Studies on the Phosphorylation of Myelin Basic Protein by Protein ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY @ 1985 by The American Society of Biological Chemists, Inc.

Vol. 260, No. 23, Issue of October 15, pp. 12492-12499,1985 Printed in U.S.A.

Studies on the Phosphorylationof Myelin Basic Protein by Protein Kinase C and Adenosine 3’:5’-Monophosphate-dependent Protein Kinase* (Received for publication, March 22,1985)

Akira Kishimoto, Kaoru NishiyamaS, Hiroyuki Nakanishi, Yuko Uratsujig,Hideaki Nomura, Yoshifumi TakeyamalI,and.Yasutomi Nishizuka From the Department of Biochemistry, Kobe University School of Medicine, Kobe 650, Japan

The substrate specificity of .protein kinase C was studied and compared with that of cyclic AMP-dependent protein kinase (protein kinase A) by using bovine brain myelin basic protein as a model substrate. This basic protein was phosphorylated at multiple sites by both of these protein kinases. In this analysis, the basic protein was thoroughly phosphorylated in vitro with [y3’P]ATP and each protein kinase, and then digested with trypsin. The resulting radioactive phosphopeptides were isolated,by gel filtration followed by high performance liquid chromatography on a reversephase column. Subsequent amino acid analysis and/or sequential Edman degradation of the purified phosphopeptides, together with the known primary sequence of this protein, revealed that Ser-46and Ser-151 were specifically phosphorylated by protein kinase C, whereas Thr-34 and Ser-115 were phosphorylated preferentially by protein kinase A. Both kinases reacted with Ser-8, Ser-11, Ser-55, Ser-110, Ser-132, and Ser-161 at various reaction velocities. Contrary to protein kinase A, protein kinase C appears to react preferentially with seryl residues that are located at the amino-terminal side close to lysine or arginine. The seryl residues that are phosphorylated commonly by these two protein kinases have basic amino acids .at both the amino- and carboxyl-terminal sides. These results provide some cluesto understanding the rationale that these kinases may show different but sometimes similar functions depending on the structure of target phosphate acceptor proteins.

Protein kinase C has recently attracted great attention in the studies on the activation of cellular functions and prolif-

* This investigation was supported in partby research grants from the Scientific Research Fund of the Ministry of Education, Science and Culture, Japan (1982-1984), the Intractable Disease Division, Public Health Bureau, the Ministry of Health and Welfare, Japan (1982-1984), a grant-in-aid ofNew Drug Development from the Ministry of Healthand Welfare, Japan (1983), the Science and Technology Agency(1983), and the Yamanouchi Foundation for Research on Metabolic Disorders (1982-1983). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ On leave from Kobe University School of Allied Medical Sciences, Kobe 654, Japan. On leave from the Department of Anesthesiology, Kobe University School of Medicine, Kobe 650, Japan. ll On leave from the Department of Surgery (1st Division), Kobe University School of Medicine, Kobe 650, Japan.

eration. Under physiological conditions the enzyme is activated synergistically by diacylglycerol and Ca2+in the presence of membrane phospholipid (1).The diacylglycerol active in this capacity may arise in the plasma membrane only transiently from the receptor-mediated hydrolysis of inositol phospholipids, and this protein kinase appears to play roles of crucial importance in transmitting various extracellular informational signals across the cell membrane (for reviews, see Refs. 2, 3). The signals relating to protein kinase C normally differ from those from a group of hormones and some neurotransmitters that produce cyclic AMP asan intracellular messenger (for reviews, see Refs. 4-6). Thus, it is likely that protein kinase C and protein kinase A’ transduce distinctly different pieces of information into thecell through their own specific protein phosphorylation. However, these two signal pathways sometimes cause apparently similar cellular responses but often potentiate each other, and the potential interaction of these pathways at either receptor levels or target proteins is becoming important for understanding the basal mechanism of signal transduction (3). To assess one of the basic problems, the substrate specificity of protein kinase C, the present studies were undertaken to identify the sites of MBP phosphorylated by this protein kinase, and to compare them with the sites phosphorylated by protein kinase A. Although the physiological significance of this protein phosphorylation is not clear, MBP appears to be an ideal model substrate, since this basic protein is phosphorylated at multiple sites by both protein kinase C (7-9) and protein kinase A (10-E),and its primary structure has been clarified (13).The results obtained seem to indicate that these two protein kinases definitely differ from each other in their substrate recognition, but are able to phosphorylate the same site in anappropriate amino acid sequence. EXPERIMENTALPROCEDURES

