Two Nucleic Acid-dependent Nucleoside Triphosphate Phosphohydrolases from Vaccinia Virus PURIFICATION AND CHARACTERIZATION

Abstract Two nucleic acid-dependent nucleoside triphosphate phosphohydrolases have been found in vaccinia virus cores. Both enzymes hydrolyze ATP to ADP and Pi in the presence of nucleic acid. The enzymes were solubilized from purified viral cores by treatment with sulfhydryl-reducing agent, salt, and sodium deoxycholate in a Tris buffer at alkaline pH. DNA-free viral extracts were prepared by adsorption of viral DNA to DEAE-cellulose columns in the presence of high salt, and the two enzymes were purified by DNA-cellulose chromatography. Phosphohydrolase I was purified 300- to 400-fold with yields of 80 to 90 % and is approximately 95 % pure. A molecular weight of approximately 61,000 was calculated for the native enzyme using Sephadex gel chromatography and sucrose gradient sedimentation. A molecular weight value of 68,000 was obtained under denaturing conditions of sodium dodecyl sulfate polyacrylamide gel electrophoresis, indicating that phosphohydrolase I is a monomeric enzyme. The smaller amount of phosphohydrolase II prevented accurate determinations of specific activity or purity. The molecular weight of the native enzyme was calculated to be 68,000 and it may also consist of a single polypeptide. Nucleic acids were required by phosphohydrolase I and II for the hydrolysis of ATP, and optimal activities were obtained at neutral pH in the presence of divalent cations. The hydrolysis of ATP resulted in the production of stoichiometric amounts of ADP and Pi. The Km values of phosphohydrolases I and II were 1.4 x 10-4 m and 6.4 x 10-4 m, respectively.

have been found in vaccinia virus cores. Both enzymes hydrolyze ATP to ADP and Pi in the presence of nucleic acid. The enzymes were solubilized from purified viral cores by treatment with sulfhydryl-reducing agent, salt, and sodium deoxycholate in a Tris buffer at alkaline pH. DNA-free viral extracts were prepared by adsorption of viral DNA to DEAE-cellulose columns in the presence of high salt, and the two enzymes were purified by DNA-cellulose chromatography.
Phosphohydrolase I was purified 300-to 400fold with yields of 80 to 90% and is approximately 95 % pure. A molecular weight of approximately 61,000 was calculated for the native enzyme using Sephadex gel chromatography and sucrose gradient sedimentation. A molecular weight value of 68,000 was obtained under denaturing conditions of sodium dodecyl sulfate polyacrylamide gel electrophoresis, indicating that phosphohydrolase I is a monomeric enzyme. The smaller amount of phosphohydrolase II prevented accurate determinations of specific activity or purity. The molecular weight of the native enzyme was calculated to be 68,000 and it may also consist of a single polypeptide.
Nucleic acids were required by phosphohydrolase I and II for the hydrolysis of ATP, and optimal activities were obtained at neutral pH in the presence of divalent cations.
The hydrolysis of ATP resulted in the production of stoichiometric amounts of ADP and Pi. The K, values of phosphohydrolases I and II were 1.4 X low4 M and 6.4 X IOh M, respectively.
A number of enzymatic activities are associated with purified preparations of vaccinia virus. These include a DNAdependent RNA polymerase (1, 2) a deoxyribonucleasc (3, 4), a nuclcositlc triphosphate phosphohytlrolase (5, 6), a protein kinase (7,8), and a polyadcnylate polymcrase (9, 10). Now of the enzymes are released upon treatment of the virion with nonionic detergents, indicating that they arc contained within the core. Rcccntly, \ve * A preliminary report of these studies was prcscntcd at the Annual Meeting of the American Society for Microbiology, 1973.
noted that the phosphohydrolase was solubilized during disruption of the vaccinia core and that the released enzyme was dependent for activity upon the presence of nucleic acids (11). The latter finding was novel since similar enzymes have not been described in either animal cells or other viruses. DNA-dependent ATPasc activities of at least two varieties have been found in prokaryotic cells. The ATP-dependent deosyribonucleases from uninfected or phage-infected bacteria exhibit DNA-dependent ATPase activities (12-21). Certain other briefly described DNA-dependent ATPascs lack dcmo&rable nuclease activities and arc of unknown function (22,23).
