Phosphorylation of Synthetic Randompolypeptides by Protein Kinase P and Other Protein-Serine (Threonine) Kinases and Stimulation

Home , Protein kinase A

Proc. Nati. Acad. Sci. USA Vol. 85, pp. 1408-1411, March 1988 Biochemistry Phosphorylation of synthetic random polypeptides by protein kinase P and other protein-serine (threonine) kinases and stimulation or inhibition of kinase activities by microbial toxins [poly(threonine,glutamate)/polymyxin B/syringomycin/consensus sequences] M. ABDEL-GHANY*, D. RADEN*, E. RACKER*t, AND E. KATCHALSKI-KATZIRt *Section of Biochemistry, Molecular and Cell Biology, Division of Biological Sciences, Cornell University, Ithaca, NY 14853; and the tDepartment of Biophysics, The Weizmann Institute of Science, 76100 Rehovot, Israel Contributed by E. Racker, October 28, 1987

ABSTRACT A synthetic random polymer of threonine polymer that is not phosphorylated by three other protein- and glutamate (1:4.4) is readily phosphorylated by protein serine (threonine) kinases. In contrast, PK-P does not phos- kinase P but not by five other protein-serine (threonine) phorylate a serine/arginine polymer that is readily phospho- kinases. A synthetic random polymer of serine and arginine rylated by PK-A (the catalytic subunit of cAMP-dependent (1:3) is readily phosphorylated by protein kinase A and protein protein kinase) and PK-C. Syringomycin and polymyxin B, kinase C but not by protein kinase P. Although the amino acid two microbial toxins, stimulate PK-P and inhibit PK-C but sequences surrounding the phosphorylated serine (threonine) have no effect on PK-A. Thus, these toxins and synthetic residue have been demonstrated in studies with small synthetic polymers serve to distinguish between PK-P and other polypeptides to be decisive factors in the rate at which they are protein-serine kinases. phosphorylated, the findings with the large synthetic polypeptides suggest that in the case of proteins the size, the tertiary MATERIALS AND METHODS structure, and particularly the electrostatic interactions are Materials and methods were as described previously (5, 6). equally or more important contributing factors. Syringomy- The random polymers poly(Thr19Glu8") (1:4.4 threonine/glu- cin, a toxin from Pseudomonas syringae, and polymyxin B, tamate ratio), poly(Ser25Glu75), and poly(Ser25Lys75) were from Bacillus polymyxa, stimulate protein kinase P, strongly synthesized by previously described methods (7). Poly(Ser25- inhibit protein kinase C, and have no effect on protein kinase Arg75), poly(Lys50Tyr50), polymyxin B, histone type III-S A. Basic polypeptides with high lysine content are phospho- (histone 1), and histone type 11-AS were purchased from rylated by ATP nonenzymatically. Sigma; histone 2B was from Worthington. Syringomycin from Pseudomonas syringae was a gift from J. Takemoto (Utah State University, Logan, UT) and J. E. DeVay (University of A flood of new protein kinases have been described in recent California, Davis, CA). 10-(3-Aminopropyl)-2-chloropheno- years (1). It seems imperative to find ways to distinguish thiazine was supplied by Smith Kline & French. Casein between them, particularly for the purpose of dissecting the kinase II from yeast and skeletal muscle was prepared complex interactions that take place in intact cells and crude through the phosphocellulose column step as described (8, 9). extracts. Of particular interest are the kinases of oncogene PK-C was prepared as described (10). PK-A and the PK-A products and plasma membrane receptors involved in signal inhibitor were generously provided by E. Fischer (University transduction. Their