Enzymes and Substrate-Protein kinase C was purified from rat brain as described by Kikkawa et al. (14). The catalytic subunit of protein kinase A was purified from rabbit skeletal muscle as described by Bechtel et al. (15). Both protein kinases were homogeneous as judged by polyacrylamide gel electrophoresis. MBP was isolated from bovine brain by the method of Oshiro and Eylar (16). The original numbering system for the amino acid sequence (13) is used throughout the present studiesin accordance with the earlier reports of phosphorylation sites (10,17),although this sequence has been later modified (18). Calfthymus H1 histone was prepared as described (19). Other Chemicals and Materials-Trypsin treated with L-l-tosylamido-2-phenylethyl chloromethyl ketone was obtained from Worthington. [y3‘P]ATP was prepared as described by Glynn and Chappell The abbreviations used are: protein kinase A, cyclic AMP-dependent protein kinase; MBP, myelin basic protein; PTH, phenylthiohydantoin; HPLC, high performance liquid chromatography.

12492

12493

Substrate Specificity of Protein Kinase C TABLE I Relative uelocitiesof phsphorylntwn of MBP andHI histone by protein kinases A and C Protein kinase A (0.05 pg) or protein kinase C (0.04 pg) wasassayed under the standard conditions except that either 200 pg/ml MBP or 200pg/ml H1 histone was employed as substrate. Incubation was carried out for 3 min a t 30 "C with or without Ca2+,phospholipid vesicle, and diolein for protein kinase C, and with or without protein inhibitor for protein kinase A. Protein kinase C

Substrate

-a

Protein kinase A

+'

-b

f b

pmolphosphate/min/mg enzyme protein

Bi

0.04 1.4 0.5 0.34 MBP 0.22 2.9 1.0 0.16 H1 histone In the absence and presence, respectively, of Ca2+,phospholipid, and diolein. In theabsence and presence, respectively, of protein inhibitor.

4 h

!2

30

$ 3

50

70

FRACTION WUIBER FIG.2. Gel filtration of radioactive tryptic phosphopeptides

\

s!

derived from MBP phosphorylated by protein kinasesA and C. The phosphopeptides were chromatographed on a Sephadex G-15 column (118 X 1 cm) equilibrated with 0.1% (v/v) trifluoracetic acid. Elution was performed with the same solution a t a flow rate of 3.6 ml/h. Fractions of 0.8 ml each were collected, and the radioactivity was measured. Fractions (I, ZI, 111, and IV as indicated by the horizontal bars) were separately pooled. A, with protein kinase C; B, with protein kinase A.

U

0

60

120

240

180

INCUBATION TIME

(

MIN 1

FIG.1. Phosphorylation of MBP catalyzed by protein kinases A and C. Either protein kinase A (1.4 pg) or protein kinaseC (0.27 pg)was incubated at 30 "cin a final volume of 0.5 ml of 20 mM Tris/HCl at pH 7.5, containing 30 mM 2-mercaptoethanol, 5 mM magnesium acetate, 0.15 mM [y3'P] ATP (2.5 X lo6 cpm), and 0.5 mg/ml MBP. The reaction of protein kinase C was started by the addition of a solution containing the final concentrations of 0.1 mM CaC12, 8 pg/ml phospholipid vesicle, and 0.4 pg/ml diolein to the above mixture. A 2 0 4 aliquot was removed at each time indicated, and the radioactivity of acid-precipitable materials was measured. Where indicated by an arrow, fresh protein kinase C (0.27 pg, 50 pl) containing 0.1 mMCaCl', 8 pg/ml phospholipid vesicle, and 0.4 pg/ ml diolein was added to thereaction mixture. 0 and 0, phosphorylaphosphorylation by tion by protein kinases C and A, respectively; 0, protein kinase A followed by the addition of protein kinase C. (20). Erythrocyte membrane phospholipid was prepared as described previously (21). Protein inhibitor of protein kinase A was prepared from rabbit skeletal muscle as described by Walsh et al. (22). 1,2Diolein, trifluoroacetic acid (spectrograde), and Sephadex G-15 were obtained from Serdary Research Laboratories, E. Merck, and Pharmacia Fine Chemicals, respectively. Reagents and solvents used for the amino acid analysis and automated sequence analysis were pur-

