Proc. Nati. Acad. Sci. USA Vol. 84, pp. 1824-1828, April 1987 Biochemistry Cauliflower mosaic coat protein is phosphorylated in vitro by a virion-associated protein kinase (ADP/casein kinase/ virus/retrold virus/reverse transcription) Jose MARTINEZ-IZQUIERDO AND THOMAS HOHN Friedrich Miescher-Institut, P.O. Box 2543, CH-4002 Basel, Switzerland Communicated by Diter von Wettstein, November 21, 1986 (receivedfor review October 20, 1986)

ABSTRACT A protein kinase has been found to be asso- branes at 40 V (constant voltage) overnight (18). Treatment ciated with particles ofthe cauliflower mosaic virus. of the nitrocellulose membranes was as described (19). This protein kinase can phosphorylate endogenous viral Rabbit anti-p37 serum diluted 1:1000 with the milk-based proteins in vitro and exchange substrates with casein kinase solution A (19) and 1% goat preimmune serum and peroxi- type H. The activity is not affected by cAMP but is enhanced dase-coupled goat anti-rabbit IgG (Bio-Rad) was used. considerably by ADP. The cofactor is either Mn2+ or Mg2+, Protein Kinase Assay. The standard in vitro reaction mix- and the phosphate donor is either ATP or GTP. Serine and ture contained 15 pug of viral particles, 50 mM Tris HCl (pH threonine residues are phosphorylated. 7.4), 0.6 mM MnCI2 (unless otherwise specified), 1 mM dithiothreitol, 2 AtM ATP, and 2.5 ACi of [y32P]ATP (5000 The structural and functional proteins of various animal Ci/mmol; 1 Ci = 37 GBq; Amersham) in a final volume of 50 , including members of the Retroviridae and Al. The reaction mixture was incubated at 370C for 30 min families, are phosphorylated by virus-asso- (unless otherwise specified), and the reaction was terminated ciated protein kinases (1-12). In contrast, among the plant by the addition of 16.7 /l of 5 x NaDodSO4/PAGE sample viruses, only cauliflower mosaic virus (CaMV) contains buffer (46) and a 4-min incubation at 90'C. After electropho- phosphorylated proteins (13), but little is known about the resis, phosphorylated proteins were visualized by staining protein kinase responsible. Various properties of CaMV are and autoradiography. For exact quantitation, the individual related to those of animal (14). CaMV is a bands were cut out, and radioactivity was measured in a double-stranded DNA virus that replicates through an RNA liquid scintillation counter. intermediate using a , which is closely Phosphoamino Acid Analysis. Analysis was as described related to the same enzyme from retroviruses. (20). In this report we show that a protein kinase is associated firmly with CaMV particles and can act on endogenous and RESULTS exogenous substrates. The activity is characterized biochem- ically and compared to other known protein kinases. Similar CaMV Structural Proteins and in Vitro Phosphorylation. results have been independently described by another group The NaDodSO4/PAGE analysis of CaMV particles reveals a complex pattern of major and minor protein bands (21, 22). (15). The pattern shown in Fig. 1, lane 1, is obtained only when fresh virus preparations are used. When stored preparations MATERIALS AND METHODS are used (e.g., for 3 weeks at 40C), smaller-sized bands Virus Isolation. CaMV, strain CM4-184 (16) was propagat- predominate (including bands at 32 and 27 kDa) resembling ed in turnips ( rapa, Just Right). Viral particles free those in ref. 23, which indicate proteolytic activity. Antise- of inclusion bodies were isolated as described (17). rum prepared against the major 37-kDa protein species Purification of Viral Capsid Proteins and Antisera Produc- (anti-p37) was used to detect related coat protein species by tion. CaMV particles were subjected to preparative NaDod- immunoblotting (Fig. 1, lane 2). Major bands at 37 and 44 kDa S04/PAGE (46). After electrophoresis the gel was soaked in (p37 and p44, respectively) as well as minor bands at 40, 57, 0.5 M KCl at 40C, and the opaque gel bands were excised and 80, 90, and 96 kDa (p40, p57, p80, p90, and p96, respectively) washed three times