Protein Kinase Activities Associated with the Virions of Pseudorabies and Herpes Simplex Virus

Transcription

1 J. gen. ViroL (1985), 66, Printed in Great Britain 661 Key words: herpesvirus/protein kinase /phosphoproteins Protein Kinase Activities Associated with the Virions of Pseudorabies and Herpes Simplex Virus By WILLIAM S. STEVELY,* MATILDA KATAN,t VALERIE STIRLING, GRAEME SMITH AND DAVID P. LEADER Department of Biochemistry, University of Glasgow, Glasgow G12 8QQ, U.K. (Accepted 3 December 1984) SUMMARY Protein kinase has been extracted in soluble form from virions of pseudorabies virus using 10~ NP40, 0.6 M-NaC1. Chromatographic analysis of the extract on DEAEcellulose and on phosphocellulose showed it to contain more than one kinase. The activity responsible for the phosphorylation of the major phosphoproteins (mol. wts , and 72000) of virions was found to be similar in its properties to the host enzyme casein kinase I1. Purified casein kinase II from ascites cells or from pig liver was able to phosphorylate heat-inactivated virions. In addition to the major phosphoproteins, active virion preparations were able to phosphorylate a minor low molecular weight phosphoprotein, incorporation into which could be stimulated by the addition of cyclic AMP to the assay. Purified host cyclic AMP-dependent protein kinase also phosphorylated this protein in heat-inactivated virions. Analysis of herpes simplex virus type 1 showed that the major phosphoproteins (VP12 and VP23) could be phosphorylated in heat-inactivated virions by added casein kinase II. One of these (VP12) together with a further minor phosphoprotein (VP 14) could be phosphorylated by cyclic AMP-dependent protein kinase. INTRODUCTION Previous reports from a number of laboratories have shown that protein kinase activities are associated with several herpesviruses including herpes simplex virus (HSV) (Rubenstein et al., 1972; Lemaster & Roizman, 1980) and pseudorabies virus (PRV) (Tan, 1975). The most detailed of these reports was on the activity (or activities) associated with HSV (Lemaster & Roizman, 1980) which was shown to partition with capsid-tegument structures, to require Mg 2 or Mn 2 and to be able to use either ATP or GTP. Only virus structural proteins were phosphorylated; exogenously added substrates were not phosphorylated. For equine herpesvirus, Randall et al. (1972) found that the kinase associated with virions could utilize added protamine or histone as substrate. For tupaia herpesvirus, Flugel & Darai (1982) reported that the activity associated with virions was not enhanced by added substrates. For PRV much less information is available though it was suggested (Tan, 1975) that the kinase associated with virus, as well as other kinases tested, could utilize exogenously added substrates such as histone, protamine, casein and phosvitin. The function of the herpesvirus kinase activity is not known; no previous report has established whether there is a single activity in virions or more than one activity, or whether the activity originates from virus or host. There have been no reports of the extraction of herpesvirus kinase in a soluble form from virions, apart from a brief preliminary note in the report by Lemaster & Roizman (1980). Extraction and fractionation is clearly fundamental to attempts to characterize these protein kinases, and the ability of the virion extracts to phosphorylate both acidic and basic proteins makes it likely that the virions of herpesviruses contain more than one protein kinase. We describe here studies on extracts of the virions of PRV which contain a l On leave of absence from the Institute for Nuclear Sciences 'Boris Kidric', Belgrade, Yugoslavia SGM

2 662 W. S. STEVELY AND OTHERS protein kinase activity able to phosphorylate the major virion phosphoproteins. We have resolved this protein kinase from other kinases present in the virion and demonstrate its similarity to the cellular enzyme, casein kinase II. METHODS Cells. BHK-21/C13 cells were maintained in monolayer cultures in modified Eagle's medium containing 10% calf serum, as previously described (Stevely, 1975). Virus. PRV was originally derived from a stock preparation (Kaplan & Vatter, 1959) and has been plaquepurified several times. Virus stocks were prepared from infected BHK monolayer cultures as previously described (Chantler & Stevely, 1973). HSV-I was the F strain, and the original stock was obtained from Y. Becker, Hebrew University, Jerusalem, Israel. Purification ofvirions and nucleocapsids. Virions and nucleocapsids were prepared from cells which had been infected for 24 