chased from LKB. Spectrapor 3 membrane tube (M? cutoff, 3,500) was obtained from Spectrum Medical Industries, Inc. PTH-amino acid derivatives were purchased from Wako Chemicals. Protein Kinase Assay-Protein kinase C was routinely assayed by measuring the incorporation of 3zPof [y3'P]ATP into H1histone in the presence of Ca2+,phospholipid, and diolein as described earlier (21). The catalytic subunit of protein kinase A was assayed with H1 histone as substrate by the method as described previously (23). High Performance Liquid Chromatography-HPLC was carried out using a Waters HPLC system consisting of one 6000A pump, one M45 pump, a 660 programmer, and a U6K universal injector. Column eluates were monitored for the absorbance a t 206 nm using a Shimadzu spectrophotometric detector, Model SPD-SA, equipped with an 8-pl cell with 1-cm light path length. All separations were performed a t room temperature on a TSK-gel ODs-12OT reverse-phase column (5 pm silica, 120-A pore, 250 X 4.6 mm; Toyo Soda Japan). A 25-p1 aliquot of phosphopeptides dissolved in 0.1% (v/v) trifluoracetic acid was injected onto the column equilibrated with the same acid. After the column was washed with 0.1% trifluoracetic acid for 12 min a t a flow rate of 1 ml/min, elution was performed using an 80-min linear gradient from 0.1% (v/v) trifluoracetic acid to 20% (v/ v) acetonitrile in0.1% (v/v) trifluoracetic acid at a combined constant flow rate of 1ml/min with column pressure between 1,500 and 2,500 p.s.i. A d y s i s Of Phosphoserine and Phosphothreonine-The radioactive MBP and phosphopeptides were dissolved in 6 N HCI, and heated for 2 h at 110 "C in a sealed glass tube. After the HCl was removed under reduced pressure, the hydrolysates were analyzed by electrophoresis on a cellulose-coated thin layer plate (24). Under the condition employed, phosphoserine, phosphothreonine, and phosphotyrosine were separated from one another. Alternatively, phosphoamino acids were separated by column chromatography, and theradioactivity was quantitated (25).

12494

Substrate Specificityof Protein KinaseC

FIG. 3. Separation of radioactive tryptic phosphopeptides on HPLC.

A 25.~1 portionof each labeled phosphopeptide fraction shown in Fig. 2 was injected onto a reverse-phase column as described under "Experimental Procedures." Fractions of 1 ml each were collected and the radioactivity was determined. The arrow shows the start of a linear gradient of acetonitrile. A and F , unfractionated tryptic phosphopeptides of MBP thatwas phosphorylated by protein kinases C and A, respectively; B, C, D, E, G, H , I, and J,phosphopeptides in Fractions IC, IIc, IIIc, IVc, Ia, IIa, IIIa, and IVa, respectively.