functions are regulated by autophos- of Washington, Seattle), and cGMP-dependent protein kinase phorylation as well as by protein-serine (threonine) kinases. and calmodulin-dependent protein kinase II by H. C. Palfrey For example, the receptor that responds to epidermal growth (University of Chicago, Chicago). factor (EGF) catalyzes a protein-tyrosine kinase activity that Protein Kinase Assays. PK-P was assayed in a final volume is inhibited by phosphorylation of Ca2"- and phospholipid- of 50 Al containing 20 mM Na Hepes (pH 7.4), 5 mM MgCl2, dependent protein kinase C (PK-C) (2). Polypeptide- 10 mM thioglycerol, 5 ,g of casein (Sigma, catalogue no. dependent protein kinase P (PK-P) stimulates or inhibits, C4765) with or without 5 ,g of histone (Sigma type III-S), dependent on the extent of its activation, the EGF receptor PK-P (e.g., 2.5 ng), and 10 ,M [y-32P]ATP (=5000 kinase (3). To obtain better insight into these complex cpm/pmol). After 30 min at 23°C, samples (20 or 30 ,ul) were interactions it would be useful to have specific activators, placed on squares of Whatman 3 MM filter paper, washed, inhibitors, and substrates such as were available in the and dried for measurement of radioactivity by scintillation elucidation of other complex pathways (e.g., glycolysis and counting (6). For purpose of comparison, casein kinase II oxidative phosphorylation). and PK-A were assayed under the same conditions (not Synthetic random polymers containing tyrosine, gluta- optimal) except that 50 ,ug of casein was present in the casein mate, and other amino acids serve as specific substrates or kinase II assay. PK-C was assayed with histone III-S as inhibitors for tyrosine-specific protein kinases (4). Distinct substrate as described (10). differences, depending on amino acid composition, were observed in the interactions of these polymers with different RESULTS protein-tyrosine kinases. An exploration of synthetic ran- Phosphorylation of Synthetic Random Polypeptides by dom polymers that contain seine or threonine was under- PK-P. Poly(Thr19Glu81), a synthetic random polymer of taken in the hope that they may also yield some clues to the threonine and glutamate (1:4.4 molar ratio), was found to be substrate specificity of serine (threonine) kinases and may be a suitable substrate for PK-P from yeast. However, because useful in studies with crude extracts. In this paper it is shown that PK-P phosphorylates a synthetic threonine/glutamate Abbreviations: EGF, epidermal growth factor; PK-P, polypeptide- dependent protein kinase; PK-C, Ca2+- and phospholipid-dependent The publication costs of this article were defrayed in part by page charge protein kinase; PK-A, catalytic subunit of cAMP-dependent protein payment. This article must therefore be hereby marked "advertisement" kinase. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 1408 Downloaded by guest on September 25, 2021 Biochemistry: Abdel-Ghany et al. Proc. Natl. Acad. Sci. USA 85 (1988) 1409

Table 1. Phosphorylation of poly(Thr9CGlu81) and casein 2500 - Histone by PK-P Activity, nmol per min per mg ) 2000 - of protein E 1500- Substrate(s) - Histone + Histone c Casein (0.5 ug) 26 670 15000- Poly(Thr"9Glu81) (1.5 ,ug) 47 117 Poly(Ser25Arg75) Poly(Thr"9Glu81) + casein 38 154 <5001 Assays were performed for 30 min at room temperature as L-..O-I~~~~~~~~~~~~~~~~-m-a ,-- 25. .