The Major Ribonucleoprotein-associated Protein Kinase of Vesicular Stomatitis Virus Is a Host Cell Protein*

Ribonucleoprotein particles (RNPs) of vesicular stomatitis virus (VSV) were fractionated by column chro- matography through Fractogel TSK HW-55F and by centrifugation through KC1 sucrose. Analyses of frac- tions for protein content and for protein kinase activity indicated that the major peak of kinase activity did not correspond exactly with any of the VSV-specific proteins. Neither anti-NS nor anti” IgG preparations inhibited protein kinase activity, and IgG did not act as an exogenous phosphate acceptor. Reconstitution of an RNP-enzyme complex did not result in a restoration of protein kinase activity. In vitro translation of VSV- specific poly(A)-containing RNA did not result in any detectable production of kinase activity. Thus, the major RNP-associated kinase is a host cell protein which is tightly bound to the RNP particle. ~~ VSV’ virion-associated host-deriveil vitro VSV-associated kinase phosphorylates and M and extent proteins and G, as nonviral VSV-associated protein 39) a functional relationship between protein kinase and direct correlation between phosphorylation and transcription contrast,

Ribonucleoprotein particles (RNPs) of vesicular stomatitis virus (VSV) were fractionated by column chromatography through Fractogel TSK HW-55F and by centrifugation through KC1 sucrose. Analyses of fractions for protein content and for protein kinase activity indicated that the major peak of kinase activity did not correspond exactly with any of the VSV-specific proteins. Neither anti-NS nor anti" IgG preparations inhibited protein kinase activity, and IgG did not act as an exogenous phosphate acceptor. Reconstitution of an RNP-enzyme complex did not result in a restoration of protein kinase activity. In vitro translation of VSVspecific poly(A)-containing RNA did not result in any detectable production of kinase activity. Thus, the major RNP-associated kinase is a host cell protein which is tightly bound to the RNP particle. ~~ VSV' has a virion-associated protein kinase activity which is CAMP independent and has been suggested to be a hostderiveil protein (6,7,24). In an in vitro reaction, the VSVassociated kinase phosphorylates mainly the virion proteins NS and M and to a lesser extent proteins L and G, as well as nonviral proteins (7,39). The function of the VSV-associated protein kinase is not known, but there is some evidence (5,20,39) suggesting that a functional relationship exists between protein kinase activity and transcription. However, no direct correlation between phosphorylation and transcription was established by these investigations. In contrast, other recent studies by  indicated that phosphorylation of ribonucleocapsid-associated proteins in vitro does not play a necessary role in VSV primary transcription.
An electron microscopic study of VSV virions following phosphorylation in vitro indicated a possible relationship between phosphorylation and virion uncoating (38).
Clinton et al. (6), using immunological techniques, tentatively identified one protein kinase associated with VSV as pp60 src. Other protein kinases are probably associated with VSV since the reaction products obtained from in vivo and in oitro reactions differ (7,15,24).
The above studies which have attempted to define the possible role(s) of the one or more kinases associated with the *This research was supported by National Institutes of Health Grant 5 ROI AI 12316 and National Science Foundation Grant PCM-81-10986. 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.
* The abbreviations used are: VSV, vesicular stomatitis virus; RNP, ribonucleoprotein; BSA, bovine serum albumin; DTT, dithiothreitol; RNP-N, ribonucleoprotein particle containing only N protein plus the RNA template; PAGE, polyacrylamide gel electrophoresis. VSV virion and RNP particle have only indirectly implicated this enzyme activity as a possible means of regulating the virus-specific RNA polymerase during replication. We still lack direct evidence for this function, and we do not know the number and location of the phosphate groups on the NS protein and the difference(s) in phosphorylation sites on active or inactive polymerase. It is clear that other negative strand viruses have virion-associated kinase activity and/or phosphoproteins as do viruses such as Rous sarcoma (4) and SV40 (36). In all cases our understanding of the role(s) of the phosphoproteins in regulation of virus replication is still unclear.