RETENTION TlME ( MIN 1 Amino AcidAnalysis-The radioactive phosphopeptides (5-20 mol of radioactive phosphate from [r-32P]ATP intoeach mol nmol each) were hydrolyzed in 5.7 N HC1 containing 0.1% (v/v) of MBP, respectively. (Fig. 1) Subsequent incorporation of phenol for 22 h at 110 "C in vacuo, and analyzed with an LKB amino acid analyzer, Model 415001,according to themanufacturer's program radioactive phosphate was extremely slow, and the enzyme did not appear to be limited. Although both kinases were for protein hydrolysate analysis. Sequentiel Edman Degradation-The amino acid sequence of some autophosphorylated (14, 30), practically no acid-precipitable radioactive peptides were determined with an LKB solid phase se- radioactivity was detected with the enzyme alone under the quencer, Model4020,by the method of Laursen (26) using the conditions employed. Acidhydrolysis of the radioactive MBP manufacturer's program. Before starting an analysis, phosphopep- revealed that protein kinase C phosphorylated exclusively tides having lysine at the C-terminal were covalently attached to pphenylenediisothiocyanate glass (Pierce No. 22350) as described by seryl residue, and the amount of phosphothreonine obtained Potter and Taylor (27), and phosphopeptides having arginine at the was less than 5% of that of phosphoserine. On the other hand, C-terminal were covalently attached to aminopropyl glass (Pierce protein kinase A phosphorylated both seryl (72%) and No. 23538) as described by Machleidt et al. (28). The resulting PTH threonyl (28%) residues. derivatives were analyzed by the HPLC system (29). Phosphopeptides from Radioactive MBP-To identify the Preparation of Lipid Vesicles-Phospholipid and diolein were seryl and threonylresidues phosphorylated by protein kinases stored separately in chloroform a t -20 "C. Prior to use lipids were C and A, MBP (10 mg; 0.6 pmol) was incubated with [ Y - ~ ~ P ] mixed first in chloroform, dried, and then suspended in 20 mM Tris/ HC1, pH 7.5, by sonication for 5 min a t 0 "C using a Kontes sonifier, ATP and either protein kinase A (56 pg) or protein kinase C (11 pg) in a 20-ml mixture under the conditions similar to Model K881440. Determinations-The radioactivity of 32Pwas determined using a those described in the legend to Fig. 1. The reaction was Packard Tri-Carb liquid scintillation spectrometer, Model 3330. Pro- stopped by the addition of an ice-cold saturated ammonium tein was determined as described previously (21). Polyacrylamide gel sulfate solution (16.4 ml,final 45% saturation). After standing electrophoresis was performed under the condition described by for 20 min in ice, the precipitated MBP was collected by Oshiro and Eylar (16) except that a slab-gel system was employed. After the electrophoresis and protein staining, the gel was dried on a centrifugation for 15 min at 20,000 x g at 4 "C, and washed with 20 ml of an ice-cold 45% saturated ammonium sulfate filter paper in vacuo for autoradiography. solution followed by centrifugation. These procedures miniRESULTS mized the enzyme protein which contaminated the radioactive Phosphorylation of MBP-Table I shows the relative rates MBP preparations.The precipitated MBP was dissolvedin 3 of phosphorylation of MBP and H1 histone catalyzed by ml of distilled water, and dialyzed overnight at 4 "C with a protein kinases A and C. For protein kinase C, MBP was a Spectrapor 3 membrane tube against a large volume of disbetter substrate than H1 histone, and the reaction required tilled water. The radioactive MBP thusprepared was electroCa", diolein, and phospholipid. The optimum condition was phoretically homogeneous, and all radioactivity co-migrated obtained in the presence of 5 mM magnesium acetate, 8 pg/ with MBP as judged by autoradiography. The recovery of ml phospholipid, 0.4 pg/ml diolein, and 10 p~ CaClz at pH MBP was 7.5 mg (0.45 pmol), and the radioactive phosphate 7.5. Protein kinase A also phosphorylated both H1 histone in amounts of 1.8 pmol and 1.3 pmol was associated with the and MBP, but H1 histone was a better substrate for this MBP that was incubated with protein kinases C and A, enzyme. K, values for MBP of protein kinases A and C were respectively. The radioactive MBP was freeze-dried, dissolved in 2 ml of M 160 pg/ml and 50 pg/ml, and those for ATP were 7 X 0.1 M ammonium bicarbonate at pH 8.0, and digested with and 1 X M, respectively. After prolonged incubation under optimum conditions, pro- trypsin (0.2mg, treated with L-1-tosylamido-2-phenylethyl tein kinases A and C incorporated approximately 2.9 and 4.0 chloromethyl ketone) for 2.5 h at 37 "C. Fresh trypsin (0.2

Substrate Specificity of Protein Kinase C

"

S

8

12495

Substrate Specificity

12496

TABLE III Amino acid sequences of “P-labeled tryptic pi2osphopeptides The amino acid sequence of each phosphopeptide is predicted from the data of amino acid analysis shown in Table 11. The number indicates the sequence from the amino-terminal end (13).

C

ofKinase Protein

4 100

Amino acid sequence

Peptide

r .

11 Ser-Lys 53 Arg-Gly-Ser-Gly-Lys 6 Arg-Pro-Ser-Gln-Arg 131 Ala-Ser-Asp-Tyr-Lys 49

be

30

Fl c

10 y

0

z

3

q T

Li

Gly-Ala-Pro-Lys-Arg-Gly-Ser-Gly-Lys-Asp-

U

Gly-His-His-Ala-Ala-Arg 156

4:

0

100

Leu-Gly-Gly-Arg-Asp-Ser-Arg

e

143

Gly-His-Asp-Ala-Gln-Gly-Thr-Leu-Ser-Lys

30

65

Thr-Thr-His-Tyr-Gly-Ser-Leu-Pro-Gln-Lys 65

10

Thr-Thr-His-Tyr-Gly-Ser-Leu-Pro-Gln-Lys 108 Gly-Leu-Ser-Leu-Ser-Arg 106

Gly-Arg-Gly-Leu-Ser-Leu-Ser-Arg 43 Phe-Phe-Gly-Ser-Asp-Arg 13

Tyr-Leu-Ala-Ser-Ala-Ser-Thr-Met-AspHis-Ala-Arg 31

His-Arg-Asp-Thr-Gly-Ile-Leu-Asp-Ser-LeuGly-Arg 114 p15

Phe-Ser-Trp-Gly-Ala-Glu-Gly-Gln-LysPro-Gly-Phe-Gly-Tyr-Gly-Gly-Arg

mg) was again added, and the mixture was incubated for an additional 2 h. Then, ammonium carbonate was removed by extensivelyophilization. The resulting radioactive tryptic phosphopeptides (1.7 pmol and 1.1 pmol of 32Pfor protein kinases C and A, respectively) were taken up with 0.1 ml of 0.1% (v/v) trifluoracetic acid, and subjected to gel filtration on a Sephadex G-15 column under the conditions specified in the legend to Fig. 2. The totalrecovery of radioactive materials was about 80%. Peak I (fractions 33-38; 0.4 pmol and 0.24 pmol of 32Pfor protein kinases C and A, respectively), peak I1 (fractions 39-47; 0.77 pmol and 0.44 wmol of 32Pfor protein kinases C and A, respectively), peak111(fractions 48-53; 0.19 pmol and 0.04 pmol of 32Pfor protein kinases C and A, respectively), and peak IV (fractions 59-67; 0.08 pmol of 32P for both protein kinases) were separately collected. These fractions were designated hereafter as la, IIa, IXIa, and IVa for the preparations obtained with protein kinase A, and IC, IIc, IIIc, and IVc for the preparations obtained with protein kinase C. These phosphopeptides were lyophilized and taken up with each0.1 ml of 0.1% trifluoracetic acid. The materials were then subjected to HPLC equipped with a reverse-phase columnunder the condition described under “Experimental Procedures.” The radioactive materials were eluted by application of a linearly increasing concentration gradient of acetonitrile. As shown in Fig. 3, the HPLC procedure separated several phosphopeptides for both preparations phosphorylated by protein kinases C and A. The radioactive phosphopeptidesthat are indicated by the same numbering in thk figure were eluted at the same retention time.

3

4

8 4

8 4

8

CYCLE NUMBER FIG,4. Sequential Edman degradation of purified radioactive tryptic phosphopeptides. The phosphopeptides (5-10 nmol) were each covalently coupled to derivatized glass beads and degraded sequentially from the amino terminus as described under “Experimental Procedures.” Trifluoracetic acid/methanol fractions (anilinothiazolinone amino acids) were evaporated and redissolved in 1 ml of 0.1% trifluoracetic acid. After the determination of radioactivity released at each cycle, anilinothiazolinone amino acid was converted to PTH amino acid by 1 N HC1 treatment. PTH amino acid was estimated as described under “Experimental Procedures.” The deduced sequence is indicated at the inset using one-letter amino acid abbreviations shown in Table11.A circled P shows the phosphorylated seryl or threonyl residue. A, PSfor protein kinase C (270 cprn); B, Ps for protein kinase A (1130 cpm); C, PI? for protein kinase C (730 cprn); D,PI1for protein kinase A (570 cpm); E,PI3 for protein kinase C (280 cpm); F , P15 for protein kinase A (220 cpm). 0, radioactivity; 0, yield of PTH amino acid from the amount of phosphopeptide coupled to glass beads.

Each phosphopeptide was lyophilized, taken up with a small volume of 0.1% trifluoracetic acid, and rechromatographed on the reverse-phase column of HPLC. Each phosphopeptide appeared as a single and symmetric peak,and the amount of the phosphopeptide purified in this way was in the range of 10-70 nmol. Phosphorylation Sites-Theradioactivephosphopeptide (5-20 nmol of 32P)was analyzedfor amino acidsas described under “Experimental Procedures.” The results shown in Table I1 together with the known primary structure of bovine brain MBP (13) indicated that the amino acid sequences of these phosphopeptides are as given in Table 111. Phosphopeptides P5,Pe, PL1, and P14appeared to be products of incomplete digestion. This incomplete digestionmay probably be caused by some steric hindrance due to the phosphorylation of the seryl or threonyl residue just near the site to be cleaved by trypsin (10, 31). Phosphopeptides P,-P6 and PL2contained 1 seryl residue and one phosphate, and it was thus concluded that radioactive phosphate was located on Ser-11, Ser-55, Ser8, Ser-132, Ser-55, Ser-161, and Ser-46, respectively. Phosphopeptides P, and PI, contained 1 seryl residue, 1 threonyl residue, and one phosphate. Upon acid hydrolysis,

Substrate Specificity

ofKinase Protein

C

12497

TABLE IV Summary of major phosphorylation sitesof MBP for protein kinasesC and A The amount of phosphate of each peptide listed was quantitated by the HPLC analysis. The amounts of phosphate covalently attached toPg, Ps, and P13 were less than 0.1 mol/mol of MBP. C and A indicate the results obtained for protein kinases C and A, respectively. Solid bars with numbering in parentheses indicate the seryl or threonyl residue which was phosphorylated. Dotted bars indicate the basic amino acid residue near the phosphorylation site. Amino acid sequence