-.75% AN, Poly(Seff,'Lys"') described in Materials and Methods with 2.5 ng of PK-P and the n indicated amounts of substrates, in the absence or presence (5 yg) of 0 10 20 30 40 histone (Sigma type III-S). Time, min the acidic polymer interacted with the basic activators of FIG. 2. Time course of the phosphorylation of histone (Sigma PK-P, the polymer substrate was added at a concentration type III-S), poly(Ser25Arg75), and poly(Ser25Lys"5) by PK-A. Assays were performed in a final volume of 50 Ml with 80 ng of PK-A, 20 ,ug below its Km. Under these conditions the specific activity of of the substrate, and 10 MM [y-32P]ATP (5000 cpm/pmol). the enzyme was lower than with casein but markedly stimulated by histone (Table 1). In the presence of poly- Nonenzymatic Phosphorylation of Basic Polypeptides by (Thr'9Glu81) and low concentrations of casein the activity ATP. In the course of investigations on the specificity of the was considerably lower than with casein alone. The time activation of PK-P by histone III, we found that synthetic course of the reaction is shown in Fig. 1. Again the enzyme polymers like poly(Ser,Lys) or poly(Lys), which are not activity was lower with poly(Thr'9Glu8l) than with casein substrates for this enzyme, were activators of PK-P. How- but considerably higher than with poly(Ser25Glu75). ever, we observed in the filter paper assay as well as in Phosphorylation of Random Polypeptides by PK-A, PK-C, NaDodSO4/PAGE a nonenzymatic interaction between his- and Casein Kinase H. Poly(Ser25Arg75) served as a suitable tones or lysine-containing synthetic polymers and radio- substrate for PK-A (Fig. 2). Whereas low molecular weight active ATP. Poly(Lys5Tyreo) and poly(Lys) gave the high- synthetic substrates serve as substrates usually at millimolar est values, poly(Ser25Lys75) and histone less. In order to concentrations, the polymer was phosphorylated at about 10 evaluate the interaction between ATP and histones in greater gM. Poly(Ser25Lys75) was a poor substrate. A comparison of detail, various commercial histone preparations were incu- other protein-serine (threonine) kinases with the synthetic bated with radioactive ATP and then analyzed by NaDod- polymers is shown in Table 2. Neither PK-C nor yeast casein S04/PAGE and autoradiography (Fig. 3). Histones exposed kinase II phosphorylated poly(Thr19Glu81) significantly. to [y-32PIATP gave consistently much darker bands than PK-A, cGMP-dependent protein kinase, and calmodulin- those exposed to [a-32PJATP. It therefore seems unlikely dependent protein kinase II were also inactive (data not that the major incorporation of radioactivity was due to shown). On the other hand, poly(Ser25Arg75) was an excel- trapping of radioactive ATP, particularly since the samples lent substrate for PK-C but not for histone-activated PK-P. were electrophoresed in the presence of NaDodSO4. Thus, it Effect of Syringomycin and Polymyxin B and Other Inhib- seemed that a preferential transfer of the terminal phosphate itors on Protein-Serine (Threonine) Kinases. Another ap- from ATP had occurred. Consistent with this proposition proach to distinguish between protein-serine (threonine) was the observation that extensive washing of the filter kinases was to test the effect of toxins and other inhibitors paper in 2 M NaCl, which would be expected to dissociate (Table 3). Syringomycin, a phytotoxin produced by Pseu- ionic bonds, did not remove the radioactivity. Moreover, and B had effect exposure of histone to 32p; instead of ATP did not give rise domonas syringae (11, 12), polymyxin no to a radioactive precipitate on the filter paper. Phosphoryl- on PK-A, had very little effect on casein kinase II, stimu- ation of histones was favored at alkaline pH and was lated PK-P markedly in the absence of histone, and inhibited dependent on time and concentration of reagents (Fig. 