isolated from hydrolyzed acceptor proteins. The lack of stimulation by a cyclic nucleotide on the phosphorylation of acceptor pro- tein@) on the cell surface does not rule out a regulatory function by the protein kinase of the plasma membrane. Instead, we propose an autoregulatory mechanism for the phosphorylation a t the cell surface. This mechanism is based upon the high sensitivity of the enzyme to Ca-ions. INTRODUCTION The presence of endogenous protein kinase at the surface of Ehrlich cells catalyzing the phosphoryla- tion of endogenous membrane protein(s) was first described 1970 (1,22) and in some following reports ( 2 , 4 , 23, 24). Schlatz & Marinetti described in 1971 the cyclic AMP-dependent phosphorylation of iso- lated rat liver plasma membrane by a partially purified protein kinase from rat liver (28). In a simi- lar experiment Kinzel & Mueller (13) were able to demonstrate the phosphorylation of endogenous acceptor proteins at the surface of intact HeLa cells in the presence of a protein kinase isolated from rat skeletal muscle and ( Y ~ ~ P ) A T P in the external medium. This reacting was also stimulated to a limited degree by cyclic AMP. Protein kinase- catalyzed membrane phosphorylations, either in- volving the endogenous protein kinase and an ex- trinsic acceptor protein or an extrinsic enzyme and the endogenous acceptor or where both enzyme and acceptor are endogenous, have subsequently been published for many membrane systems (for refer- ences, see ref. 4). Few of these protein kinase- mediated reactions have been cyclic AMP- independent ( 5 , 6, 10, 21, 31, 34) including the plasma membrane phosphorylation system for the Ehrlich cell (4, 23, 24). Many cyclic AMP-dependent protein kinases are, however, known to be activated by more than one 3’3’-cyclic nucleotide. Hence, cyclic GMP-specific kinases have in recent years been found in pig lung (16), guinea pig fetal lung (19), bovine adrenal glands (30), rat pancreas (33) and cerebellum (1 l ) , mammalian smooth muscle (7) and heart tissue (15, 17). Furthermore, it has been claimed that mem- brane-associated, cyclic GMP-dependent pro- tein kinases might mediate the effects of cholinergic agents in a manner quite analogous to the actions of particulate and soluble cyclic AMP-dependent pro- tein kinases in mediating cellular responses to polypeptide hormones, adrenergic agents and neurotransmitters (26). Cyclic GMP-stimulated protein kinase differs from that stimulated by cyclic AMP as regards some physicochemical properties (19, 20, 32). Neverthe- less, enzymatic properties for both types of protein kinases are the same as regards the requirements of binding steps at 3 different sites in the molecule, namely the binding of cyclic nucleotide for activa- tion, the binding of ATP and Mg2+ and the binding of protein substrate for the reaction (14). Smooth muscle is the only tissue in which any endogenous substrate proteins for the cyclic GMP- dependent protein kinases have so far been found (9), and it was of interest to extend the work on the I -772x54 Upsula J M r d Sci 82 2 Gunnar Ronquist et al. Table I. Effect of dibuturylcyclic G M P on the en- dogenous phosphorylation of the surface mem- brane of Ehrlich cells The figures denote the phosphoryl group transfer in nmol/lx lo8 cells per min at 37°C into endogenous acceptor protein SerP ThrP Peak 1 (a) Incubation of intact Ehrlich cells with exogenous (y”P) ATP and without cyclic nucleotide in the medium 0.190 0.040 0.003 ( b ) as ( a ) but in the presence of 1 X dibuturyl- cyclic GMP 0.130 0.040 0.003 endogenous protein kinase(s) as well as the en- dogenous acceptor proteins at the surface of Ehr- lich cells so as to encompass even a study of the possible