h. Cultures were infected with virus at an input multiplicity of 20 p.f.u./cell. To label proteins, infected cells were incubated between 4 and 24 h post-infection in medium containing one-fifth the usual concentration of methionine and 4 ~tci/ml [35S]methionine. Nucleocapsids were isolated from purified nuclei exactly as previously described (Stevely, 1975). Virions were isolated as follows. Infected cell pellets were suspended in 2 vol. 1 mm-phosphate buffer ph 7.4 and disrupted with four strokes of a Dounce homogenizer. The cytoplasm was then separated from the nuclei by centrifugation at 500 g for 10 min. Samples (2 ml) of the supernatant were layered on 36 ml sucrose density gradients (15 to 40~, w/w) made up in 1 mm-phosphate buffer. The gradients were centrifuged for 1 h at r.p.m, in the Beckman SW27 rotor. After centrifugation, the diffuse light-scattering band of virions was aspirated with a needle and syringe. For some experiments this virus suspension was diluted tenfold with 10 mm-tris-hc1 ph 7.2 and centrifuged at r.p.m, in the Beckman SW27 rotor to pellet the particles. These were resuspended in a small volume of 10 mm-tris-hcl for subsequent use. In other instances virus particles were further purified as previously described (Stevely, 1975). Briefly, the virus suspension from the sucrose density gradient was made 0.5 M with respect to urea and sonically treated for 5 s. This material was made 50% (w/w) with respect to sucrose by the addition of solid sucrose. Samples (10 ml) of this solution were placed in 38 ml tubes and discontinuous sucrose gradients were formed by the successive layering of 40, 30 and 20% (w/w) sucrose (made up in 10 mm-tris-hcl buffer) on top of the 50% sucrose-virus layer. These gradients were centrifuged at r.p.m, for 18 h in the Beckman SW27 rotor. The bulk of the virions floated to the 40 to 500/0 interface and were aspirated by needle and syringe. This virus suspension was diluted fourfold with 10 mm-tris-hcl buffer, layered over 10 ml of 10% sucrose in 2 M-urea, 10 mm-tris-hc1 buffer and centrifuged at r.p.m for 2 h in the SW27 rotor. The virus pellets were resuspended in a small volume of 10mM-Tris HCI buffer for subsequent use. We found no significant difference in the properties of the kinase associated with the virions of virus prepared directly from the rate sedimentation gradients and that from virus subjected to the further flotation purification. Extraction oj kinase Jkom virus particles. Virus particles were suspended in 10 mm-tris-hc1 ph 7.5. Aliquots (0.2 ml) containing 2 to 4 mg of protein were made to 10% with respect to NP40 (Sigma) and 0.6 M with respect to NaC1. The suspension was allowed to stand at room temperature for 30 min after which it was diluted tenfold either with 10 mm-tris-hcl ph 7-5 or with the same buffer containing 0-6 M-NaC1. This diluted material was centrifuged at g for 2 h. The supernatant was dialysed against kinase assay buffer and aliquots were stored at - 80 C. Protein kinase assay. The standard kinase assay buffer contained 50 mm-magnesium acetate, 0.1% NP40, 1 mmdithiothreitol and 50 mm-tris HCI ph 8-0. Each assay contained 1 to 5 ~tci ['y-3-'p]atp. The reaction volume was 200 ~tl and all reactions were carried out at 37 C for 30 min. Aliquots (usually 20 tal) were applied to 2.5 cm discs of Whatman No. 1 filter paper which were then washed four times successively in large volumes of 10% TCA saturated with sodium pyrophosphate, 10% TCA, ethanol and finally ether. In several experiments the reaction mixture was modified to test the effect of ph, divalent cations etc. Details are provided in the text where appropriate. Polyacrvlamide gel electrophoresis. The procedures were as described previously (McGrath & Stevely, 1980). Enzymes. Rabbit muscle type I and type II cyclic AMP-dependent protein kinases were purchased from Sigma. Casein kinases I and II were a gift from M. J. McGarvey and had been purified from Krebs II ascites cells (McGarvey & Leader, 1983). A cytoplasmic high-speed supernatant had been prepared and the kinases purified by column chromatography successively on DEAE-cellulose and phosphocellulose. An additional highly purified preparation (only a single protein contaminant) of casein kinase II purified from pig liver was a gift from Dr O.-G. Issinger, Institut ftir Humangenetik, University of the Saarland, Homburg, F.R.G. The same pattern of virus protein phosphorylation was observed for casein kinase II from both sources. BHK kinases were obtained by column chromatography of a high-speed supernatant on DEAE-cellulose essentially as for Krebs II ascites cells.