The purpose of this study was to develop methods for purifying the L/NS VSV RNA polymerase and RNP-N to decide if the RNP-associated kinase was a VSV protein.
Furthermore, the purified enzyme subunits from our strain of VSV Indiana would then be amenable to structural analyses such as the topology of the phosphopeptides in active versus inactive NS protein. Since published methods (9, 25) for enzyme purification failed to produce pure L/NS enzyme subunits or pure RNP-N template, we have established the purification procedure reported here. Our results indicate that the RNP-associated kinase is a host cell protein which is tightly bound to the RNP particle.

Celk and
Virus-HeLa S3 cells and VSV Indiana were grown as previously described (17) with the exception that 10% calf serum (Irvine Scientific) was used in the medium instead of fetal calf serum. Baby hamster kidney cells were grown as the HeLa cells except for the addition of 2% fetal calf serum and 10% tryptose phosphate broth (29.5 g/liter). Purified virus preparations were stored at -70 "C in 10% dimethyl sulfoxide.
Isolation of RNPs from Virions-RNPs were isolated by mixing virions with an equal volume of lysis solution (3.8% Triton N 101, 0.8 M NaCI, 1.2 mM DTT) and subsequent centrifugation through a discontinuous glycerol gradient (12). The pellet of RNPs was resuspended in 10 mM Tris, pH 8.0 containing 5 mM Dm.
Fractionation of RNPs-The RNPs, which contained the proteins L, NS, N, and some M, were made 2 M CsCl and 10 mM Tris, pH 7.4, and were incubated for 20 min at 20 "C. The suspension of RNP proteins and molecular weight marker proteins were layered onto a column (20 X 540 mm) containing Fractogel TSK HW-55F (EM Reagents) which had been equilibrated with 1 M KCI, 10 m M Tris, pH 8.0, 1 m M DTT, and 15% sucrose. The column was eluted at 20 "C at a flow rate of 0.8 ml/min. Fractions of 1 ml were collected and stored a t 4 "C until assayed.
Protein peaks, identified by Coomassie blue or silver nitrate staining of the proteins separated in polyacrylamide gels (23), were dialyzed against 10 mM Tris, pH 8.0, 1 mM DTT, and concentrated by placing the dialysis tubing in crystalline sucrose until the volume was reduced by the desired amount.
The L and NS proteins, which eluted from the column together as a complex, were further separated by layering the dialyzed proteins over a 5-25% sucrose gradient containing 1 M KC1, 10 mM Tris, pH 8.0,l mM DTT, with a 0.2-ml cushion of 80% sucrose and centrifuging 15283 by guest on March 24, 2020 http://www.jbc.org/ Downloaded from at 35,000 rpm for 20 h at 4 "C using an SW 60 rotor. Nineteen fractions were collected from each gradient and were assayed for protein content by polyacrylamide gel electrophoresis and silver nitrate staining and for protein kinase activity.
The Fractogel peak fraction which corresponded to partially purified RNPs was purified further by isopycnic centrifugation. The fractions were combined and centrifuged through either 26-33% CsCl or 15-47% Renografin at 35,000 rpm for 18 h 4 "C using a n SW 41 rotor (32). The band of RNPs was collected, diluted with 10 mM Tris, pH 8.0, and centrifuged at 40,000 rpm for 2 h at 4 'C using an SW 41 rotor. The pellet of RNPs was resuspended in 10 mM Tris, pH 8.0, containing 5 mM DTT, and stored at -70 "C.
Protein Kinase Assay-Protein kinase activity was assayed according to previously described methods (24,39). In addition to aliquots of column or gradient fractions, the reaction mixture (100 pl) contained 10 mM Tris, pH 8.0, 10 mM MgCl,, 5 mM DTT, 1 X M ATP, 6 pCi of [y3'P]ATP (New England Nuclear; specific activity, 8.7 Ci/mmol), and 20 pg of casein as an exogenous phosphate acceptor (8). After incubation of the protein kinase reaction mixtures at 37 "C for 45 min, aliquots were precipitated with 10% trichloroacetic acid containing 50 mM monosodium phosphate and 50 mM sodium pyrophosphate, precipitates were collected on GF/C Whatman filters, and radioactivity was determined in a Beckman liquid scintillation spectrometer. The remainder of the reaction mixture was mixed with gel sample buffer (20% glycerol, 2% DTT, 6% SDS, 0.125 M Tris, pH 6.8, and 0.004% bromphenol blue) and boiled for 1 min.