cyclic GMP dependency. The present re- sults demonstrate the lack of influence of cyclic GMP in the concentration range 1 x M on the protein kinase system at the outer surface of Ehrlich cells. The steric dispositions of the regu- lator to the catalytic subunit have been investi- gated by means of dibuturyl cyclic GMP. The reac- tion was also studied with an exogenous phosphoryl group acceptor protein as well as with a purified membrane fraction. M- 1 x MATERIAL AND METHODS Materials. All chemicals were of analytical grade, (y3’P)ATP was purchased from NEN chemicals, GmbH, Frankfurt am Main, Germany. Unlabeled ATP (sodium form), cyclic and dibuturyl cyclic GMP as well as phosvi- tin were all obtained from Sigma Chemical Company, St. Louis, Mo., USA. Preparation of tumor cells. The Ehrlich ascites tu- mour cells were grown for 8-10 days in 5-week-old male Swiss albino mice obtained from the Anticimex Breeding Farm, Norrviken, Stockholm. The tumour cells were separated by centrifugation of the ascitic fluid, which had been diluted without delay 10-fold with ice-cold Krebs-Ringer bicarbonate medium, in order to diminish the tendency of cell agglutination. The cells were then washed twice in the Krebs-Ringer bicarbonate medium. Final washing was performed with a medium of 130 mM NaCl and 25 mM KCI for experiments with intact cells. For the preparation of a membrane fraction the cells were finally washed in a sucrose-(0.250 M)-tris-glycylglycine (0.016 M) buffer, pH 8.0, also containing 2 mM CaCI,, followed by centrifugation (25). The washings of intact cells were carried out in the International Refrigerated Centrifuge by acceleration for 10-15 sec up to at most Upsalu J M e d Sc; 82 5700 rpm followed by retardation for about 2 min. All preparatory steps were carried out at 4°C. Preparation of plasma membrane fraction. The proce- dure described by Ronquist & Christensen (25) was fol- lowed exactly. The final membrane fraction was suspended in 130 mM NaCl and 25 mM KCI in a concen- tration corresponding to about 20 mg membrane protein per ml. Incubation procedure. Incubations were performed di- rectly after the preparation of the cells and the membrane fraction. About 1 x lo8 cells or a membrane fraction cor- responding to 20 mg membrane protein were suspended in an incubation volume of 9.5 ml consisting of 130 mM NaCl and 25 mM KCI. It contained in addition 350 pmoles of Tris-acetic acid buffer, pH 7.5, 20 pmoles of MgCI,, 10 pmoles of NaZHPO, (unlabeled), 5 pmoles of glutathione (reduced form) and 0.5 pmoles of ( Y ~ ~ P ) A T P . In some experiments, 2.5 pmoles of phosvitin were also included. Dibuturylcyclic GMP, when present, was used in final concentratims of 1 x M- 1 x lo-’ M. The commercial phosvitin preparation was free from protein kinase activi- ty. Incubation was performed at 37°C and teminated after 1 min by rapid centrifugation. The entire centrifugation procedure did not exceed 30 sec. The supernatant repre- senting the external medium including the acceptor pro- tein in most experiments was immediately precipitated with trichloroacetic acid (10 % final concentration). In order to completely precipitate the low molecular weight acceptor protein, 50 mg of bovine albumin were added to the precipitating agent and the cell pellets were precipi- tated with trichloroacetic acid. The precipitation was al- lowed to proceed for at least 24 h at 4°C. The precipitates were washed once with 10% trichloroacetic acid con- taining 5 mM of unlabeled orthophosphate and 1 mM of ATP. I n order to remove nucleic acid and phozpholipids, the precipitate was treated according to Schneider (29). Par- tial hydrolysis of the Schneider protein was performed using the method of Lipmann (18). Labeled