with 0.5 M NaCl. Pieces of gel (1 cm3) were detected on the immunoblots, suggesting that these were placed in the elution chamber of an ISCO electropho- protein species may indeed share common sequences. Most retic sample concentrator. Elution of p37 protein was at 3 W likely p37 and p44 (and perhaps p40) are derived from p57 by for 15 hr in 0.05% NaDodSO4/0.1 M glycine/0.0125 M proteolytic cleavage, and p80 and p96 are dimers of p37 and Tris HCl, pH 8.3. After completion of extraction, the 200-gl of p44, respectively (21, 24). sample was collected, and the sample well was washed twice When viral particles were incubated with [y-32P]ATP in the with 100 /l of water; sample and washings were pooled and presence of Mn2+ or Mg2+, they were phosphorylated. The stored at -20'C. labeled phosphate incorporated was resistant to DNases and Anti-p37 sera was raised in rabbits by i.m. injection of 100 RNases but sensitive to proteinase K. Electrophoresis re- ,tg ofpurified p37 emulsified in Freund's complete adjuvant. vealed that the phosphorylated proteins were coat protein Further booster doses were given under similar conditions (in species identified as pp44 (75%), pp57 (8%), and pp96 (15%) Freund's incomplete adjuvant) after 4, 7, 9, and 10 weeks. in Fig. 1, lane 3. Proteins p37, p40, and p80 remained Serum samples were collected 10 days after the last injection. essentially unphosphorylated, although p37 represents 60% Immunoblotting. Proteins were electrotransferred from of the capsid protein. This phosphorylation pattern corre- NaDodSO4/polyacrylamide gels to nitrocellulose mem- sponds to the one obtained from viral particles labeled in vivo with 32p (ref. 13 and unpublished data) and strengthens the The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviations: CaMV, cauliflower mosaic virus; ORF, open reading in accordance with 18 U.S.C. §1734 solely to indicate this fact. frame.

Downloaded by guest on September 24, 2021 1824 Biochemistry: Martinez-lzquierdo and Hohn Proc. Natl. Acad. Sci. USA 84 (1987) 1825

1 2 3 pp96 (labeled pair). The stability of the label to prolonged

E 0800 37 1 2 ._0 4- (U o ~~~~~~~~96- E a:.1 otpoen fCM.Via atce eesbetdt -0 -70 M 400 -5 E

40 80 120 Time, min B

M 30- OE 0 -6 600 U) E a20+ 02a 0 400

a) o 10 =a 200 (0 co 0 0L

6 40 Cation, mM Time, min FIG. 2. Properties ofCaMV protein kinase and the peptide-phosphate bond. For all experiments standard protein kinase reaction conditions were used with the variations indicated. Effects were assayed as follows. The reaction mixtures derived from 15 gg ofvirus particles were boiled in NaDodSO4 and transferred quantitatively to lanes of NaDodSO4/polyacrylamide gels and electrophoresed. The pp44 bands were cut out, and their radioactivity was determined by scintillation counting. (A) Stability of the ester bond. The phosphorylated viral particles were incubated with 0.1 M HCO at 370 (A), 1 mM unlabeled ATP (r) or, after adjustment of the buffer to calf intestine alkaline phosphatase reaction buffer, with 10 units of calf intestine alkaline phosphatase (+). (Inset) Autoradiogram ofviral particles before (lane 1) and after (lane 2) calfintestine alkaline phosphatase treatment; the numbers in the right and left margins represent protein sizes in kDa (for calf intestine alkaline phosphatase: 70 kDa; ref. 25). (B) Divalent cation requirements. The kinase reaction was performed in the presence of the following divalent cation chlorides at concentrations shown: Mn2+ (0), Mg2+ (+), Co2+ (o), Zn2+ (A), and Ca2+ (x). (C) Inhibitors. The kinase reaction was performed in the presence of quercetin (o, values x 0.1 AuM), heparin (+, values in Iug/ml), or pyrophosphate (o, values x 10 mM) at concentrations indicated. (D) Time course of phosphate incorporation. The kinase reaction was performed under standard conditions except that 5.0 (A), 2.0 (o), 0.5 (+), or 0.25 (n) AM ATP final concentration was used. Downloaded by guest on September 24, 2021 1826 Biochemistry: Martinez-Izquierdo and Hohn