Samples were applied to discontinuous Tris-glycine-buffered SDSpolyacrylamide gels consisting of a 10% resolving gel and a 4% stacking gel (12). Gels were dried and exposed to Kodak SB-5 film at Protein Quantitation-Proteins were quantified by densitometric scanning of silver nitrate stained bands in a polyacrylamide slab gel, or protein concentration were determined colorimetrically with Bio-Rad protein dye reagent measuring absorbance at 595 nm, using ovalbumin as a standard.
Reconstitution of Purified L and NS Protein with the RNP-N Template-Purified L and NS proteins from sucrose gradients were mixed with RNP-N template in quantities which were determined to be saturating for bound L/NS enzyme.' The protein kinase activity of this reconstituted RNP particle was assayed as described above using casein as an exogenous phosphate acceptor. The RNP-N template which was used for these experiments was isolated by isopycnic centrifugation of intracellular RNPs from VSV-infected cells in Renografin and was shown by silver staining to be free of all proteins except N.
Treatment of Protein Preparations Exhibiting Protein Kinase Activity with Anti-VSV Protein IgG Fractions-To determine the effect of specific anti-VSV protein IgG preparations on RNP-associated protein kinase activity, column fractions containing protein kinase activity were also assayed in the presence of purified IgG. The conditions of preincubation of fractions with IgG preparations were the same as those described previously (12).
Purification ofRNA-Baby hamster kidney cells infected with VSV at a multiplicity of infection of 8 plaque-forming units/cell were harvested at 5 h postinfection as described previously (30). The infected cytoplasmic extract was centrifuged through a 20 to 40% CsCl gradient, and the RNA pellet containing VSV messenger RNA was resuspended in 0.1 M NaCl, 10 mM Tris, pH 7.4, 1 mM EDTA, 0.05% SDS, and precipitated with ethanol (32). Poly(A)-containing mRNA was isolated on an oligo(dT)-cellulose column (27) and precipitated with ethanol.
Polylysine Affinity Chromatography-A reticulocyte lysate reaction mix, with or without added VSV mRNA, was chromatographed on a column (7.0 X 55 mm) of agarose poly-L-lysine (P-L Biochemicals). The column fractions were eluted with 10 ml of 0.1 M KCl, 10 mM Tris, pH 7.4, then 10 ml of 0.5 M KCI, 10 mM Tris, pH 7.4, and finally a 10-ml gradient of 0.7-1.5 M KC1,lO mM Tris, pH 7.4. The fractions were dialyzed overnight against 10 mM Tris, pH 8.0,l mM DTT, and a sample was assayed for protein kinase activity using casein as the phosphate acceptor, as described above.

Fractionation of VSV RNPs by Fractogel Column
Chromatography-Virion RNPs were incubated in the presence of 2 M CsCl and chromatographed through a column of Fractogel TSK HW55F. Fractions from the Fractogel column were analyzed in SDS-PAGE gels. Fig. lA shows a silver stain of this SDS-PAGE gel, indicating the elution order of the virionspecific proteins from the high salt-stripped virion RNPs. The first protein peak eluted from the column in the void volume represented partially purified RNPs which contained N protein (RNP-N) and trace amounts of L, NS, and M proteins (fractions 20-24). The next peak eluted (fractions 40-46) contained L and NS proteins and was assumed to be a complex of L and NS due to the coelution of these two proteins. This peak was followed by BSA in fractions 42-48, a contaminant in the BSA protein solution in fractions 46-50, and finally, protein M in fractions 52-58. There was no free L or NS observed despite the presence of 1 M KC1 and DTT in the column buffer. In another experiment (data not shown) myosin (Mr = 200,000) and BSA were added to the suspension of RNPs, and the mixture was eluted from the Fractogel column. Myosin eluted after the RNP-N peak and prior to the L/NS complex, and BSA eluted as described above.