3 Pseudorabies virus protein kinase 663 Laser densitometry. Incorporation of 32p into individual proteins was estimated by laser densitometry of autoradiographs using an LK laser densitometer and LKB 2220 recording integrator. Column chromatograph), ofvirion extracts. Columns 1 cm in diameter were packed either with phosphoceltulose or with DEAE-cellulose to a height of 2 cm. Four mg of protein extracted from virions was applied to the columns in 1.5 ml of the appropriate starting buffer. Twenty-five ml of starting buffer was used to remove non-binding material from the column. This was followed by 100 ml of an appropriate salt gradient in the starting buffer. Columns were run at 15 ml/h and fractions of about 1-3 ml were collected. The conductivity of fractions was measured to allow calculation of the salt concentration. For the phosphocellulose chromatography the starting buffer was 50 mm-tris-hcl ph 7.5, 250 mm-naci, 1 mm-edta, 10 mm-2-mercaptoethanol, 10~ glycerol. The salt gradient was from 0.25 to 1.25 M-NaCI. For the DEAE-cellulose chromatography the starting buffer was 20 mm-tris HC1 ph 7.5, 1 mm-edta, 10 mm-2-mercaptoethanol, 10~o glycerol. The salt gradient was from 0 to 0-4 M-KCI. RESULTS Purity of virion preparations Since the conclusions in this report depend in part on the purity of the virion preparations, we have assessed this by a method that allows direct comparison both with our own earlier work (Stevely, 1975) and with that of Spear & Roizman (1972). The procedure for isolating virus particles differs from the methods used in these two previous reports only in the substitution of a sucrose density gradient for a dextran gradient in the initial stage. In these earlier studies the enrichment of viral protein with respect to host protein during the course of purification was estimated as follows. Uninfected cells were labelled using 14C-amino acids for 48 h; infected cells were labelled using 3H-amino acids from 5 to 24 h post-infection. The harvested cells were mixed and the ratio of viral 3H to cellular 14C counts showed a 20-fold enrichment in the final flotation-purified virus compared to the ratio in the initial homogenate. As discussed in these reports, this is the lower limit of the degree of purification of virion proteins with respect to cellular proteins (for example, not all 3H-labelled viral proteins are present in pure virus particles). The total enrichment was estimated to be of the order of 200-fold. We have carried out a similar analysis for this present study. Cells were labelled with 3H and with [35S]methionine. The 3H was added for the 48 h before infection. Shortly before adding virus the cells were washed several times with non-radioactive medium. From 4 to 24 h postinfection [3SS]methionine was added to the cells. The ratios of 35S : 3H counts were subsequently measured and found to be as follows : homogenate 0.4 : 1, virus pelleted and resuspended from the sucrose density gradient 6.5 : 1 and virus pelleted and resuspended from the final flotation gradient 16.5 : 1. These figures are in good agreement with those previously reported and indeed indicate a somewhat better overall enrichment. We concude, therefore, that our virion preparations are comparable in purity to those previously reported by ourselves and others. Properties of the endogenous phosphorylation of detergent-treated virions Virus was assayed for endogenous protein kinase activity. The activity was substantially dependent on the presence of the non-ionic detergent N P40 in the assay mixture (Table 1). This dependence was also found for the additional activity stimulated by the exogenous substrate casein (Table 1). A similar dependence is also found when protamine or histone is used as the added substrate (data not shown). (We find that for freshly harvested virus the activity is almost completely dependent on detergent but that following storage at - 20 C some activity is found in its absence.) Virus structural proteins phosphorylated in these assays are shown in Fig. 1 (a), which records an autoradiograph of virus particles subjected to electrophoresis following autophosphorylation. The proteins phosphorylated were the same as those phosphorylated in virus particles purified from cells labelled with 32p~ (Stevely, 1975). Longer exposure times revealed several additional phosphorylated structural proteins. Analysis of the phosphoamino acids of the total assay protein gave phosphoserine as the major product with some phosphothreonine also present (data not shown).

4 664 W. S. STEVELY AND OTHERS (a) 1 2 (b) ~ ~ ~:, ~ Fig, 1. (a) Autoradiogram of a 10 % polyacrylamide gel separation of PRV virion polypeptides following either growth in the presence of [3sS]methionine or kinase assay in the presence of [32p]ATP. Lane 1, [35S]methionine-labelled virions; lane 2, 32p-labelled virions after kinase assay under standard conditions. (b) Effect of cyclic AMP on PRV virion kinase activity. Autoradiogram of a 15~ polyacrylamide gel separation of virion polypeptides is shown. Lane 1, standard virion kinase assay; lane 2, virion kinase assay in the presence of 10 ~tm-cyclic AMP. Table 1. Effect of the addition of NP40 to the assay of protein kinase in PR V virions* 32p incorporated (c.p.m.) 32p incorporated (c.p.m.) in the presence of 5 ~tg/ml heparin - NP N P40 (0.1 ~) NP40 + casein (100 lag) NP40 + casein (100 lag) * Twenty-five ixg of protein as virions was added to each assay. The endogenous PRV activity was able to utilize GTP as well as ATP as the source of phosphate. Using 32p-labelled nucleoside triphosphates of known specific activity in assays containing the same amount of virion protein, 41.5 pmol of phosphorus per mg of protein was incorporated using ATP and 20.5 pmol per mg protein using GTP. Phosphorylation required Mg 2+. Of the other divalent metal ions tested, only Mn 2 showed any ability to support kinase activity (results not shown). When the concentration of Mg z+ ions in the assay was varied, the kinase showed a broad peak of activity between 50 mm and 150 mm (Fig. 2). This contrasts with the behaviour of kinase in the form of a soluble extract (see Fig. 2 and below).