Proc. Natl. Acad Sci. USA 84 (1987) by two-dimensional separation of the products by electro- Table 1. In vitro characteristics of CaMV-associated phoresis on cellulose thin-layer plates (27) confirmed the protein kinase expected bonds: 84% of the radioactivity/was present as Additions to standard reaction mixture phosphoserine; 15% was present as phosphothreonine; and Activity, % <2% was present as phosphotyrosine. When isolated pp44 None 100 was analyzed, only phosphoserine was found. GTP and Mg2+ instead of ATP and Mn2+ 60 The phosphate label is probably located at the outside of cAMP at 5 ;LM 110 the virus particle since it could be removed by treatment of cGMP at 5 I&M 108 intact virus particles' with calf intestine alkaline phosphatase AMP at 2 AuM 123 (Fig. 2A). NaDodSO4 treatment immediately after calf intes- ADP at 0.2 AM 159 tine alkaline phosphatase addition preserved nearly the full ADP at 2 ,uM 341 amount of phosphate label on pp44, showing that dephos- ADP at 10 AM 542 phorylation did not continue after virus dissociation. The Guanosine at 2 ,uM 87 measurement of phosphate on the electrophoretically sepa- Putrescine at 100 jAM 84 rated proteins rather than in total protein trichloroacetic acid Spermidine at 100 ,uM 92 precipitation is necessary, since calfintestine alkaline phos- Spermine at 100 jIM 76 phatase is apparently autophosphorylated during dephospho- KCl at 100 mM 47 rylation of the substrate as had been reported for bacterial alkaline phosphatase (Fig. 2A Inset; ref. 28). pH 6.0 and 7.4 and between 37 and 45TC. Temperatures Properties of the Protein Kinase Activity. Two types of >450C inhibited the reaction, and the enzyme became totally observations showed that the kinase activity is strongly inactivated after 10 min at 680C. associated with the virus particles, i.e., with its substrate: (t) The effect of several nucleotide analogues and other Extensive dilution ofthe particles did not reduce the activity. substances on the reaction was measured using the same (it) The specific activity remains unchanged upon consecu- assay (Table 1). The activity was independent of cyclic tive purification steps. In sucrose- and CsCl-density gradient nucleotides and AMP either in the presence ofATP or ofGTP centrifugation, substrate and enzyme copurify, and little as phosphate donors. ADP had significant stimulatory effect activity is lost (Fig. 3). in the range tested (0.2-10 juM). Of several known protein In general, protein kinases use divalent cations as cofac- kinase inhibitors tested only pyrophosphate (29) and those tors. In our standard assay the CaMV-associated protein specific for casein kinase II, namely heparin and quercetin kinase showed a strong preference for Mn2+ as the divalent (30, 31), had a significant effect. As little as 10 ttM pyro- cation with an optimal concentration between 0.4 and 0.7 mM phosphate caused an 18% inhibition, which increased to 95% (Fig. 2B). Mg2' can substitute for Mn2+, but the optimal with 500 ttM pyrophosphate. The inhibitory effect ofheparin concentration is at 10 mM and only 60% of the Mn2+- and quercitin was significant, although much lower with supported incorporation can be attained. With Co2+, Ca2+, or CaMV-associated kinase than with casein kinase II. For Zn2+ at concentrations between 0.5 and 10 mM the activity instance, heparin at 0.2 jig/ml inhibits casein kinase II by was very low. 80% (32), while in our experiments heparin at 10 jig/ml was CaMV protein kinase was active over broad pH and required to inhibit the CaMV-associated enzyme to the same temperature ranges. Optimal activity was observed between extent (Fig. 2C). In contrast to casein kinase type 11(30), no stimulatory effect with polyamines or monovalent cations Fraction was observed (Table 1). 