These results showed that even after treatment of the RNPs with 2 M CsCl and 1 M KC1, some L and NS proteins remain attached to the template (fractions 20-24), and, in addition, the complex of L and NS was not dissociated under these salt conditions. Based on the molecular weights of myosin, BSA, and M protein, the apparent molecular weight of the L/NS complex was 75,000. This was somewhat unexpected since the M , of L protein from PAGE analysis is approximately 180,000, and the M, of NS protein, derived from the cDNA sequence (28), is 25,000. If the L/NS complex were at a minimum L,NS, as was predicted for the active complex (25), the apparent M, of 75,000 on the Fractogel column was quite low. This low M , was obviously not due to degradation, since PAGE analysis of this peak revealed intact L and NS proteins. Densitometric scans of silver nitrate-stained gels of the L/NS complex indicated a purity of greater than 95% (data not shown).
Protein kinase activities of the Fractogel column fractions, shown in Fig. lA, were assayed using casein as an exogenous phosphate acceptor. Two major and one minor peak of protein kinase activity were observed (Fig. 1B). The first corresponded to the RNP-N template. The second small peak around fraction 42 corresponded to the L/NS complex as determined by a silver-stained polyacrylamide gel. The third and largest peak of protein kinase activity did not correspond precisely with any of the peaks of VSV-specific proteins; the peak of this activity eluted after the L/NS complex and slightly before the peak of M protein. There were no other detectable silver-stained proteins visible in this region.
The relative heights of the peaks of kinase activity varied somewhat from experiment to experiment and depended on the initial amount of RNPs used as starting material. When the RNP concentration was very dilute, stripping of the template appeared to be more efficient and the peak of kinase activity associated with the RNP-N template decreased with a concomitant increase in the third peak of kinase activity.
Effect of Monospecific Anti-VSV Protein IgG Preparations on the RNP-associated Protein Kinase Activity-Since the Fractogel peak fraction which gave maximum protein kinase activity (see Fig. 1B  of anti" IgG and anti-NS IgG were mixed with a constant amount of protein from the Fractogel peak; and protein kinase activity was assayed in duplicate reactions except that casein was omitted from one set of reactions. Fig. 2A shows the Coomassie blue-stained polyacrylamide gel of the kinase reactions, and Fig. 2B shows the autoradiogram of the products on the same gel. In reactions where casein was omitted, the protein kinase activity was not above background since IgG did not act as an exogenous phosphate acceptor in this reaction. In no case did either anti" IgG or anti-NS IgG have any effect on the RNP-associatedprotein kinase activity when casein was used as an exogenous phosphate acceptor. These results strongly support the notion that the major protein kinase activity is a host protein which is present a t as an exogenous phosphate acceptor. Duplicate samples from in vitro reactions were trichloroacetic acid-precipitated, filtered, and radioactivity was determined by a liquid scintillation spectrometer. 180K, for example, 180,000.

Effect of Isopycnic Centrifugation of Partially Purified RNPs in
CsCl or in Renografin-The Fractogel peak fraction which corresponded to partially purified RNPs and which contained significant protein kinase activity was centrifuged either in CsCl (3 successive gradients) or in Renografin. Virion RNPs, purified by either procedure, still contained small amounts of L and NS in addition to N protein but showed no protein kinase activity above backgrounds (data not shown). Therefore, the kinase activity associated with the RNP fraction from the Fractogel column was dissociated with further purification and likely represents incomplete stripping of contaminating proteins by the high salt treatment. This interpretation is confirmed by the presence of small amounts of L, NS, and M proteins in the preparation (Fig. 1A) and the variability levels too low to be detected by the silver-staining technique. of the amount of kinase activity recorded as a function of The order of the lanes in A and B was reversed through photography. 18OK, for example, 180,000.