5 Pseudorabies virus protein kinase lo d.~. 6 v 4 ~ 2-0 Y'. I I I I I I I I I I Magnesium acetate (mm) Fig. 2. Effect of Mg -' on PRV kinase activity. A, Normal assay of virion protein kinase; O, assay of the soluble extract. Standard conditions as defined in Methods were used and the same substrate proteins, present in the extract, were phosphorylated as in the virion assay. The activity showed a dependence on ph rather similar to that reported for HSV (Lemaster & Roizman, 1980) and the dependence on added KCI or NaC1 indicated maximum activity at 20 mm (results not shown). The kinase activity was, not surprisingly, found to be temperature-sensitive. When virions were kept at 60 C for various lengths of time before assay at 37 C, the kinase activity declined. After 15 min at 60 C, subsequent assay gave no incorporation of 32p above that of control blank assays. This decline in incorporation was not selective for any particular substrate protein: incorporation into all structural proteins was equally reduced (results not shown). This was of practical importance, as it allowed heat-inactivated virions to be used as a substrate when assaying the virion protein kinases in soluble extracts of the virus. Purified nucleocapsids show no endogenous protein kinase activity. Addition of cyclic AMP to the assay did not give a significant increase in the total incorporation of 32p. However, autoradiograms of the phosphorylated particles did show stimulation in the phosphorylation of a low molecular weight protein of about (Fig. 1 b). Analysis of autoradiograms by laser densitometry indicated that there was a fourfold stimulation of the phosphorylation of this protein in the presence of cyclic AMP. The major and mol. wt. phosphoproteins together accounted for approximately 86 ~ of the total incorporation in the absence of cyclic AMP and 81 ~o in its presence, while incorporation into the mol. wt. species rose from 2.5~ to 8~. Properties of soluble PR V protein kinase A number of procedures were used in attempts to isolate a soluble kinase from virus particles. It was found that the NP40 and salt treatment detailed in Methods gave reproducible results and reasonable recovery of activity. Using virus that had been labelled during infection with [3H]thymidine it was found that more than 90~o of the DNA remained in the final pellet. Measurements of protein indicated that 60 ~o of the protein was released from the particles. It is difficult to estimate enzyme yield. However, protamine was found to act as an exogenous substrate in the standard assay and could also function as a substrate for ~he extracted kinase. Comparison of protamine kinase activities indicated that approximately 50~ of the protamine kinase activity was present in the extract (data not shown). The residual pellet still autophosphorylated the same virus proteins as those phosphorylated by complete particles (data not shown). It seems likely that there is no selective extraction of one particular kinase activity by this procedure.