Peptide

Amount of phosphate C A

mol/mol MBP

Preferable sites

forkinase protein

C

Preferable sites for protein kinase A Common sites for both

protein kinases

Phe-Phe-Gly-&(lB)-Asp-Arg-Gly Gly-Thr-Leu-Ser(l51)-Lys~fi~-Phe "-

P12

His-&g-Asp-T~(34)-Gly-Ile-Leu Ser-&g-Phe-Le+5)-Trp-Gly-Ala

P14

_"Lys-Arg-Pro-&(8)-Gln-Arg-Ser _" Ser-Gln-Arg-Ser(ll)-&$f$Leu _" -

Pa P1

Ly~-~kg-Gly-~(55)-Gly-Lys-A~~ Arg-Gly-Leu-~(llO)-Le~-k&-&~ Gly-&g-Ala-&(132)-Asp-Tyr-Lys Gly-Arg-Asp-Ser(l61)-Arg-Ser-Gly "-

"

"_

__ a ND,

P7

PI5

P2

+ p5

Plo + P11 P 0.1 4 PS

0.3 0.4

ND" ND"

0.1 0.1

0.6 0.3

0.6 0.2 0.7 O.5 0.5 0.3

0.5 0.2 0.5 0.5

0.3

not detectable.

FIG. 5. Relative rates of phosphorylation of major sites in MBP catalyzed by protein kinases A and C. After incubation as indicated under the standard conditions, the radioactive tryptic phosphopeptides, 1.3 nmol MBP equivalent, were separated directly by HPLC asdescribed in the legend to Fig. 3. The radioactivity of each phosphopeptide was determined. A , for protein kinase C; B , for protein kinase A. 0,Ser8 (Pa); A, Thr-34 (P14); 0, Ser-55 (Pz plus P5); 0, Ser-110 (Ploplus Pll); X, Ser-115 (P1& 0, Ser-151 (P,).

0

15

30

INCUBATION TIME radioactive phosphoserineandphosphothreonine were recovered from P7 and p,,, respectively. Thus, the radioactive phosphate was located on Ser-151 and Thr-34. Phosphopeptides Ps-Pll and P13, that contained more than 2 seryl or threonyl residues, were subjected to sequential Edman degradation. As shown in Fig.4, the amino acid sequences of these phosphopeptides coincided with those predicted from the amino acid analysis. The highest release of 32Pwas observed at the second cycle for Ps, the sixth cycle for P9, the third cycle for Plo,the fifth cycle for P,,, and the sixth cycle for P13,indicating that thephosphate was located

15

90 0 (

30

90

MIN 1

on Thr-66,Ser-70, Ser-110, Ser-110, and Ser-18, respectively. The theoretical amino acid sequence of P15 should contain tryptophan (TablesI1 and 111), but theresidue was most likely destroyed during the acid hydrolysis. The sequential Edman degradation of PI5 indicated that radioactive phosphate was located on Ser-115. Specificityand Reaction Rates-The amounts of radioactive phosphate incorporated into the seryl and threonyl residues are summarized in Table IV. Although MBP was fully phosphorylated under the present conditions, the radioactive phosphate covalently attached to each site varied from less than

Substrate Specificity of Protein Kinase C

12498

0.1 to 0.7 mol/mol of the protein, but the precise reason of this poor stoichiometry is not known. Presumably, the reaction rates toward someof these sites are extremely slow. Ser46 and Ser-151were specific for protein kinase C, while Thr34 and Ser-115 were phosphorylated almost specifically by protein kinase A, and the reaction rates by protein kinase C were negligible.Other sites such as Ser-8, Ser-11, Ser-55,Ser110, Ser-132,and Ser-161were phosphorylated commonly by protein kinases A and C. The amounts of phosphate in Ser18 (in PIS),Thr-66 (in Ps), andSer-'70 (in P9)were less than 0.1 mol/mol, and theirreaction rates for these protein kinases, if any, were extremely slow. In another set of experiments the relative rates of phosphorylation toward each of the major sites were estimated. To do so, MBP was phosphorylated by either protein kinase A or C for various periods of time, and the radioactive MBP was isolated and digested by trypsin. Then, the radioactive phosphopeptides were separated and quantitatedby HPLC under the conditions described above. The results are summarized in Fig. 5 . Protein kinase C reacted most rapidly with Ser-8, and thisseryl residue wasalmost fully phosphorylated within 30 min. The enzyme phosphorylated Ser-55, Ser-110,and Ser151 preferentially in this order. On the other hand, protein kinase A phosphorylated most rapidly Ser-110,and also Thr34, Ser-8, Ser-55, and Ser-115at thereaction rates indicated. It was concluded that protein kinases C and A were able to phosphorylate some common sites, but their reaction rates were distinctly different. DISCUSSION