In this report we describe the purification, some physical prop-erties, and a preliminary characterization of the nucleic aciddependent nucleoside triphosphatc phosphohydrolase activities from vaccinia virus. In the following communication (24) we will dcscribc the nucleotitlc substrate and nucleic acid cofactor specificities of the purified enzymes.

MITERIALS ASD METHODS
Cell a,rd Virus Slraifls-Vaccinia virus (strain WR) was purified from HeLa cells by the procedure of Joklik (25) which included sedimentation through a sucrose cushion and two sucrose gradient ccntrifugations as previously described (26). Virus grown in the presence of [3H]lcucine was purified in a similar manner.
After drying, the chromatography plates were washed in beakers with deionized water, dried, and developed with 1 M CH&OOII-4 M LiCl (8:2, v/v) as described by Randerath and Randerath (27). After chromatography the plates were dried and the ADP and ATP spots visualized under ultraviolet light, cut out, and placed in scintillation vials. LiCl (0.5 ml of a 2 13 solution) was added to elute the nucleotidcs which were counted after addition of 5 ml of a toluene-based scintillation fluid containing Triton X-100 (28). By definition, 1 enzyme unit will form 1 nmole of ADP from ATP in 1 min. 1. DNA-cellulose chromatography of nucleic acid-dependent phosphohydrolase activity from vaccinia virus. Phosphohydrolase activity from a low salt DEAlS-cellulose column was applied directly to a DNA-cellulose column (18 packed ml containing 280 pg of denatured calf thymus DNA per packed ml) and developed with a 350-ml gradient of 0.05 to 0.25 M NaCl in Buffer A ("Materials and Methods"). Fractions of 5.7 ml were collected at a flow rate of 12 ml per hr. Aliquots (0.1 ml) were removed from each fraction and counted for [3H]leucine (O-O). Aliquots (5 11) were assayed for nucleic acid-dependent phosphohydrolase activity (0-e).
NPH I and NPH II, phosphohydrolase I and II, respectively. 0.9-cm diameter, (18 packed ml, 280 pg of DNA per ml) and equilibrated with Buffer A containing 0.05 M NaCl. The flowthrough from the low salt DEAE-cellulose column was applied directly to the DNA-cellulose column and after extensive washing with Buffer A containing 0.05 M NaCl, the column was developed with a gradient composed of Buffer A containing 0.05 and 0.25 M NaCl as the limiting salt concentrations.
The results of such a DNA-cellulose column chromatography are shown in Fig. 1. About 40 to 50% of the protein did not adsorb to the column as monitored by following the distribution of [3H]leucine-labeled viral material. Several radioactive protein peaks were eluted between 0.1 and 0.25 in NaCl.
In other experiments no additional protein was eluted with salt concentrations as high as 1 M. The column fractions were then assayed for phosphohydrolase activity and the results are presented in Fig. 1. Some phosphohydrolase activity appeared in the DNAcellulose flow-through.
The amount of this unadsorbed phosphohydrolase activity was greatly recluced in subsequent esperiments by using longer columns.
Two peaks of nucleic aciddependent, phosphohydrolase activity were eluted from the DNA-cellulose column on application of the salt gradient. The first peak, by far the major enzymatic component, was eluted at a salt concentration of approximately 0.1 M and was coincident wit.11 a [W]leucine-labeled protein peak. A second, minor peak of phosphohydrolase activit,y was eluted at approximately 0.17 M. This peak of activity was not coincident with a [3H]leucine-labeled protein peak indicating the presence of multiple proteins in that region.
Greater resolution of the minor phosphohydrolase activity has been obtained in subsequent purifications by increasing the total volume of the gradient used to develop the DNA-cellulose column.