RNP concentration.
containing the L and NS complex were combined, dialyzed, Separation of L and NS Proteins-Since small amounts of and concentrated. They were then applied to a 5-2576 sucrose kinase activity were associated with the NS and L complex, gradient containing 1 M KCI, 10 mM Tris, pH 8.0, and 1 mM this complex was dissociated and the separate proteins ana-DTT. After centrifugation, fractions were collected, and each lyzed for kinase activity. The fractions from the column fraction was analyzed for protein content and for protein by guest on March 24, 2020 http://www.jbc.org/ Downloaded from kinase activity. Fig. 3A shows a silver nitrate stain of an SDSpolyacrylamide gel of the sucrose gradient fractions and demonstrates the complete separation of the L (fractions 1-4) and N S (fractions 9-12) proteins. Each of these fractions was then analyzed for protein kinase activity; Fig. 3B shows that the major protein kinase activity was found in fractions 10-15, as determined using casein as an exogenous phosphate acceptor. This latter broad peak of protein kinase activity did not correspond precisely with the peak of NS protein; the maximum kinase activity was always found in fractions representative of material of smaller size in SDS gels than NS protein.
No other silver-stained bands were visible in the region of this peak from the gradient.
A small amount of protein kinase activity was also found in fractions 1-4 which contained L protein. This kinase activity represented only 1-2% of the total kinase activity of the fractions from Fractogel chromatography. It may represent protein aggregates which sediment with L protein.
Since it was such a small percentage of the total activity, no attempt has been made to characterize it further.

The Effect of Reconstitution on Protein Kinase Actiuity-
Since the capping, methylation, transcription, and polyadenylation reactions carried out by the L/NS enzyme require RNP-N template (IO), it was conceivable that there also is a VSV-specific kinase activity which requires the presence of A L/NS enzyme and RNP-N template to function. To test this hypothesis, a reconstitution experiment was performed. Table I shows the protein kinase activities of purified L and NS proteins and mixtures of purified L and NS proteins combined with RNP-N template. These activities were compared with the protein kinase activity of detergent-solubilized VSV using an amount of VSV which was similar in protein content to the purified preparations of the proteins. All mixtures containing NS protein possessed slight protein kinase activity; however, this activity did not increase when NS and L protein were mixed with RNP-N template using conditions which resulted in reassociation of these proteins with the template,2 indicating that formation of an RNP-enzyme complex did not enhance kinase activity.

In Vitro Translation of VSV Poly(A)-containing RNAs-In
order to further test whether the VSV-associated kinase activity was a VSV protein or a host protein, VSV-specific poly(A)-containing RNA was purified from infected baby hamster kidney cells and was then translated in a rabbit reticulocyte lysate mixture containing 164 pg of RNA. We reasoned that the poly(A)-containing VSV mRNA, which would direct the in vitro synthesis of at least microgram amounts of VSV proteins in the reticulocyte lysate, might be detectable as an increase in the kinase activity of the lysate if one of these four VSV proteins were a protein kinase.   I Protein kinase actioities of purified proteins RNP-N was purified by banding RNPs from VSV-infected cells in Renografin as described in the text. NS and L proteins were prepared by Fractogel column chromatography and 1 M KC1 sucrose centrifugation as described in the text. Protein kinase activities were determined by using casein as an exogenous phosphate acceptor and 0.4 pg of RNP-N, 0.6 pg of NS protein, and 0.6 pg of L protein per reaction. After incubation for 45 min at 37 "C, duplicate samples were trichloroacetic acid-precipitated, filtered, and radioactivity determined. Activity of each sample was compared with solubilized VSV virions of similar protein content. column was used, therefore, as a crude means of separating proteins in the reticulocyte lysate. Either reticulocyte lysate alone (Fig. 5A) or lysate containing translated VSV protein (Fig. 5B) was fractionated on agarose poly-L-lysine; the pooled fractions were dialyzed and tested for kinase activity using casein as a phosphate acceptor. There was no increase in total activity in the lysate that contained in uitro translated VSV proteins, and the fractions that were enriched for VSV proteins showed less activity. The products of these reactions were also resolved on SDS-polyacrylamide gels, and there were no detectable differences in the phosphorylated proteins from the fractions shown in Fig. 5, A and B (data not shown).