4). PK-C. In the presence of histone, both syringomycin and Boiling the radioactive histone for 10 min in 1 M HCI polymyxin inhibited PK-P. PK-A inhibitor had no effect on eliminated the radioactivity from the protein. The radioac- PK-P or on the other protein kinases, whereas 10-(3- tive polymers were, however, more stable in 0.1 M HCO than aminopropyl)-2-chlorophenothiazine inhibited all but PK-A. Table 2. Phosphorylation of synthetic polymers by PK-P, casein 2500 kinase II (CK-II), and PK-C 2000 Poly(Thr9Glu81 ) Activity, nmol per min per mg of protein PK-P CK-II 1500 Substrate - H + H Yeast Muscle PK-C C.) 1000 Casein Casein 25 590 1.9 13.3 2.7 Poly(Ser25Glu75) Histone 0 0 0.1 0.6 66 500 Poly(Thr19Glu81) 15 119 0.3 NT 0.9 12 58 0.3 0.6 0.3 0 Poly(Ser25Glu75) 20 40 60 Poly(Ser25Arg75) 21 18 NT 0.4 27 Time, min Assays were as described in Materials and Methods with the following additions. For PK-P (2.5 ng): 5 ,ug of casein without or FIG. 1. Time course of PK-P activity. Assays were performed with S ,g of Sigma histone type III-S (H); S Mg of histone alone; or as described in Materials and Methods with 2.5 ng of PK-P, 5 ,ug of S Ag of poly(Thr19Glu8l), 2.5 Ag of poly(Ser25Glu75), or 5 Mg of casein, 1.5 Mg of poly(Thr'9Glu8"), or 2.5 Mug of poly(Ser25Glu75), in poly(Ser25Arg75) without or with 20 Ag of histone. For CK-II (200 the absence or presence of histone (Sigma type III-S). Values in the ng): 50 Ag of casein without or with 5 ,g of histone; 5 ,g of synthetic absence of histone III were subtracted from the values in its polymer. For PK-C (100 ng): 5 Mg of casein, 5 ,ug of histone, or 5 Mg presence. Note that values for casein have been divided by 10. of synthetic polymers. NT, not tested. Downloaded by guest on September 25, 2021 1410 Biochemistry: Abdel-Ghany et al. Proc. Natl. Acad. Sci. USA 85 (1988) Table 3. Effect of inhibitors on protein kinases Activity, nmol per min per mg of protein PK-P CK-II Inhibitor - H + H PK-A Yeast Muscle PK-C None 6 390 28 1.6 13 130 Syringomycin (10 ,ug) 95 100 33 1.4 12 5 PK-A inhibitor (100 units) 8.4 390 2.5 1.7 12 130 10-(3-Aminopropyl)-2-chlorophenothiazine (180 tLM) 60 180 26 0.7 7 5.6 Polymyxin B 10 ,g 77 400 39 NT 12 18 40 Ig 86 300 23 NT 13 9 Assay mixtures with PK-P (2.5 ng), PK-A (80 ng), casein kinase II (CK-II) from yeast (200 ng) or from skeletal muscle (400 ng), and PK-C (25 ng) were incubated at 230C for 30 min. PK-P was assayed in the absence or presence (5 ,ug) of Sigma histone type III (H). Inhibitors were added immediately prior to assay. Histone (5 ug) was used as substrate for PK-A and PK-C, casein (5 Ag) for PK-P, and casein (50 Ag) for CK-II. NT, not tested. in 0.1 M NaOH when heated to 80°C for 10 min. Attempts to magnitude lower than those of the small peptides. This reverse the reaction with ATP, glucose, or glycerol as suggests that in the case of proteins the size, tertiary acceptors have thus far failed. structure, and particularly the electrostatic interactions with the enzyme (13, 14) may be equally or more important for DISCUSSION catalytic activity than the so-called ''consensus sequences." There are several advantages in the use of synthetic random By synthesis of large polypeptides with different amino acid polypeptides in the studies of protein kinases. One, particu- mixtures and different charge distributions, it may be possi- larly obvious in the case of protein-tyrosine kinases, is that ble to evaluate quantitatively the contributing components. enzyme assays can be performed in the presence of excess PK-P has some properties in common with yeast casein protein-serine kinases by a rapid