The two phosphohydrolase activities rechromatographed as distinct species at the expected salt concentrations and will be referred to as phosphohydrolase I and phosphohydrolase II.
2. Purification of phosphohydrolase I and phosphohydrolase II as monitored by SDS polyacrylamide gel electrophoresis. Gels A to C show the polypeptide composition of fractions containing approximately 200 phosphohydrolase units derived from purified cores, high speed soluble material from deoxycholate-extracted cores, and low salt DEAE-cellulose flowthrough material, respectively. Gels D and E represent approximately 4000 phosphohydrolase I and 2000 phosphohydrolase II units of enzyme purified from DNA-cellulose columns. These samples were concentrated by trichloroacetic acid precipitation as described under "Materials and Methods." Similar but fainter bands were obtained with unconcentrated material. Gel F shows the polypeptide composition of material eluting from the DNAcellulose column immediately after phosphohydrolase II. An equivalent volume of sample was concentrated for electrophoresis by trichloroacetic acid precipitation as that of phosphohydrolase I and phosphohydrolase II. Electrophoresis was from top to bottom.
Pur$cation of Phosphohydrolase I and Phosphohydrolase II as Moniiored by SDS Polyacrylamide Gel Electrophoresis-Standard SDS 7.5% polyacrylamide gels were used to monitor the degree of purification of the phosphohydrolase enzymes. Fig. 2 shows a typical set of results.
Gel A in Fig. 2 indicates the polypeptide composition of purified viral cores. Gel B shows the polypeptide composition of soluble material estracted from cores by deoxycholate treatment.
The major structural viral polypeptides 4a and 4b, which compose approximately 32yo of the total viral protein, remain insoluble with this treatment.
Gel C indicates the polypeptide composition of the flow-through material from the low salt DEAE-cellulose column.
The effective removal of several high molecular weight polypeptides should be noted. Gels D and E show the rather extensive purification of phosphohydrolase I and phosphohydrolase II, respectively, which is achieved by DNA-cellulose affinity chromatography. Purified preparations of phosphohydrolase I contain a single stained 3276 and prepared for electrophoresis on 7.5% SDS polyncrylamide gels as described under "Materials and Methods." The gel was frozen, sliced into l-mm sections, and dissolved with II,O*, and the fractions were counted for radioactivity.
band, while phosphohytlrolase II frcqucntlg contains two very closely spaced bands. USC of a mow shallow salt gradient to develop the DNA-cellulose column or rechromatography of ~~l~osphol~ydrolase II on a second I)N:\-cellulose column appeared to result in a diminution of the faster migrating polypeptide observed in Gel fl without loss of l)hosl)liohydrolase II activity.
Further work is necdcd, Hoover, to clarify the relationship of the two bands. Gel fi' contains the polypeptidcs cluted immediately following I)liosl)hollytlrolasc I I on DNAcellulose.
Subsequent studies have shown that this material contains polyadcnylatc polymcrasc activity (10). Gels A to C wcrc loaded with approximately 200 enzyme units per gel, whereas more than 4000 and 2000 cnzymc units of purified I)liosphohydrolase I and l)hospl~ol~ytlrolase II were applied on Gels D and R, respectively.
In order to determine the relative purity of l)hospliohydrolase I, approximately 4000 enzyme units of purified [3H]lcucinelabclcd phosphohydrolase I were concentrated by trichloroacetic acid prccipitstion and applied to 7.5$; SI)S polyacrylamide gels. After electrophorcsis the gel was &cd into 1-mni sections and dissolved with hydrogen pcrosidr, and [311]lcuci~le-labclcd material was counted.
The result of sucsh a dctcrmination is shown in Fig. 3. The purity of phosl)hohydrolasc I is in the order of 95% as estimated by SIX polyacrylamidc gel clcctrophoresis. A similar determination was not done for phosphohydrolasc II since the relatively low amount of radioactivity found in this minor enzymatic component would have madc a reliable estimate of purity difficult.