phos- phorylserine and phosphorylthreonine and phosphopep- tides were isolated by ion exchange chromatography, as described previously (23). A radiometric determina- tion of the different fractions from column chromato- graphy was carried out in a Nuclear Chicago Scintilla- tion Counter by measuring the Cerenkov radiation. RESULTS During the separation process two other (32P)- labelled phosphopeptides are recovered in addi- tion t o the (32P)-labelIed phosphorylserine and phosphorylthreonine. Those peptides have been called peak I and 2 , respectively ( 2 3 ) . Table I il- lustrates the lack of stimulation by IX M di- buturylcyclic GMP upon the endogenous mem- brane-associated protein kinase on phosphoryla- tion of the endogenous phosphoryl group acceptor in the surface membrane of intact Ehrlich cells. Instead a slight decrease of the degree of Cyclic 3',5'-GMP independent protein kinase 3 Table 11. Effect of various concentrations of di- buturylcyclic GMP on the phosphorylation of an exogenous acceptor protein by a protein kinase as- sociated with the surface membrane of Ehrlich cells The figures denote the phosphoryl group transfer in nmol/lx lo8 cells per min at 37°C into acceptor protein SerP ThrP Peak 1 ( a ) Phosvitin as exogenous phosphoryl group acceptor together with intact Ehrlich cells 0.580 0.225 0.103 ( b ) as ( a ) plus 1 x IO-'M 0.540 0.231 0.098 dibuturylcyclic GMP ( c ) as ( a ) plus 1 x M dibuturylcyclic GMP 0.473 0.197 0.076 (d) as ( a ) plus 1 x 10-4M dibuturylcyclic GMP 0.482 0.173 0.081 phosphorylation is seen in the presence of the cyclic nucleotide. This observation is in agreement with earlier findings with cyclic AMP (4, 23). The ali- phatic cyclic nucleotide is assumed to penetrate the plasma membrane. Thus, providing the cyclic GMP-binding subunit is in trans-position to the catalytic subunit, it would nevertheless be possible for the added cyclic nucleotide t o react. However, in experiments with intact cells the possibility of binding to the receptor site by endogenous cyclic nucleotide(s) cannot be ruled out. The lack of stimulation by different concentra- tions of dibuturyl cyclic GMP on the phosphoryla- tion of phosvitin as exogenous phosphoryl group acceptor in the presence of intact Ehrlich cells is shown in Table 11. Instead, a slight inhibitory ac- tion is again seen, especially at the two higher con- centrations. The incorporation ratio between phosphorylserine and phosphorylthreonine re- mained constant and about 2 . 5 , regardless of the absZnce or presence of different concentrations of dibuturylcyclic GMP in the incubation media. As mentioned before, the possibility of a reaction between cellular cyclic GMP and the receptor site of the regulatory subunit on the inner surface of the intact membrane could not be ruled out. In such a case additional cyclic GMP would have no or very little effect. Therefore, experiments were also performed on a purified membrane preparation mainly consisting of vesicles (25). It is evident from Table 111 that n o stimulatory effect is exerted by 1 x M dibuturylcyclic GMP in the medium also containing phosvitin and the membrane vesicles from Ehrlich cells. Instead, a small inhibitory effect is seen, consistent with the findings for the intact cells. The incorporation ratio between phosphoryl- serine and phosphorylthreonine was increased about ~~ 3 times compared with that for intact cells. DISCUSSION We propose that the protein kinase of the plasma membrane of Ehrlich cells is not dependent on either cyclic GMP in accordance with the present work or cyclic AMP (4, 23). Furthermore, at least some of this protein kinase is associated with the outer surface of the Ehrlich cell membrane. An external protein