filter-paper assay. As kinase II (8, 9), such as sensitivity to acidic polysaccharides pointed out previously (4), small polypeptides that have been and activation by basic compounds. However, whereas custom-synthesized according to the sequence around the casein kinase II was isolated from a soluble yeast extract and phosphorylation site of tyrosine in natural substrates have did not require a detergent for stability, we isolated PK-P by much higher Km values than the synthetic, large polypep- extraction of yeast membranes with Triton X-100 and tides. Of particular interest are selective differences in the showed that the detergent was required for stability of the choice of optimal substrates observed with various protein- enzyme (6). Furthermore, phosphorylation of casein by tyrosine kinases. For example, a tyrosine/glutamate poly- purified yeast casein kinase II was not stimulated by histone mer was an excellent substrate for the placental insulin (Sigma type III-S) under our conditions of assay (Table 2), receptor, whereas a tyrosine/alanine/glutamate polymer whereas PK-P exhibited almost complete dependency. The was not. For the EGF receptor the tyrosine/alanine/gluta- enzymes also showed different substrate specificity. Poly- mate polymer was a better substrate than poly(Tyro,Glu). In (Thr'9Glu8") was rapidly phosphorylated by PK-P but not by the case of serine (threonine) protein kinases, caseins or casein kinase II. Finally, PK-P and PK-C were influenced by histones have served as excellent traditional substrates. The syringomycin, whereas casein kinase II was not. In the synthetic random polypeptides described here are not only absence of histone, syringomycin, as well as polymyxin, a convenient substrates but have Km values that are orders of known inhibitor of PK-C (15), markedly stimulated the

Mr 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 x 1- 66.0- - 45.0- ^ 36.0- 4 4wh 29.0- _W f.. 24.0- 0 20.1- 5

.I 14.2-S

FIG. 3. Phosphorylation of histones by ATP. Various histones were incubated for 20 min at room temperature in 20 mM Tricine (pH 9.3) containing 200 ,uM [y-32PJATP (177 cpm/pmol) or (a-32P]ATP (218 cpm/pmol). The_phosphorylated histones were subjected to NaDodSO4/PAGE and stained with Coomassie blue (lanes 1-8), and the autoradiogram (lanes 9-16) was scanned with a Bio-Rad 620 scanning densitometer. The ratio of the densitometer reading for [y-32P]ATP vs. [a-32P]ATP was 3.8 for histone 11-AS (Sigma), 5.3 for histone 2B (Worthington), and 1.7 for histone 111-S (Sigma). Lanes 1 and 9, [a-32P]ATP; lanes 2 and 10, [y-32P]ATP; lanes 3 and 11, histone 11-AS plus [a-32P]ATP; lanes 4 and 12, histone 11-AS plus [y-32P]ATP; lanes 5 and 13, histone 111-S plus [y-32P]ATP; lanes 6 and 14, histone 111-S plus [y-32P]ATP; lanes 7 and 15, histone 2B plus [a-32P]ATP; lanes 8 and 16, histone 2B plus [y-32P]ATP. Lane at left contained standard proteins used as molecular weight markers. Downloaded by guest on September 25, 2021 Biochemistry: Abdel-Ghany et al. Proc. Natl. Acad. Sci. USA 85 (1988) 1411 A B C 200 - 200 l

E 100 100- 0. CY, .I o . I . I - . 0 400 800 0 20 40 0 20 40 ATP, ,uM Poly(Lys), /g Time, min

FIG. 4. Effect of ATP concentration, poly(Lys) concentration, and time on phosphorylation of poly(Lys) by ATP. Assay mixtures contained 20 mM Tricine (pH 9.3). (A) At various ATP concentrations, with 25 Ag of poly(Lys) for 30 min at room temperature. (B) At various concentrations of poly(Lys) for 30 min at room temperature in the presence of 200 .uM [y-32PATP (1000 cpm/pmol). (C) For various times, with 25 jtg of poly(Lys) and 200 ,uM [y-32P]ATP.