6 666 (a) 1 ], w. S. STEVELY AND OTHERS 3 (b) " O ii o, ; f 2 '2. i Fig. 3. Phosphorylation of heat-inactivated PRV virions by (a) soluble extract or (b) casein kinase II purified from Krebs 11 ascites cells. (a) Lane 1, complete assay including soluble extract; lane 2, complete assay including soluble extract but in the presence of 1 ~tg/ml heparin; lane 3, assay in the absence of soluble extract. (b) Lane 1, complete assay including casein kinase II; lane 2, complete assay including casein kinase II but in the presence of I /ag/ml heparin ; lane 3, assay in the absence of heatinactivated virions; lane 4, assay in the absence of casein kinase II. Table 2. Protein kinase activity of extracts of PRV and HSV 3zp incorporated (c.p.m.) ~k Added substrate PRV extract* HSV extract* None None but in the presence of 0-2 ~tg/ml heparin Protamine (100 ~g) Histone (100 lag) Casein (100 ~tg) Heat-inactivated homologous virus (20 ~tg) * Twenty ~tg of extract protein was added to each assay. The extract was found to have properties similar in many respects to the kinase associated with the particle. However, it is of interest that the rather high Mg 2+ dependence found for the activity of complete particles was not characteristic of the soluble extract (Fig. 2). A similar finding has been noted for HSV (Lemaster & Roizman, 1980). In that study Lemaster & Roizman cited unpublished data showing that preliminary studies with solubilized enzyme gave an Mg 2 optimum near to 5 mm and suggested that resistance to high Mg 2+ concentrations is related to the environment of the kinase within particles. Release of the kinase in soluble form

7 Pseudorabies virus protein kinase 667 I I I I I I 1 12 /.f I, )< ~,t"i" ~'/'t O-2 _~ >_ 0 < 462 I ~..',, I "'~ """"1... i"" I I I Fraction number Fig. 4. DEAE-cellulose chromatography of a soluble extract of PRV virions. Fractions were eluted at increasing concentrations of KCI (--) and were assayed for activity against protamine ( ), casein (---) and heat-inactivated virions (...). therefore results in the more normal Mg 2+ dependence profile. They have proposed that such behaviour could reflect changes in the juxtaposition of substrate proteins to the insoluble enzyme rather than the intrinsic properties of the enzyme. Accurate assessment of the properties of the kinase is thus more likely to be obtained using extracted enzyme. The extracted kinase was found to phosphorylate several different added substrates (Table 2). The phosphorylation of heat-inactivated particles by the extract was similar in terms of proteins phosphorylated and the extent of phosphorylation to that of endogenous phosphorylation and was found to be sensitive to heparin (Fig. 3a). Fig. 3(a) shows that the mol. wt., the mol. wt. and the mol. wt. species were all phosphorylated by the extract (lane 1) and that their phosphorylation was sensitive to heparin (lane 2), thus indicating the action of casein kinase II (see below). Lane 3 shows that the phosphorylation in lane 2 was not due to residual activity in the heat-treated virions. Partial purification of the kinase Since the extract was able to phosphorylate casein, histone and protamine it seemed likely that it contained more than one kinase activity. Accordingly, extract was applied either to DEAE-cellulose (Fig. 4) or to phosphocellulose (Fig. 5). Fractions from these columns were tested for activity against casein, histone, protamine and inactive PRV. For the heat-inactivated virus, electrophoresis and autoradiography were carried out. It is clear from these experiments that the extract contained several kinase activities and that the major phosphorylation of virus structural protein was by an activity which also phosphorylated casein, could use GTP as a substrate and which was inhibited by heparin (Fig. 6). These properties are consistent with those of casein kinase II, as were the migration characteristics of the activity on the two chromatography media (Hathaway & Traugh, 1982). We have not as yet fully identified the other kinase activities, but at least two of these appear to resemble cellular enzymes. The heparin-resistant peak of casein kinase activity in Fig. 6 clearly resembles casein kinase I, and the first peak of protamine kinase activity in Fig. 5 which catalyses the Ca 2+- and phospholipid-dependent phosphorylation of histone H1 (M. Katan, unpublished results) has the properties of protein kinase C (Takai et al., 1979). We have carried out an extensive analysis of the protein kinase activities present in post-ribosomal high-speed

8 668 W. S. STEVELY AND OTHERS ~... I I I I I I " 1.2 x U o. Ii u -.11 I Ill 'Ill 6JIIi 0.8 g ~0.4 ~ z 0.0,,,,,, Fraction number Fig. 5. Phosphocellulose chromatography of a soluble extract of PRV virions. Fractions were eluted at increasing concentrations of NaC1 (--, upper curve) and were assayed for activity against protamine ( ), histone (---) and casein (---, lower curve), -, r ' 0.8 ~ I'.\... I 7o.6 x "" 0.5 d 0.3 "r. =~ t ' i Fraction number Fig. 6. Substrate specific ties of fractions eluted by NaC1 (...) from phosphocellulose. Fractions from the separation shown in Fig. 5 were further assayed. 0, Activity against casein using ATP;, activity against casein using GTP : i, activity against casein using ATP but in the presence of I pg/ml heparin. The lower panel shows the phosphorylation of heat-inactivated virions by the fractions, i i supernatants from normal and infected cells (M. Katan, unpublished results). These studies have confirmed the identities of the kinases discussed here and have enabled us to estimate total casein kinase activities during infection.