kinase activity in rat C-6 glioma cells was claimed in a recent report to be cyclic AMP-dependent (27). In the experimental system used, histone was the external phosphoryl group acceptor. N o data were presented concerning any phosphorylation of a possible endogenous ac- ceptor protein of the rat glioma cells, nor was any other acceptor protein used (27). We have also studied the protein kinase at the outer surface of human glioma cells (3), as well as of Ehrlich cells (24). In both types of cells we found a stimulation by cyclic AMP but only when histone was the ex- ogenous phosphoryl group acceptor. Furthermore, in the case of Ehrlich cells the maximal phosphory- lation with histone as the exogenous acceptor was 8 times lower than with phosvitin as the exogenous Table 111. Effect of dibuturylcyclic GMP on the phosphorylation of an exogenous acceptor protein by a protein kinase associated with a plasma membrane prepparpation from Ehrlich cells The figures denote the phosphoryl group transfer in nmol/mg Schneider protein of membrane vesicles per min at 37°C into acceptor protein. For comparison with Tables 1 and 2, 1 x lo8 Ehrlich cells correspond to about I5 mg Schneider protein from membrane vesicles prepared from Ehrlich cells ~ SerP ThrP Peak 1 Peak 2 ( a ) Phosvitin as exogenous phosphoryl group ac- ceptor together with membrane vesicles from Ehrlich cells 0.451 0.069 0.022 0.004 ( b ) as ( a ) plus 1 x 10-fiM dibuturylcyclic G M P in incubation medium 0.440 0.046 0.022 0.004 4 Gunnar Ronquist et al. acceptor under otherwise same experimental condi- tions (24). Also, the amount of phosphorylation of histone as the exogenous acceptor did not exceed that of the endogenous acceptor of the Ehrlich cells (24). The endogenous phosphorylation of Ehrlich cell membranes is not stimulated by cyclic AMP (23). We have therefore interpreted the effect of cyclic AMP with histone as the exogenous acceptor protein as secondary, e.g. interfering with the basic histone-protein interaction with the netto-negative charges of the intact cells (24). However, the ex- istence of at least two or more protein kinases at the Ehrlich cell surface with different properties vis-d- vis cyclic nucleotides cannot be excluded. The experiments with diffferent concentrations of cyclic GMP in the medium (Table 11) showed a very constant phosphorylserine to phosphoryl- threonine ratio throughout the experiments, thus favouring the view of homogeneity as regards pro- tein kinase activity in the surface membrane of Ehr- lich cells. The labelled phosphorylserine is always in excess of labelled phosphorylthreonine ( 2 , 4, 23, 24). This observation has been confirmed to be valid also for HeLa cells and SV3T3-ts cells in a recent work (12). A phosphorylation and dephosphorylation cycle of membrane proteins have been suggested to oc- cur, thereby changing their charge and conforma- tion (8). Furthermore, such changes might affect the interaction of the membrane proteins with the neighbouring lipids or lipoprotein complexes result- ing in the aggregation of intramembrane particles and relative increase in the exposed free lipid bilayer phase of the membrane (8). A cyclic nucleotide dependency of the protein kinase in the surface membrane does not appear to be obligatory. Instead we propose an autoregula- tory mechanism based upon the actual Ca’+- concentration in the microenvironment of the surface membrane. We found a strong inhibitory effect by Ca2+ on the endogenous protein kinase at the membrane surface of Ehrlich cells (23). Maxi- mal activity was obtained with MgZ+, Na+ and K + with no Ca2+ in the system. When half the amount of Mg2+ was stoichiometrically exchanged for Ca2+ the protein kinase activity