2 4 6 8 10 12 C 2 4 6 8 10 12 C Kinetics and Extent of Phosphorylation. The incorporation of phosphate into CaMV coat proteins had two phases, probably reflecting different phosphorylation sites. A rapid early rate was observed, followed by a slower rate between ._ 30 and 120 min (Fig. 2D). Saturation had not yet been reached (l) C-) by 120 min. 0-: I -0~*_ From the relation between the ATP concentration and the initial phosphate incorporation rate, the protein kinase affin- ity constant and the maximal rate ofphosphate incorporation for ATP could be determined using the Lineweaver-Burk w...s~A plot. At optimal divalent cation concentration (0.6 mM Mn2+), the values determined for ATP as substrate and (I,a) 0 derived from virus particles at 300 jig/ml were: Km = 0.36 C.) AuM and V,,,, = 42.4 fM/min. With GTP as phosphate donor cn # the same proteins were phosphorylated as with ATP (data not _ _ _a 40 -.%.. am w- am shown). However, the preferred divalent cation for GTP was Mg2e at an optimal concentration of 10 mM. The affinity of Immunoblot Autoradiogram the protein kinase for GTP was high, although lower than for ATP, as has been described for the casein kinase II activities FIG. 3. Copurification of protein kinase with CaMV viral parti- (30). The estimated value for the Km was 2.3 jiM. cles. Samples (2 mg) of viral particles were separated by CsCl- and The amount of phosphate incorporated per molecule of sucrose-gradient centrifugation, and fractions were collected from pp44 was rather low. At the initial phase approximately one the top. Every second fraction was assayed for kinase activity by phosphate was incorporated per 650 molecules of pp44, incubation in the kinase reaction mixture, electrophoresis, and perhaps because most of the available substrate sites are autoradiography (Right) and for antigenicity by immunoblotting with already phosphorylated and the phosphate is not anti-p37 antibodies (Left). In both, CsCl- (Upper) and sucrose- generally gradient centrifugation (Lower), the incorporation of radioactive exchanged. In fact, as discussed above, the radioactive label phosphate into coat proteins strongly correlated with the amount of could not be chased with cold ATP, and preliminary exper- coat protein detected by anti-p37 antibodies [lanes 4, 6, and 8 (CsC1) iments indicate that more extensive phosphorylation occurs and lanes 6, 8, and 10 (sucrose)]. Lanes C contain the viral when viral particles had been pretreated with calf intestine preparation before centrifugation. alkaline phosphatase. Downloaded by guest on September 24, 2021 Biochemistry: Martinez-lzquierdo and Hohn Proc. Natl. Acad. Sci. USA 84 (1987) 1827 A B Like most of the virus-associated protein kinases CaMV 1 2 3 4 1 2 3 4 5 6 protein kinase is cyclic nucleotide independent, is inhibited 10 Eby pyrophosphate, and phosphorylates serine and threonine residues. In these respects it also resembles the casein kinases. These exist as types I and II, which can be @131-96 distinguished by their preference for the phosphate donor. - 57 The ubiquitous type I uses only ATP, while type II, which so F -_44Eb far has not been found in , and CaMV protein kinase use either ATP or GTP. In fact, casein kinase type II and CaMV protein kinase can exchange their substrates and are inhibited by the same effectors, heparin and quercetin, i.. although casein kinase is more sensitive. Unlike casein IL_ kinase type II, the CaMV protein kinase is not activated by K+ or polyamines. Its activation by ADP, on the other hand, Twrr has similarities to a reductase kinase (37) and a phosphoryl- ase kinase (38) in which ADP was identified as allosteric P w activator. Casein kinase type II phosphorylates serine and threonine residues in the neighborhood of acidic amino acids. Phos- FIG. 4. Cross specificities. (i4) Exogenous substrates for the phoserine was, in fact, found in pp44, and phosphoserine and CaMV-associated protein kinase. The khnase reaction mixture con-' ' tained the following exogenous substrates at 5 ,&g: lysine-rich phosphothreonine were found in total viral protein. If the (Type III-S, Sigma; lan4e 1); arginine-rich histones (Type specificities of casein kinase are valid for CaMV protein II-A, Sigma; lane 2); none (lane 3); and casein (lane 