9 Pseudorabies virus protein kinase 669 x,,,o e- L~ =_ t t n i i i u [-400 (a) Casein kinase II / lo ~ ~ - ~ ' ~ }-200 ~ = - Casein l/'\...f kinase I,4 \ ". 0 "T i i I... i (b) Fraction number ~O~o...,...O~o/ \ I 80 L0 0 I I i I I I Time after viral infection (h) Fig. 7. Casein kinase II activity during PRV infection. (a) DEAE-cellulose chromatography of the casein kinase activities of a high-speed post-ribosomal fraction from uninfected cells. The column was eluted with an increasing concentration of KCI ( ) and assayed for activity against casein (0). (b) Variation in casein kinase II activity during infection, Equivalent amounts of protein from cells infected for different lengths of time were fractionated as in (a). The total activity in the hatched area was estimated for each sample and used as a measure of casein kinase II. Casein kinase H activity in infected cells We have examined the total casein kinase II activity in infected cells. A post-ribosomal highspeed supernatant from cells at various times of infection was subjected to DEAE-cellulose chromatography in a manner similar to that described for virion extracts. The same amount of protein was applied to the column for each time point. Fig. 7 (a) shows the profile of casein kinase I and II activities observed in uninfected cells. Because of overlap with casein kinase I, the sum of the activity in the three peak fractions (indicated by the hatched area) was used for quantitative comparison of cellular casein kinase II activity. This is shown for various times of infection in Fig. 7(b) and indicates that there was no change in total activity of the enzyme during the first 10 h of infection, by which time the virus growth cycle is virtually complete (Tyler et al., 1974). Phosphorvlation of PR V by exogenous kinases Heat-inactivated PRV particles (treated for 15 min at 60 C) were used as substrate for a variety of kinases. Purified type I and type II cyclic AMP-dependent protein kinases from rabbit muscle, purified casein kinases I and II from Krebs II ascites cells, and kinases purified from normal and infected BHK cells were all employed. Assays were carried out using these various kinases and autoradiographs of the proteins phosphorylated in each case were examined. Not all of these are shown. Only casein kinase II or mixed kinase fractions known to contain this activity in addition to others phosphorylated the major viral phosphoproteins (Fig. 3 and 8). This phosphorylation was similar in terms of the number of proteins phosphorylated and the extent of phosphorylation to that which we observed using our standard endogenous kinase assay (Fig. 8). The one major exception to this was the protein of approx mol, wt. to which we referred above. This protein was only a minor phosphorylated species in the

10 670 W. S. STEVELY AND OTHERS ~ L LL ~.... U F Fig. 8. Phosphorylation of PRV by added kinase. Autoradiogram of a 15% polyacrylamide gel separation of phosphorylated polypeptides is shown Lane 1, assay of active PRV virions: lane 2, heatinactivated virions phosphorylated by cyclic AMP-dependent protein kinase type II; lane 3, heatinactivated virions phosphorylated by casein kinase II ; lane 4, autophosphorylation of heat-inactivated virions; lane 5, casein kinase II autophosphorylation; lane 6, cyclic AMP-dependent protein kinase type II autophosphorylation. Virion polypeptides phosphorylated by added kinase are indicated to the left of the appropriate lanes. endogenous assay but was strongly phosphorylated by type I and type II cyclic AMP-dependent kinases from rabbit muscle. In these figures we have included lanes which show the autophosphorylation of the kinase preparations. These phosphorylations were, of course, of polypeptides present as contaminants in the enzyme preparations. The additional phosphorylation of virion proteins is clearly distinguishable from these and is indicated at the side of the appropriate lanes. Densitometry of lane 2 of Fig. 8 and comparison of this with lane 5 showed that the autophosphorylation of the cyclic AMP-dependent kinase accounted for 40 / of the incorporation seen in lane 2. The incorporation into viral proteins in this lane was principally into the mol. wt. species. This accounted for 6 0 ~ of the viral protein phosphorylation. Some incorporation into other viral proteins did occur but to a limited extent, no single species having more than 1 0 ~ of the incorporation found in the tool. wt. band. This contrasts with the phosphorylation by casein kinase II shown in lane 3. Comparing lane 3 with the autophosphorylation of the casein kinase II preparation shown in lane 4 indicates that