phorylates the synthetic short peptide, Arg-Arg-Lys-Ala-SerGly-Pro-Pro-Val, although the reaction rate appears to be extremely slow (35). This peptide has the same amino acid sequence as that found in H1 histone, and this seryl residue (Ser-35in H1 histone) is known to be phosphorylated rapidly by protein kinase A (19). Further exploration is needed to clarify whether the principle mentioned above may be extended to other protein substrates of protein kinase C. Martenson et al. (36) have described that MBP isolated from rabbit brain contains many phosphate molecules covalently attached to several seryl and threonyl residues. The primary structure of rabbit MBP is not identical with that of bovine MBP,but some of these sites, Ser-7, Ser-56, and Ser163, appear to correspond to Ser-8, Ser-55, and Ser-161 in bovine MBP, respectively, that are phosphorylatedby protein kinase C in vitro. Turner et al. (32) have proposedthat Ser115 in bovine MBP is a preferred and probably major phosphorylation site for protein kinase C, and have discussed a possibility that the phosphorylation of this seryl residue is related to the antigenicity of this protein to induce experimental allergic encephalomyelitis (fora review, see Ref. 37). However, the present studies have revealed that this particular seryl residue appears to be phosphorylated by protein kinase A but not by protein kinase C. The reason for this discrepancy betweenthe two laboratories is not known. Acknowledgments-We would like to thank Dr. Hideki Tachibana (The Graduate School of Science and Technology of KobeUniversity) for the amino acid sequence analysis. We would like also to thank Sachiko Nishiyama and Sachiko Fukase for skillful secretarialassistance.

Wise et al. (7) have described first that protein kinase C REFERENCES phosphorylates rapidly MBP in uitro, and incorporates 4-5 mol of phosphate into 1 mol of this basic protein. Subse1. Kishimoto, A., Takai, Y., Mori, T., Kikkawa, U., and Nishizuka, quently, Turner et al. (32) havesuggestedby fingerprint Y.(1980)J. Bwl. Chem. 255,2273-2276 analysis that protein kinase C shows an intramolecular sub2. Nishizuka, Y. (1984) Nature 308,693-697 3. Nishizuka, Y . (1984) Science 225,1365-1370 strate specificity distinctly different from that of protein kinase A. Extending these observations the present studies 4. Michell, R. H. (1975) Biochim. Biophys. Acta 415,81-147 5. Nishizuka, Y. (1983) Trends Biochem. Sci. 8,13-16 have identified all seryl and threonyl residues in MBP that 6. Berridge, M. J., and Irvine,R. F.(1984) Nature 312,315-321 may be phosphorylated in vitro by both of these protein 7. Wise, B. C., Glass, 0. B., Chou, C.-H. J., Raynor, R. L., Katoh, kinases. The results show a substantial difference in their N., Schatzman, R. C., Turner, R. S., Kibler, R. F., and Kuo, J. substrate specificities. Previous works by Krebs and his asF. (1982) J. BioL Chem. 257,8489-8495 8. Sulakhe, P. V., Petrali, E. H., Daris, E. R., and Thiessen, B. J. sociates and others (for a review, see Ref. 33) have pointed (1980) Biochemistry 19,5363-5371 out that the primary structure of the vicinity of the amino 9. Minakuchi, R., Takai, Y., Yu, B., and Nishizuka, Y. (1981) J. acyl residue to be phosphorylated is one of the important Biochem. (Tokyo) 89,1651-1654 determinant factors for the substrate recognition, and that 10. Carnegie, R. P., Kemp, B. E., Dunkley, P. R., and Murray,A. W. protein kinase A reacts normally with the seryl and threonyl (1973) Biochem. J. 135,569-572 residues that are located at the carboxyl-terminal side close 11. Miyamoto,E., and Kakiuchi,S. (1974) J. Bwl. Chem. 249,27692777 to lysine or arginine. For bovine MBP, Carnegie et al. (10) haveshown that Ser-11, Thr-34, Ser-55, and Ser-110 are 12. Steck, A. J., and Appel, S. H. (1974) J. Biol. Chem. 249,54165420 preferable sites of phosphorylation for protein kinase A. The 13. Eylar, E. H., Brostoff, S., Hashim, G., Caccam, J., and Burnett, present analysis has identified Ser-115, Ser-132,and Ser-161 P. (1971) J. Biol. Chem. 246,5770-5784 as additional sites to be phosphorylated by protein kinase A. 14. Kikkawa, U., Takai, Y., Minakuchi, R., Inohara, S., and NishiAllof these seryl and threonyl residues are located at the zuka, Y.(1982) J. Biol. Chem. 257, 13341-13348 carboxyl-terminal side close to basic amino acids. Contrary 15. Bechtel, P.J., Beavo, J. A., and Krebs,E. G. (1977)J. Bid. Chem. 252,2691-2697 to protein kinase A, protein kinase C appears to favor hyY., and Eylar,E. H. (1970)Arch. Biochem. Biophys.1 3 8 , droxyl amino acids that are located at the amino-terminal 16. Oshiro, 392-396 sideclose to lysine or arginine. Both Ser-46 and Ser-151, 17, Chou, F. C.-H., Chou, C.-H. J., Shapira, R., and Kibler, R. F. which are specifically phosphorylated by protein kinase C, (1976) J. Biol. Chem. 251,2671-2679 have a basic amino acid at the carboxyl- but not the amino- 18. Brostoff, S. W., Reuter, W., Hichens,M., and Eylar,E. H. (1974) J. BioLChem. 249,559-567 terminal side. It is worth noting that all seryl residues in MBP which are phosphorylatedcommonly by protein kinases 19. Hashimoto,E., Takeda, M., Nishizuka, Y., Hamana, K., and Iwai, K. (1976) J. Bid. Chem. 251,6287-6293 C and A have basic amino acidsat both amino- and carboxyl- 20. Glynn, I. M., and Chappell,J. B. (1964)Biochem. J. 90,147-149 terminal sides. In epidermal growthfactor receptor, Thr-654, 21. Takai, Y.,Kishimoto, A., Iwasa, Y., Kawahara, Y., Mori, T., and that is located at the amino-terminal side close to arginine, Nishizuka, Y.(1979) J. Bid. Chem. 254,3692-3695 has been recentlyidentified as thetarget site of protein kinase 22. Walsh, D. A., Ashby, C.D., Gonzalez, C., Calkins,D., Fischer, E. H., and Krebs,E,G. (1971) J. Biol. Chem. 246,1977-1985 C (34). However, it is reported that protein kinase C phos-