Thus, there was no evidence of a VSV-specific kinase activity among the in uitro translated VSV proteins.

DISCUSSION
The RNP of VSV is a large enzyme-template complex and has been shown previously to be involved in transcription (10, E ) , in the synthesis and addition of the cap group and methylation of each of the VSV-specific mRNAs (1, 21, and in replication (13, 26, 33, 37). In addition, there is an associated diphosphokinase activity (29,35) and a CAMP-independent protein kinase activity (18,24,39). The switch from transcriptive functions to replication probably involves some as yet undefined modification of the L/NS enzyme complex and a probable role for host factors in this switch (26,33) which leads to a suppression of transcription and associated capping and methylation to synthesis of full length (+) and (-) strands assembled into RNP particles. Direct evidence that the L/NS enzyme complex is involved in both transcription and replication was recently obtained through the use of monospecific antisera against the enzyme subunits which were were shown to affect both in uitro transcription and replication (12,14). This change in RNA products brought about by regulation of L/NS enzyme molecules also results in the abolition of the recognition of the two base intercistronic stop signals by the enzyme and the synthesis of complete complementary strands which are predominantly (-) strand (32).
A possible mechanism proposed for L/NS modification was the phosphorylation of the L/NS enzyme by the RNP-associated protein kinase (39), although Schnitzlein and Reichmann (31) have suggested that the L/NS subunit ratio was increased under conditions which favored transcription over replication.
To test the possibility that protein phosphorylation regulated L/NS enzyme function, we wanted to determine whether the protein kinase activity was one of the functions of the virus RNP proteins or the result of an associated host protein.
Published methods (9,16,22,25) for purifying the VSV RNP proteins did not, in our hands, yield proteins of sufficient purity, when analyzed by silver nitrate staining, to allow us to determine this. Previous methods have used high salt and Renografin to remove L and NS proteins from the RNP template and column chromatography with DEAE-Sephadex and phosphocellulose and centrifugation to separate and purify the VSV RNP proteins. When we isolated virion RNPs by centrifuging virions, lysed by Triton N101, in a high salt buffer through glycerol, the RNPs were not completely free of M protein (12). In addition, application of high salt-dissociated proteins to DEAE-Sephadex or phosphocellulose columns did not result in pure preparations of the individual RNP proteins, using RNPs prepared from our strain of VSV.
In view of this, we have developed an alternative purification method for our strain of VSV using Fractogel TSK which is comprised of a hydrophilic vinyl polymer which is stable in high concentrations of salt (11). We applied RNPs to a column of Fractogel, using a KC1 concentration which was sufficiently high to dissociate most of the L and NS proteins from the RNP-N template but not high enough to cause release of N protein from the RNP-N template. In 1 M KC1 the Fractogel column allowed separation of partially purified RNPs, an L/ NS complex, and M protein (see Fig. lA). The L/NS complex thus obtained was greater than 95% pure, as determined by silver nitrate staining. The L and NS proteins were subsequently separated from each other by high speed centrifugation through sucrose gradients containing l M KC1 (see Fig.  3A).
Using proteins with known molecular weights as standards, the apparent molecular weight of the L/NS complex, determined by elution from the Fractogel column, was approximately 75,000. This apparent molecular weight was obviously surprising since the molecular weight of L protein alone is about 180,000. We could not determine the number of NS molecules per L molecule in this complex because of the small amount of material in our preparations. Naito and Ishihama (25) showed that the active transcriptase probably was a LINSl heterodimer, but did not attempt to determine the molecular weight of this complex. The apparently low molecular weight that we observed for our L/NS complex was not due to degradation of either subunit since SDS-polyacrylamide gels of the complex components (Fig. 1A) never showed breakdown. This aberrant elution pattern is likely due to one or more factors such as enzyme configuration in 1 M KCl, charge effects of the Fractogel column, etc.