11 P s e u d o r a b i e s virus p r o t e i n k i n a s e : ~ i ~ ~..~? Fig. 9. Phosphorylation of HSV by added kinase. Autoradiogram of a 15~'~ polyacrylamide gel separation of phosphorylated polypeptides is shown. Lanes 1 and 2, assay of active HSV virions; lane 3, autophosphorylation of heat-inactivated virions; lane 4, heat-inactivated virions phosphorylated by cyclic AMP-dependent protein kinase type II; lane 5, heat-inactivated virions phosphorylated by casein kinase II; lane 6, casein kinase II autophosphorylation; lane 7, cyclic AMP-dependent protein kinase type II autophosphorylation. Virion polypeptides phosphorylated by the added kinase are indicated to the left of the appropriate lanes. incorporation into viral proteins is almost entirely into the and mol. wt. proteins. For this enzyme the overall pattern of viral protein phosphorylation was similar to that in lane 1 which illustrates the autophosphorylation of active virus particles in the absence of any added kinase. Lane 4 proves that the incorporation into viral proteins seen in lanes 2 and 3 was not due to residual activity in the heat-treated virions. Casein kinase II is known to be especially sensitive to inhibition by heparin which has no effect at the levels used on casein kinase I, the type I and II c A M P - d e p e n d e n t protein kinases, protease-activated kinase I, and the hemin-controlled repressor ( H a t h a w a y et al., 1980). The phosphorylation of heat-inactivated virus particles by casein kinase II was inhibited by heparin (Fig. 3). Inhibition by heparin was also found for endogenous virion phosphorylation and phosphorylation of added casein (Table 1). We also found that purified nucleocapsids did not act as substrate for either the extracted kinase from virions or any of the other kinases we have tested above (results not shown). C o m p a r i s o n with H S V Experiments comparing the properties of HSV protein kinase with those described here for P R V have been carried out. The pattern of phosphorylation in the endogenous assay was similar to that reported by Lemaster & Roizman (1980). The two most prominent bands (Fig. 9, lane 1)

12 672 W. S. STEVELY AND OTHERS were of approximate molecular weights and and correspond to virion proteins 12 and 23 respectively. In addition, other proteins phosphorylated to a lesser degree included virion proteins l, 4 and 14. HSV inactivated in the same manner as described for PRV was phosphorylated by added casein kinase II. However, again at least one HSV protein (virion protein 14) was phosphorylated by cyclic AMP-dependent histone kinase. A second (virion protein 12) could be phosphorylated either by exogenous casein kinase II or exogenous cyclic AMP-dependent kinase (Fig. 9). Given the different substrate specificities of these enzymes, this may reflect the phosphorylation of separate sites on the protein. The HSV phosphorylation was sensitive to heparin and an HSV extract prepared like that of PRV was able to utilize the same range of substrates as PRV (Table 2). Preliminary experiments indicated a chromatographic profile for an HSV extract similar to those shown for PRV. In addition, we found that an extract from PRV was able to phosphorylate inactivated HSV and an extract from HSV was able to phosphorylate inactivated PRV (data not shown). DISCUSSION The experiments described indicate that for PRV the kinase responsible for the phosphorylation of the major structural substrates of and tool. wt. is likely to be located in the tegument of the virus. The response to added NP40 both for endogenous phosphorylation and also for that of added casein makes it unlikely that a kinase loosely adsorbed to the membrane surface is responsible. In addition, the activity remains with particles through both rate sedimentation and flotation centrifugation followed by pelleting and resuspension. The discrepancies between reports of the ability of kinase associated with virus to utilize exogenous substrates (see Introduction) seem likely to be the result of difficulties in the penetration of the substrates to the kinase sites in the particles. We believe that the change in Mg 2+ dependence between particle-associated kinase and extracted kinase points in that direction. We therefore expect that studies of extracted kinase activities from other herpesviruses will also show that they can phosphorylate substrates other than virus structural proteins. Given this, there are two major conclusions to be drawn from the results reported. First, it is clear that PRV (and also HSV) contains not just a single kinase activity but several, and second, that the activity responsible for the phosphorylation of the major phosphorylated species of virus structural proteins is analogous to host casein kinase II. That virus particles contain more than one protein kinase activity is clearly established by their differing substrate specificities and chromatographic elution patterns. On DEAE-cellulose (Fig. 5), a protamine kinase activity with no casein kinase activity eluted at approximately 0.08 M-KCI whereas on phosphocellulose (Fig. 6 and 7), casein kinase activities with no contaminating protamine kinase activity eluted between 0.4 and 0.65 M-NaCl. That at least two casein kinase activities are present is shown by the sensitivity of only one part of the activity to heparin and its ability to use GTP as well as ATP. As already discussed, these properties have allowed us to identify this as casein kinase II. Since we found no increase in total cellular casein kinase II activity in the course of infection we have every reason to assume that the enzyme in the virus