decreased by about 50% and when all Mg2+ was replaced by Ca2+, more than 80% of the activity was lost. Therefore, in a maxi- mally dephosphorylated state of the surface membrane of the cell, the CaZ+ content of the microenvironment of the cell surface might be low and the protein kinase activity high, while in a maximally phosphorylated state Ca2+ ions are elec- trostatically attracted and therefore the protein kinase activity is lower. ACKNOWLEDGEMENTS This investigation was supported by a grant from the Swedish Medical Research Council (Project No B76- 13X- 00228- 12A). REFERENCES I . 2. 3. 4. 5. 6. 7. 8. 9. 10. 11 Agren, G . & Ronquist, G . : Isolation of 32P-labelled phosphorylserine from Ehrlich mouse-ascites tumour cells suspended in an isotonic medium containing 32P- labelled adenosine triphosphate. Acta Physiol Scand - Isolation of S2P-labelled phosphorylserine and phosphorylthreonine from Ehrlich mouse-ascites tumour cells, suspended in an isotonic medium con- taining 32P-labelled nucleoside triphosphates or inor- ganic pyrophosphates. Acta Chem Scand 25: - (32P)Phosphoryl transfer by endogenous protein kinase a t the glia and glioma cell surface in culture into extrinsic acceptor proteins. Acta Physiol Scand 92:430432, 1974. - Phosphorylation of endogenous membrane pro- teins by endogenous protein kinase at the outer surface of Ehrlich cells. Upsala J Med Sci 81: 129- 134, 1976. Andrew, C., Roses, A . D., Almon, R. R. & Appel, S. H.: Phosphorylation of membranes: Identification of a membrane-bound protein kinase. Science 182: Bacalao, J. & Rieber, M . : On the properties of membrane-associated protein kinase from Chinese hamster ovary cells. FEBS Lett37:37-41, 1973. Casnellie, J. E. & Greengard, P.: Guanosine 3’3’ monophosphate-dependent phosphorylation of en- dogenous substrate proteins in membranes of mam- malian smooth muscle. Proc Natl Acad Sci USA Gazitt, Y . , Ohad, I. & Loyter, A.: Phosphorylation and dephosphorylation of membrane proteins as a possible mechanism for structural rearrangement of membrane components. Biochim Biophys Acta Greengard, P.: Possible role for cyclic nucleotides and phosphorylated membrane proteins in postsynaptic actions of neurotransmitters. Nature 260: 101-108, 1976. Ho, R., Russell, T., Asakarva, T. & Snyder, R. P.: Influence on adipocyte plasma membrane bound pro- tein kinase by feedback regulator. J Cyc Nucleo Res 1:349-358, 1975. Hofman, F. & Sold, G . : A protein kinase activity from rat cerebellum stimulated by guanosine-3’S’-mono- phosphate. Biochem Biophys Res Commun 49: 1100-1107, 1972. 79: 125-128, 1970. 2931-2934, 1971. 927-929, 1973. 71: 1891-1895, 1974. 436: 1-14, 1976. Upsalu J Mrd S c i 82 12. 13. 14. 15. 16. 17. 18. 19. 20. 21 22 23 24 25 Kinzel, V., Kiibler, D., Mastro, A. M. & Rosengurt, E.: Phosphoproteins of the cell surface as generated by endogenous or exogenous protein kinase: stability of the 32P-labelled product. Biochim Biophys Acta Kinzel, V. & Mueller, G.: Phosphorylation of surface proteins of H e L a cells using an exogenous protein kinase and ( Y ~ ~ P ) ATP. Biochim Biophys Acta 322: 337-351, 1973. Kobayashi, R. & Fang, V.: Studies on cyclic GMP- dependent protein kinase properties by blue dextran- sepharose chromatography. Biochem Biophys Res Commun69: 1080-1087, 1976. Kuo, J. F., Davis, C. W. & Tse, J.: Depressed cardiac cyclic GMP-dependent protein kinase in spontane- ously hypertensive rats and its further depression by guanethidine. Nature 261: 335-336, 1976. Kuo, J. F . , Kuo, W-N., Shoji, M., Davis, C. W., Seery, V. L . & Donelly, T. E.: Purification and gen- eral properties of guanosine 3’3’-monophosphate- dependent protein kinase from guinea pig fetal lung. JBiol Chem251: 1759-1766, 