4). Molecular kinase, the acidic environments around the serine and thre- sizes (in kDa) of the phosphorrylated coat protein species are onine residues, both present at the N terminus and at the indicated on the margin. (B) Phosj horylation of CaMV coat proteins presumptive C terminus of CaMV capsid protein pp44, are by exogenous kinases. Casein kingase II (from rabbit skeletal muscle; possible sites. Proteins pp44 and p37 were subjected to kindly provided by B. Hemmings; lanes 1 and 2); catalytic subunit of N-terminal amino acid sequencing by J. Hofsteenge, S. cAMP-dependent protein kinase (source as above; lanes 3 and 4); and Hile S small ribosomal subunit 6 (S6) kinase (kindly provided by P. Jeno; Stone,StermmalJ.M.-I., and T.H. (unpublishedsequncing bylJ.results). While p37 has lanes 5 and 6). Viral particles wiere heated to 68°C for 10 min to resisted analysis so far, perhaps due to a blocking of the N inactivate the CaMV protein kinaLse. Samples (15 ltg) of these were terminus as observed with capsid proteins, the first tested as substrates for the protein kinase in question (lanes 2,4, and nine amino acids of the N terminus of pp44 could be 6); control substrates were 5 pyg oif casein (lanes 1 and 3) and 5 Ag of determined (Fig. 5). This provides the final proof for the 40S ribosomal proteins (lane 5). TIle kinase reaction for casein kinase earlier evidence (24) that CaMV open frame II and cAMP-dependent protein kinase was performed as for CaMV reading (ORF) protein kinase, except that Mg2+ iwas used instead of Mn2+. The S6 IV encodes the coat protein and shows that pp44 is derived kinase reaction mixture contained150 mM Mops, 20 mM dithiothrei- from the original translation product by N-terminal cleavage. tol, 10 mM MgCl2, 30 ,uM ATP (cointaining 5 1ACi of [y32P]ATP), and The C terminus of pp44 remains uncertain but a cleavage at 21 ,ug of 40S ribosomal subunit. C', casein. The CaMV protein pp4 this terminus also should be considered (24). Sites at the N ic1sineliratpAuLav;au hvuy arrr%,uwsin.TinmLaneslanav1-1Aenzyme auopnospnoryauon ana phosphorylation of impurities is also evident. M A E SILDRTINRFWYNLGEDCLSESQFDLM 30 I R L M EE S L D G D Q I I D L T S L P S D N L Q V E Q V M 60 4 * Specificity of CaMV Kinase Activity and CaMV Kinase T TT D TDSDII5E EI5EFLLAIGEIISEDIESDSE 90 Substrate. The CaMV protein kinase can also phosphorylate some heterogenous substrates, namely lysine- and arginine- P E F E Q V R M D R T G G T E I PK E E D G E G P S R Y N E 120 rich histones and casein (Fig. 4A). After heat inactivation (10 R KR K T P E D R Y F P T Q P K T I P G Q K Q T S M G M L N 150 min, 68°C) of CaMV protein kinase, CaMV particles were I D C Q I N RR T L I DD W A A E I G L I V K T N R E D Y L 180 susceptible to phosphorylation by added casein kinase type D PE T I L L L M E H K T S G I A KEL I R N T R W N RTT 210 II. pp44 was labeled, but p37 was not labeled (Fig. 4B), G D I I E Q V I N A M Y T M F L G L NY _ D N K V A E K I D 240 resembling the situation with CaMV protein kinase. No coat protein phosphorylation was observed either with the cata- E Q E K A K I R M T KL Q L F D I C Y L E E F T C DY E K N 270 lytic subunit of cAMP-dependent protein kinase or with S6 Y K T E L A D F P G Y I N Q Y L K I P I I G E K A L T R 300 kinase. F R E A N G T S I Y S L G F A A K I V K EE L S K I C D L 330

KK Q K K L K K F N K K CC S I G E A S V E Y GG K K T S 360

DISCUSSION K K K Y H K R Y K K R Y K V Y K P Y KK K K KF R S G K Y F 390 These studies and the independent one of Menissier de K P K E K K G _ K R K Y C PK G K K DIC R C W I C N I E G H 420 Murcia et al. (15) show that CaMV virions contain both the Y A N E CIP N R Q S EK A H I L QQ A E N L G L Q P V E E 450