Substrate Specificity 23. Takai, Y., Kishimoto, A., Inoue, M., and Nishizuka, Y. (1977) J. Biol. Chem. 252,7603-7609 24. Hunter, T., and Setton, B. M. (1980) Proc. Natl. Acad. Sci. U. S. A. 7 7 , 1311-1315 25. Criss, W. E., Yamamoto, M., Takai, Y., Nishizuka, Y., and Moris, H. P. (1978) Cancer Res. 38.3532-3539 26. Laursen,' R. A. (1971) Eur. J. Biochem. 20,89-102 27. Potter, R. L., and Taylor, S.S. (1979) J. Biol. Chem. 254,24132418 28. Machleidt, W., Wachter, E., Scheulen, M., and Otto, J. (1973) FEBS Lett. 37,217-220 29. Bridgen, P., and Cross, G. A. M. (1976) Nature 263,613-614 30. Erlichman, J., Rosenfeld, R., and Rosen, 0.M. (1974) J. BWZ. Chem. 249,5000-5003

ofKinase Protein

C

12499

31. Carnegie, P. R., Dunkley, P. R., Kemp, B. E., and Murray, A. W. (1974) Nature 249,147-150 32. Turner, R. S., Chou, C.-H. J., Mazzei, G. J., Dernbure, P., and Kuo, J. F. (1984) J. Neurochem. 43,1257-1264 33. Krebs, E. G., and Beavo, J. A. (1979) Annu. Rev. Biochem. 4 8 , - . 923-959 84. Hunter, T., Ling, N., and Cooper, J. A. (1984) Nature 3 11,480483 35. O'Brian, C. A., Lawrence, D. S., Kaiser, E. T., and Weinstein, I. B. (1984) Biochem. Biophys. Res. Commun. 124,296-302 36. Martenson, R. E., Law, M. J., and Deibler, G. E. (1983) J. Biol. Chem. 258,930-937 37. Day, E. D. (1981) in Contemporary Topics in Molecular Zmmunology (Inman, F.P., and Mandy, W. J., eds) Vol. 8, pp. 1-39,

Plenum Press, New York