phosphorylation of casein by PK-P. The use of synthetic 1. Hunter, T. (1987) Cell 50, 823-829. polypeptides combined with inhibitor studies should allow 2. Cochet, G., Gill, G. N., Meisenhelder, J., Cooper, J. A. & us to differentiate between different protein kinases and Hunter, T. (1984) J. Biol. Chem. 259, 2553-2558. 3. Abdel-Ghany, M., Kole, H. K. & Racker, E. (1987) Proc. assess a possible role of PK-P in the phosphorylation of Natl. Acad. Sci. USA 84, 8888-8892. several membrane components-e.g., in Xenopus laevis 4. Braun, S., Raymond, W. E. & Racker, E. (1984) J. Biol. oocytes (16), in clathrin-coated vesicles (17, 18), and of LDL Chem. 259, 2051-2054. receptor. In the latter case we established, together with J. 5. Racker, E., Abdel-Ghany, M., Sherrill, K., Riegler, C. & Goldstein and collaborators, that PK-P, which phosphoryl- Blair, E. A. (1984) Proc. Natl. Acad. Sci. USA 81, 4250- ates LDL receptors, is not identical to the enzyme from 4254. bovine adrenal cortex (19). 6. Yanagita, Y., Abdel-Ghany, M., Raden, D., Nelson, N. & Racker, E. (1987) Proc. Natl. Acad. Sci. USA 84, 925- The phosphorylation by protein kinases of synthetic poly- 929. peptides that contain only two amino acids also raises the 7. Sela, M., Fuchs, S. & Arnon, R. (1962) Biochem. J. 85, question of the significance of both in vitro and in vivo 223-235. phosphorylation of proteins. Emphasis must therefore be 8. Meggio, F., Grankowski, N., Kudlicki, W., Szyszka, R., placed on induced changes in function. We have recently Gasior, E. & Pinna, L. A. (1986) Eur. J. Biochem. 159, 31-38. observed that PK-P alters the activity of several protein- 9. DePaoli-Roach, A. A., Ahmad, Z. & Roach, P. J. (1981) J. tyrosine kinases. Of particular interest is the observation that Biol. Chem. 256, 8955-8962. and activators of PK-P of 10. Woodgett, J. R. & Hunter, T. (1987) J. Biol. Chem. 262, polylysine, polymyxin, histone, 4836-4843. different strength, influence the protein-tyrosine kinase activ- 11. Surico, G. & DeVay, J. E. (1982) Physiol. Plant Pathol. 21, ity of the solubilized EGF receptor in a graded manner (3). 39-53. The observed nonenzymatic phosphorylation of histones 12. Bidwai, A. P., Zhang, L., Bachmann, R. C. & Takemoto, by ATP needs further exploration. The observation that the J. Y. (1987) Plant Physiol. 83, 39-43. terminal phosphate of ATP is preferentially transferred to 13. Wells, J. A., Powers, D. B., Bott, R. R., Graycar, T. P. & the peptide is important for the elimination of a noncovalent Estell, D. A. (1987) Proc. Natl. Acad. Sci. USA 84, 1219-1223. interaction. A nonenzymatic interaction between histones 14. Russell, A. J. & Fersht, A. R. (1987) Nature (London) 328, 496-500. and ADP-ribose or ribose 5-phosphate was demonstrated 15. Wise, B. C., Glass, D. B., Chou, C.-H. J., Raynor, R. L., previously (20). However, the observed lability of the phos- Katoh, N., Schatzman, R. S., Turner, R. S., Kibler, R. F. & phoryl group in the polymer to 0.1 M NaOH and greater Kuo, J. F. (1982) J. Biol. Chem. 257, 8489-8495. stability to 0.1 M HCI are not consistent with the properties 16. Gatica, M., Allende, C. C., Antonelli, M. & Allende, J. E. described for phosphorylated lysine (21) and more consistent (1987) Proc. Natl. Acad. Sci. USA 84, 324-328. with a phosphorylated serine residue. 17. Schook, W. J. & Puszkin, S. (1985) Proc. Natl. Acad. Sci. One should consider the possibility that nonenzymatic USA 82, 8039-8043. phosphorylation of histones by ATP may take place in vivo 18. Usami, M., Takahashi, A., Kadota, T. & Kadota, K. (1985) J. in view of the high intracellular ATP concentration. Biochem. (Tokyo) 97, 1819-1822. 19. Kishimoto, A., Brown, M. S., Slaughter, C. A. & Goldstein, We wish to thank Benjamin Hilton for excellent technical assis- J. L. (1987) J. Biol. Chem. 262, 1344-1351. tance in the preparation of the kinases and Leon Heppel and Ross 20. Kun, E., Chang, A. C. Y., Sharma, M. L., Ferro, A. M. & Resnick for critical review of this paper. This work was supported Nitecki, D. (1976) Proc. Natl. Acad. Sci. USA 73, 3131-3135. by Public Health Service Grant CA08964 from the National Cancer 21. Smith, D. L., Chen, C.-C., Bruegger, B. B. & Holtz, S. L. Institute. (1974) Biochemistry 13, 3780-3785. Downloaded by guest on September 25, 2021