particle is a host enzyme. This similarity to the host enzyme lies in its use of GTP as a substrate, its inhibition by heparin and its chromatographic behaviour on DEAE-cellulose and on phosphocellulose. Again, the fact that authentic casein kinase II phosphorylates the same structural proteins in heat-inactivated virus particles is evidence in favour of the host enzyme being implicated. These results are not inconsistent with any previous reports of herpesvirus particle kinase activities (Rubenstein et al., 1972; Randall et al., 1972; Tan, 1975; Lemaster & Roizman, 1980; Flugel & Darai, 1982) but represent a considerable extension of previous work. That more than one kinase activity should be found in an enveloped virus particle is not surprising given previous reports for Rous sarcoma virus (Weis & Faras, 1983) and avian myeloblastosis virus (Rosok & Watson, 1979). In the report by Tan (1975), comparison of Semliki Forest virus, Sindbis virus, Sendai virus, vaccinia virus, rabies virus, vesicular stomatitis virus and pseudorabies virus showed that in most cases there was kinase activity against several different

13 Pseudorabies virus protein kinase 673 added substrates, again indicating the likely presence of more than one kinase. That some or all of these activities might be of host origin and perhaps related to membrane activities would also be possible given the reported specific incorporation of host cell surface proteins into budding vesicular stomatitis virus particles (Lodish & Porter, 1980). The absolute amounts of these enzymes in the virions are likely to be small and might well have escaped detection as host 'contaminants' of purified virus in previous studies of virion proteins. It is of interest that casein kinase II which we have implicated in virus protein phosphorylation can be detected in nuclei, membranes, ribosomes and mitochondria as well as in post-ribosomal cytoplasmic supernatants of normal cells (Hathaway & Traugh, 1982). The authors are grateful to Mr D. Mease for technical assistance, to Dr M. J. McGarvey for the gift of casein kinases I and II, and to Dr O.-G. Issinger for the gift of casein kinase II. Drs D. P. Leader and W. S. Stevely are grateful to the Medical Research Council for the award of a project grant. REFERENCES CHANTLER, J. K. & STEVELY, W. S. (1973). Virus-induced proteins in pseudorabies-infeeted cells. I. Acid-extractable proteins of the nucleus. Journal of Virology 11, FLUGEL, R. M. & DARAI, G. (1982). Protein kinase and specific phosphate acceptor proteins associated with tupaia herpes virus. Journal of Virology 43, HATHAWAY, G. M. & TRAUGH, J. A. (1982). Casein kinases - multipotential protein kinases. Current Topics in Celt... Regulation 21, HATHAWAY, G. M., LUBBEN, T. H. & TRAUGH, J. A. (1980). Inhibition of casein kinase II by heparin. Journal of Biological Chemisto' 255, KAPLAN, A. S. & VATTER, A. E. (1959). A comparison of herpes simplex and pseudorabies viruses. Virology 7, 39~ 407. LEMASTER, S, & ROIZMAN, n. (1980). Herpes simplex virus phosphoproteins. II. Characterisation of the virion protein kinase and of the polypeptides phosphorylated in the virion. Journal of Virology 35, LODISH, H. F. & PORTER, M. (1980). Specific incorporation of host cell surface proteins into budding vesicular stomatitis virus particles. Cell 19, McGARVEY, M. J. & LEADER, D. P. (1983). The ribosomal proteins phosphorylated in vitro by protein kinase activities from Krebs II ascites cells. Bioscience Reports 3, McGRAa-n, B. M. & S'rEVELY, W. S. (1980). The characteristics of the cell-free translation of mrna from cells infected with the herpes virus pseudorabies virus. Journal of General Virology 49, RANDALL, C. C., ROGERS, H. W., DOWNER, D. N. & GENTRY, G. A. (1972). Protein kinase activity in equine herpes virus, Journal of Virology ROSOK, M. J. & WATSON, K. F. (1979). Fractionation of two protein kinases from avian myeloblastosis virus and characterisation of the protein kinase activity preferring basic phosphoacceptor proteins. Journal of Virology 29, RUBENSTEIN, A. S., GRAVELL, M. & DARLINGTON, R. (1972). Protein kinase in enveloped herpes simplex virions. Virology 50, SPEAR, P. G. & ROIZMAN, 8. (1972). Proteins specified by herpes simplex virus. V. Purification and structural proteins of the herpes virion. Journal oj Virology 9, STEVELY, W. S. (1975). Virus-induced proteins in pseudorabies-infected cells. II. Proteins of the virion and nucleocapsid. Journal o[ Virology 16, TAKAI, Y., KISHIMOTO, A., IWASA, Y., KAWAHARA, Y., MORI, T. & NISHIZUKA, Y. (1979). Calcium-dependent activation of a multifunctional protein kinase by membrane phospholipids. Journal of Biological Chemistry 254, TAN, K. 8. (1975). Comparative study of the protein kinase associated with animal viruses. Virology 64, TYLER, S. J., HAY, J. & STEVELY, W. S. (1974). Nuclear protein synthesis in hamster kidney cells: the effect of pseudorabies virus infection. Arehiv J'ftr die gesamte Virusforschung 44, WEIS, J. H. & FARAS, A. J. (1983). Purification and characterisation of two protein kinases associated with Rous sarcoma virus. Biochemistry 22, (Received 18 September 1984)