PuriJicafion of Phosphohgdrolase I&The relative purification of the nucleic acid-depentlcnt phosl)hohSclrolase I during the various steps is outlined in Table I. Removal of protein during the formation of cores from whole virions resulted in a doubling of specific activity.
An activation of phosphohydrolase activity occurred when the cores were disrupted by deosycholatc treatment,. Since only 25 to 350/, of the protein in the viral core was solubilizcd by deosycholatc treatment, a further increase in specific activity of the phosphohydrolase was obtained after high speed centrifugation. An  raphy.
The total increase in specific activity achieved during the purification procedures was 300-to 400-fold and between 80 and 90% of the initial activity was recovered.
Approximately 160 pg of purified phosphohydrolase I were recovered from 72 mg of purified virus. This suggests that about 0.2 to O.S%, of the total viral protein is phosphohydrolase I. The range of this estimate is dependent upon whether one uses the values of NP-40.treated total virus or deosycholate-disrupted cores as the starting point for phosphohydrolase activity in calculating the recovery of enzyme. The latter is perhaps a more reliable starting point.
No precise estimates regarding the increase in specific activity of phosphohydrolase II were obtained because of the small physical amounts of this enzymatic component. However, examination of the polyacrylamide gels indicated a similar order of purification.

Determination of Molecular Weights
Sephadez Gel Filtration-Purified preparations of [Wlleucinelabeled phosphohydrolase I were chromatographed on columns of Sephades G-200 as described under "Materials and Methods." Protein standards were also analyzed on the same column under comparable conditions.
The results are shown in Fig. 4. Phosphohydrolase I eluted at the same volume as bovine serum albumin which was compatible with a Stokes radius of 3.5 nm (55). The enzymatic activity coincided with the [Wlleucine radioactivity indicating the absence of any detectable impurities that could be separated by gel filtration.
I'hosphohydrolase II also eluted from Scphadex G-200 at the same volume as bovine serum albumin and is not shown.
Sucrose Gradient Centrifugation-The technique described by Martin and Ames (35) was used to determine the sedimentation behavior of phosphohydrolase I and phosphohydrolase II. The enzymes were sedimented on 5 to 20% gradients with standards of known sedimentation coefficients.
The results of such analyses are shown in Fig. 5 for phosphohydrolase I and Fig. 6 for phosphohydrolase II. In other experiments it was demonstrated that the included protein standards did not affect the sedimentation of phosphohydrolase I or phosphohydrolase II. A partial separation of phosphohydrolasc I and phosphohydrolase II was observed when the two enzymes were run on the same gradient with bovine serum albumin and assayed with divalent cations and cofactors specific for each enzyme (24) A value of approximately 68,000 was calculated for phosphohydrolase I by comparison to standards of known molecular weight (Fig. 7). The similar molecular weights obtained for phosphohydrolase I under native and denaturing conditions indicate that the enzyme is a monomer. Phosphohydrolase II has not been obtained in a homogeneous form. The band which is predominant in our most highly purified preparations has an electrophoretic mobility, however, almost identical to that of phosphohydrolase I. Thus, phosphohydrolase II probably also exists as a monomeric enzyme.

Conditions for Optimal Enzyme Activities
Nucleic Acid-The stimulatory effects of nucleic acids on the activities of phosphohydrolase I and phosphohydrolase II will be described in detail in the accompanying communication (24). The amount of commercial salmon sperm DNA contained in the standard reaction mixture described under "Materials and Methods" provided maximum activity.
pH-Using a variety of buffer systems bot.h phosphohydrolase I and phosphohydrolase II demonstrated a very broad pH optimum around neutrality. A pH 7.0 morpholinopropane sulfonic acid buffer (pK, = 7.2) was selected. No difference in activity of either enzyme was observed at several concentrations of buffer tested.