After obtaining purified enzyme subunits, we could then assay these purified VSV proteins for protein kinase activity plus and minus RNP-N template. We showed that the major peak of protein kinase activity did not correspond exactly with any of the VSV-specific proteins either eluted from the columns or isolated on sucrose gradients. The kinase activity associated with the RNP-N template was probably the result of the same kinase since the decrease of activity associated with RNP-N template resulted in a concomitant increase in this major peak of kinase activity.
Reconstitution of purified L and NS proteins with the RNP-N templae did not restore protein kinase activity indicating that formation of an RNP-enzyme complex was not necessary for kinase activity (Table I). These data strongly support the notion that the major VSV-associated protein kinase activity is a host-derived protein, as was suggested previously (18,22). This was further substantiated by the results showing no increase over background in protein kinase activity when VSV-specific mRNAs were translated in vitro to produce VSV proteins (Fig. 5).
Whatever the nature of this cellular protein kinase, it is present in small amounts since we were unable to detect a polypeptide which corresponded with the activity by silverstaining polyacrylamide gels of active preparations. In addition, it was tightly associated with the virion RNP since considerable kinase activity was always found in the partially purified RNP peak ( Fig. 1) even after incubation with 2 M CsCl and subsequent fractionation over Fractogel in the presence of 1 M KCl. Whether or not this association is a specific one is impossible to determine at this time.
The protein kinase activity does not use IgG as a phosphate acceptor (Fig. 2). IgG preparations specific for NS and M proteins had no effect on either endogenous protein kinase activity or activity with casein as an exogenous phosphate acceptor. Phosphorylation of the src protein, ppm, also was not inhibited by incubation with specific antisera (8,21).
Since N S protein has been shown to be phosphorylated shortly after synthesis in infected cells and is packaged into infectious virus as a phosphoprotein, we found in rather surprising that the RNP-associated kinase activity was, in all likelihood, a host cell protein. This finding seemed puzzling particularly since VSV seems to replicate quite well in such a broad spectrum of vertebrate and invertebrate cells. This broad host cell range, albeit not a natural in vivo range, would suggest that such cell kinases, if necessary to VSV enzyme function, are ubiquitous. Perhaps this matter could be partially resolved by further examination of kinase activity associated with VSV RNPs derived from a broad spectrum of cells in culture.
With enzyme subunits and RNP-N template greater than 95% pure we have begun fine structural studies of the L/NS enzyme to determine whether or not phosphorylation of one or both subunits is required for enzyme regulation and/or activation (5,15,20,25,34). We are presently utilizing our purified enzyme preparations to establish the exact number of phosphate residues per NS and L proteins and their locations on the NS molecule. These proteins were derived from virion RNPs, actively transcribing RNPs, intracellular pools of free NS proteins, etc.
This L/NS complex involved in VSV transcription and replication which shows remarkable changes in template recognition and product formation very likely has enzyme counterparts in all (-) strand animal virus systems such as influenza, measles, arenaviruses, bunyaviruses, etc., which all contain large multifunctional RNA polymerase molecules. Understanding the means by which these enzymes are modified and regulated in transcription and replication will be important to our understanding of the pathogenic mechanisms of these agents.
These large multifunctional enzymes are also somewhat similar to the well studied fatty acid synthase system of animal cells which is a seven-enzyme complex, and the first three enzymes responsible for the synthesis of orotic acid of the pyrimidine biosynthetic pathway in mammals (19) which is a trimer of a 200,000-dalton protein. The VSV RNA polymerase, which is a heterodimer, is unique in the sense that the enzyme in transcriptive mode possesses five or more activities all of which are suppressed by an as yet undetermined mechanism to switch to a replicative enzyme. These many activities are likely directed by active sites located on two polymerase subunits, L and NS, a fact which implies an extreme example of the conservation of genetic information.