1976. Kuo, J-F., Lee, T-P., Reyes, P. L . , Walton, K . G., Donnelly, T. E. & Greengard, P.: Cyclic nucle- otide-dependent protein kinases X . An assay method for the measurement of guanosine 3’ ,5’-monophos- phate in various biological materials and a study of agents regulating its levels in heart and brain. J Biol Chem247: 1 6 2 2 , 1972. Lipmann, F.: Uber die Bindung der Phosphosaure in Phosphoproteinen. I. Mitteilung: Isolierung einer phosphorhaltigen Aminosaure (Serinphosphosaure) aus Casein. Biochem Z262:3, 1933. Nakazava, K. & Sano, M.: Partial purification and properties of guanosine 3‘,5’-monophosphate-de- pendent protein kinase from pig lung. J Biol Chem Nishiyama, K., Katakami, H., Yamamura, H . , Takar, Y . , Shimomura, R. & Nishizuka, Y.: Func- tional specificity of guanosine 3’,5‘-monophos- phate-dependent and adenosine 3’,5’-monophos- phate-dependent protein kinases from silkworm. J Biol Chem2.50: 1297-1300, 1975. Prasad, K . N., Fogleman, M., Gaschler, M., Sinha, P. K . &Brown, L.: Cyclic nucleotide-dependent protein kinase activity in malignant and cyclic AMP-induced “differentiated” neuroblastoma cells in culture. Bio- chem Biophys Res Commun68: 1248-1254, 1976. Ronquist, G. & Agren, G.: Isolation of 32P-labelled phosphorylthreonine from Ehrlich mouse-ascites tumour cells suspended in an isotonic medium con- taining 32P-labeiled adenosine triphosphate. Acta Chem Scand 24: 728-729, 1970. - Isolation of 32P-labelled phosphorylserine and phosphorylthreonine from Ehrlich mouse-ascites tumour cells suspended in different isotonic media containing 32P-labelled adenosine triphosphate. Acta Chem Scand28: 1169-1174, 1974. - (32P)phosphoryl transfer by endogenous protein kinase at the Ehrlich cell surface into extrinsic ac- ceptor proteins. Upsala J Med Sci 79: 138-142, 1974. Ronquist, G. & Christensen, H.: Amino acid stimula- 434: 281-285, 1976. 250:7415-7419, 1975. Cyclic 3 ‘ 3 ‘-GMP independent protein kinase 5 26. 27. 28. 29. 30. 31. 32. 33. 34. tion of alkali-metal-independent ATP cleavage by an Ehrlich cell membrane preparation. Biochim Biophys Acta323: 337-341, 1973. Rubin, C. S. & Rosen, 0. M.: Protein phosphoryla- tion. Annu Rev Biochem44: 831-877, 1975. Schluger, E-J. & Kohler, G.: External cyclic AMP- dependent protein kinase activity in rat C-6 glioma cells. Nature 260: 705-707, 1976. Schlatz, L. & Marinetti, G . V.: Protein kinase mediated phosphorylation of the rat liver plasma membrane. Biochem Biophys Res Commun45: 51-56, I97 1. Schneider, W. C.: Phosphorous compounds in animal tissues. I . Extraction and estimation of desoxy- pentose nucleic acid. J Biol Chem 151:293-303, 1945. Shima, S . , Mitsunaga, M., Kawashima, Y., Taguchi, S. & Nakao, T.: Studies on cyclic nucleotides in the adrenal gland. 111. Properties of cyclic AMP- and GMP-dependent protein kinases in the adrenal gland. Biochim Biophys Acta341: 5 6 4 4 , 1974. Steiner, M.: Endogenous phosphorylation of platelet membrane proteins. Arch Biochem Biophys 171: 245-254, 1975. Takai, Y . , Nishiyama, K., Yamamura, H. & Nishizuka, Y .: Guanosine 3’,5’-monophosphate- dependent protein kinase from bovine cerebellum. Purification and characterization. J Biol Chem 250: 4690-4695, 1975. van Leemput, M., Camus, J . & Christophe, J.: Cyclic nucleotide-dependent protein kinases of the rat pancreas. Biochem Biophys Res Comrnun 54: 182- 190, 1973. Weller, M., Virmaux, N. & Mandel, P.: Light- stimulated phosphorylation of rhodopsin in the retina: The presence of a protein kinase that is specific for photobleached rhodopsin. Proc Natl Acad Sci USA 72:381-385, 1975. Received August 28, 1976 Address for reprints: Gunnar ¥, M.D. Institute of Physiological and Medical Chemistry Box 575 S-75123 Uppsala Sweden Upsulu J Med Sci 82