Requirements of Nonionic Detergent and Reducing Reagents-Neither purified phosphohydrolase I nor phosphohydrolase II showed any requirement for added nonionic detergents (NP-40, Triton X-100) or sulfhydryl-reducing reagents (DTT). These reagents, however, were retained in the phosphohydrolase assay for comparative purposes since the early fractions, particularly whole virions and cores, required these reagents for maximal activity.
The requirement for these reagents during the early purification steps was probably due to the disruptive effect of these reagents on the virus.
Requirement for Divalent Cation-Both phosphohydrolase I and phosphohydrolase II required a divalent cation for activity. This is shown in Fig. 8. No activity with either enzyme was observed in the absence of added Mg2+. Both enzymes showed optimal activity when Mg2+ was present in the range of 1 to 5 rnM.
The effect of other divalent cations either alone or with Mg2+ was investigated.
The data in Table II indicate that in the absence of Mg2f, Mn2+ was utilized quite efficiently by both enzymes. Ca2f could support the activity of phosphohydrolase I but not that of phosphohydrolase II. Neither Cu2f nor Zn 2f could substitute for Mg2+ as the divalent cation.
When these cations at 1 mM concentrations were present in the enzyme reaction in addition to 1 mM Mg"+, it was observed that Zn2f was inhibitory to both enzymes. Addition of Cu2+ inhibited phosphohydrolase I activity but allowed considerable phospho-  9 shows the rate of ATP hydrolysis by phosphohydrolase I and phosphohydrolase II as a function of ATP concentration. Double reciprocal plots of the data are shown in the inset. K, values of 1.4 X 10e4 M and 6.4 x 10e4 M were calculated for phosphohydrolase I and phosphohydrolase II, respectively. It should be noted that both enzymes were inhibited by high concentrations of ATP.
This experiment was carried out using a constant amount of Mg2+ while varying the ATP concentration. Similar K, values and inhibitions at high ATP concentrations were also obtained when the Mg2+:ATP ratio was kept constant.
EJect of Time and Enzyme Concentration on Phosphohydrolase I and Phosphohydrolase II Activity-Both phosphohydrolase I (Fig. 10) and phosphohydrolase II (Fig. 11) hydrolyze ATP without a lag period.
Both enzymes show an initial linear response with respect to time and the hydrolysis of ATP, within certain limits, was proportional to the amount of enzyme added (insets,Figs. 10 and 11). These figures also show the dependence of both phosphohydrolase I and phosphohydrolase II activity on the presence of added DNA as illustrated by the low levels of endogenous activity in the absence of added DNA.
Stoichiometry of Phosphohydrolase I and Phosphohydrolase II Reactions-The stoichiometry of ATP hydrolysis to ADP and and Pi was studied using both [3H]ATP and [yJ2P]ATP. AS can be seen in Table III the amount of labeled substrate hydrolyzed was equal to the amount of ADP and Pi that was formed.
This indicates that under standard assay conditions the products of ATP hydrolysis were not consumed by further reactions.
Side Reactions-A number of side reactions that might occur with ATP have been looked for. As expected from the stoichiometry data, hydrolysis of ADP does not occur. The enzymes also fail to hydrolyze AMP.
The enzymes fail to produce detectable ATP from ADP and inorganic phosphate and no exchange reactions were observed. Under standard conditions of assay, no formation of polynucleotides by either enzyme was observed.
Eflects of Various Substances on Enzyme Activity-The data in Table IV indicate that the end products of ATP hydrolysis, Pi and ADP, at equimolar concentrations with ATP are not very inhibitory to either phosphohydrolase I or phosphohydrolase II. Likewise PPi and AMP did not extensively inhibit either phosphohydrolase I or phosphohydrolase II. The phosphonic acid analogs of ATP with a methylene bridge between either the /3--y-or a$-phosphates were not very inhibitory to either phosphohydrolase I or phosphohydrolase II. Phosphohydrolase I was more sensitive to the inclusion of monovalent ions in the assay mixture than was phosphohydrolase II. In fact, phosphohydrolase II showed a small stimulation of activity at low monovalent ion concentration.