P YE K enzyme and the substrate of a protein kinase reaction. The GVQ E V FILE Y E E E E ETLSTEES DD G[ Fs 480 observation is so far specific for plant viruses, but not for viruses in general. Several animal viruses have protein kinase T|IE|D|S D S D 488 activities. Included are (9) hepadnaviruses and retroviruses FIG. 5. Amino acid sequence of gene encoding ORF IV protein. (6, 33), which together with the caulimoviruses and The amino acid sequence corresponding to the CaMV strain retrotransposons [yeast Ty (34), Drosophila copia (35), etc.] CM4-184 (39) ORF IV is shown with residues 76-455 representing form the group of retroid elements (15). Although the kinase pp44. Its N terminus is as determined by microsequencing (se- activity in retrotransposons is not yet identified, at least for quenced portion marked with arrows; J. Hofsteenge and S. Stone, Ty it is likely to exist, since its structural proteins are personal communication), but its C terminus is only tentatively phosphorylated (36). Thus another property unites this group assigned by estimation from its electrophoretic mobility. Consensus phosphorylation sites for casein kinase II are boxed with solid lines; in addition to their replication through reverse transcription the RNA binding motif(14) is boxed with dashed lines; and serine and and their similar genome organization (14). threonine residues are underlined. Downloaded by guest on September 24, 2021 1828 Biochemistry: Martinez-Izquierdo and Hohn Proc. Natl. Acad Sci. USA 84 (1987) terminus and at the C terminus ofthe thus defined pp44 would 9. Petit, M. A. & Pillot, J. (1985) J. Virol. 53, 543-551. meet the specific preference of casein kinase type II (40) for 10. Wilson, M. E. & Consigli, R. A. (1985) Virology 143, 526-535. Ser/Thr-Glu/P-Ser-Glu/Asp (acceptor amino acids are bold 11. Hassauer, M., Scheidtmann, K. H. & Walter G. (1986) J. type). Other possible sites on the ORF IV translation product Virol. 58, 805-816. 12. Lamb, R. A. & Choppin, P. W. (1977) Virology 81, 382-397. are located outside p44, and two ofthese might be responsible 13. Hahn, P. & Shepherd, R. J. (1982) Virology 116, 480-488. for the threonine phosphorylation of total viral protein. 14. Bonneville, J. M., Futterer, J., Gordon, K., Hohn, T., Mar- It was reported that phosphorylation of virus capsid tinez-Izquierdo, J., Pfeiffer, P. & Pietrzak, M. (1986) UCLA proteins of certain retroviruses regulates their affinity for Symp. New Series 48, in press. RNA. Interestingly in murine type C retroviruses, the affinity 15. Menissier de Murcia, J., Geldreich, A. & Lebeurier, G. (1986) is decreased (41), while in avian retroviruses it is increased J. Gen. Virol. 67, 1885-1891. upon phosphorylation (18). The relevant phosphorylation site 16. Howarth, A. J., Gardner, R. C., Messing, J. & Shepherd, in avian sarcoma virus is located close to the RNA binding R. J. (1981) Virology 112, 678-685. motif ofthe capsid protein. This motif consists ofa pattern of 17. Gardner, R. C. & Shepherd, R. J. (1980) Virology 106, 159-161. three cysteine and one histidine residues (14, 42), that are 18. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. thought to form a complex with a divalent cation ("finger"; Acad. Sci. USA 76, 4350-4354. ref. 43). The Cys/His motif exists also in CaMV (Fig. 5) with 19. Johnson, D. A., Gautsch, J. W., Sportsman, J. R. & Elder, one of the possible phosphorylation sites nearby. Thus J. H. (1984) Gene Anal. Tech. 1, 3-8. phosphorylation may have a similar function in CaMV, 20. Hunter, T. & Sefton, B. M. (1980) Proc. Natl. Acad. Sci. USA although one would then have to explain why the phosphate 77, 1311-1315. bond is accessible to calf intestine alkaline phosphatase. 21. Al Ani, R., Pfeiffer, P. & Lebeurier, G. (1979) Virology 93, In any case phosphorylation of proteins provides a general 188-197. means of posttranslational regulation due 22. Burger, J. G. & DuPlessis, D. (1983) J. Virol. Methods 7, to protein modifi- 11-19. cation. Thus it might also be involved in virus assembly (44), 23. Hull, R. & Shepherd, R. J. (1976) Virology 31, 217-220. control of protein processing (6, 45), or stabilization/ 24. Franck, A., Guilley, H., Jonard, G., Richards, K. & Hirth, L. destabilization of the capsid (7, 10). It could also be that the (1980) Cell 21, 285-294. main functional target is not capsid protein, but other viral 25. Besman, M. & Coleman, J. E. (1985) J. Biol. Chem. 260, proteins, such as reverse transcriptase (33), virus protease, or 11190-11193. even certain host proteins, and that it plays some regulative 26. Bitte, L. & Kabat, P. (1974) Methods Enzymol. 30, 563-590. role there. Further study in this field is required since the role 27. Martin-Perez, J. & Thomas, G. (1983) Proc. NatI. Acad. Sci. of protein phosphorylation is not satisfactorily explained for USA 80, 926-930. 28. Roberts, C. H. & Chlebowski, J. F. (1985) J. Biol. Chem. 260, any of the virus-associated protein kinases. 7557-7561. Our work was initiated by the observation of sequence 29. Wilson, M. E. & Consigli, R. A. (1985) Virology 143, 516-525. homologies of CaMV ORF I with ATP binding sites of 30. Hathaway, G. M. & Traugh, J. A. (1983) Methods Enzymol. various kinases, especially those present in tumor genes 99, 317-331. (unpublished data), and our working model was that the ORF 31. Maxwell, S. A. & Arlinghaus, R. B. (1985) Virology 143, I protein phosphorylates the CaMV capsid. However, the 321-333. kinases we have compared are either cAMP dependent or 32. Hathaway, G. M., Luben, T. H. & Traugh, J. A. (1980) J. preferentially phosphorylate tyrosine, unlike the CaMV- Biol. Chem. 255, 8038-8041. associated kinase. Also CaMV ORF I protein is only loosely 33. Lee, S. G., Miceli, M. V., Jungmann, R. A. & Hung, P. P. associated with virus particles can (1975) Proc. Natl. Acad. Sci. USA 72, 2945-2949. and be separated by 34. Mellor, J., Malim, M. H., Gull, K., Tuite, M. F., McCready, sucrose- or CsCl-gradient centrifugation, while the CaMV S., Dibbayawan, T., Kingsman, S. M. & Kingsman, A. J. protein kinase remains associated. (1985) Nature (London) 318, 583-586. We thank J. Menissier de Murcia for communication of her results 35. Emori, Y., Shiba, T., Kanaya, S., Inouye, S., Yuki, S. & prior to publication, J. Hofsteenge and S. Stone for protein sequenc- Saigo, K. (1985) Nature (London) 315, 773-776. ing and discussions concerning protein chemistry, P. Jeno and B. 36. Mellor, J., Fulton, A. M., Dobson, M. J., Roberts, M. A., Hemmings for discussions concerning protein kinases and for vari- Wilson, W., Kingsman, A. J. & Kingsman, S. M. (1985) Nu- ous protein kinase samples, P. Jeno, B. Hemmings, J. Futterer, and cleic Acids Res. 13, 6249-6263. W. Filipowicz for critically reading this manuscript, and our other 37. Harwood, H. J., Brandt, K. G. & Rodwell, V. W. (1984) J. colleagues for many discussions and collaborations. This work was Biol. Chem. 259, 2810-2815. supported by an EMBO fellowship to J.M.-I. 38. Cheng, A., Fitzgerald, T. J. & Carlson, G. M. (1985) J. Biol. Chem. 260, 2535-2542. 1. Akusjaervi, G., Philipson, L. & Petterson, U. (1978) Virology 39. Howarth, A. J., Gardner, R. C., Messing, J. & Shepherd, 87, 276-286. R. J. (1981) Virology 112, 678-685. 2. Roby, C. & Gibson, W. (1986) J. Virol. 59, 714-727. 40. Chan, P.-K., Aldrich, M., Cook, R. G. & Busch, H. (1986) J. 3. Howard, C. R. & Buchmeier, M. J. (1983) Virology 126, Biol. Chem. 261, 1868-1872. 538-547. 41. Sen, A., Sherr, C. J. & Todaro, G. J. (1977) Cell 10, 489-496. 4. Bell, J. C., Brown, E. G., Takayesu, D. & Prevec, L. (1984) 42. Covey, S. N. (1986) Nucleic Acids Res. 14, 623-633. Virology 132, 229-238. 43. Vincent, A. (1986) Nucleic Acids Res. 14, 4385-4391. 5. Paoletti, E. & Moss, B. (1972) J. Virol. 10, 417-424. 44. Garrcea, R. L., Ballmer-Hofer, K. & Benjamin, T. L. (1986)J. 6. Yoshinaka, Y. & Luftig, R. B. (1982) Virology 116, 181-195. Virol. 59, 185-188. 7. Yoshinaka, Y., Shames, R. & Luftig, R. B. (1983) J. Gen. 45. Naso, R. B., Karshin, W. L., Wu, Y. H. & Arlinghaus, R. B. Virol. 64, 95-102. (1979) J. Virol. 32, 187-198. 8. Fu, X., Phillips, N., Jentoft, J., Tuazon, P. T., Traugh, J. A. & 46. Hames, B. D. & Rickwood, D., eds. (1981) Gel Electrophore- Leis, J. (1985) J. Biol. Chem. 260, 9941-9945. sis of Proteins: A Practical Approach (IRL, Oxford). Downloaded by guest on September 24, 2021