Neither cyclic adenosine 3' : 5'-monophosphate nor S-adenosylmethionine had any effect on either phosphohydrolase I nor phosphohydrolase II at various concentrations tested (data not presented).
Stability of Puri$ed Enzymes-Both enzymes have been stored at -20" for several months without loss of activity even when frozen as dilute solutions directly from the DNA-cellulose (data not presented). Inset shows a double reciprocal plot of the data. phohydrolase II were incubated in a standard assay. At various FIG. 10 (center). Phosphohydrolase I activity as a function of times of incubation aliquots were removed and ATP hydrolysis time and enzyme concentration.

MINUTES
Increasing amounts of purified determined as described under "Materials and Methods." The phosphohydrolase I were incubated in a standard enzyme assay. largest enzyme dose (5.0 units) was also tested in the absence of At various times of incubation aliquots were removed and hy-added DNA as indicated in the figure. The inset shows a measure drolysis of ATP measured as described under "Materials and of relative phosphohydrolase II activity (ordinate) as a function Methods." The largest enzyme dose (17.5 units) was also tested of microliters of enzyme added (abscissa) for the first 15 min of in the absence of added DNA as indicated in the figure. The incubation. 0 32Pi was separated from [y-32P]ATP using the standard chromatographic procedure described under "Materials and Methods" and was located by radioautography.
Two enzymes which split ATP to ADP and Pi were purified from vaccinia virus cores. Bot.h enzymes are stimulated by nucleic acid and extensive,purification was achieved by DNAcellulose chromatography.
Phosphohydrolase I, the more abundant enzyme, was purified to near homogeneity and represented approximately 0.2 to 0.6% of the total viral protein. The correspondence of molecular weights determined by sucrose gradient sedimetation, gel filtration, and SDS polyacrylamide gel electrophoresis indicates that the enzyme exists as a monomer with a molecular weight of about 68,000. From the recovery and molecular weight values we estimate that each virion contains between. 100 and 300 molecules of phosphohydrolase I. A turnover number of approximately 1 x lo4 molecules of ATP per min per enzyme molecule was estimated.
Similar estimates were not made for phosphohydrolase II which has a similar molecular weight because the low amounts did not allow reliable estimates of enzyme concentration.
As described by Siegel and Monty (58)  a Control activity for phosphohydrolase I and phosphohydrolase II was 6.4 and 4.7 nmoles of ADP/5 min, respectively. using the molecular weights derived from sucrose gradient sedimentation and the Stokes radii derived from Sephadex gel filtration. Frictional ratios of approximately 1.35 and 1.29 for phosphohydrolase I and phosphohydrolase II, respectively, were calculated, suggesting axial ratios of about 7 : 1 and 6 : 1 for these enzymes if considered as prolate ellipsoids (59). One must bear in mind that deviations of the partial specific volumes of these enzymes from the assumed value of 0.73 ml g-1 or unusual hydrations could introduce rather large errors.
With the exception of reverse transcriptase (60), purification of other enzymes associated \vith animal viruses has not been reported. This has undoubtedly been due to the relatively small physical amounts of purified virus available as starting material and the insoluble particulate nature of viruses. The procedure that WC have developed has been found sufficient for the solubilization and purification of the tlvo nucleic acid-dependent phosphohydrolases reported here from relatively small amounts of purified virus (70 to 80 mg). This proccdurc also has been used to solubilize the deosgribonuclease (41), polyadenylate polymerax (lo), and protein kinasc (61) from vaccinia virions.
The question of whether phosphohydrolase I and phosphohgdrolasc II arc virus-induced enzymes must be approached in several ways. Our preliminary unpublished csperiments indicate that DNA-dependent l~llos~~hol~ydrolase activity first appears in the cytoplasm of infected cells several hours after in fcction.
A similar activity Ivas not detcctecl in the cytoplasm of uninfected cells. In addition, further characterization of the t\vo enzymes indicate that they arc unlike any previously described (12-23).
Isolation of specific temperature-sensitive mutants mould probably be an unambiguous way of demonstrating the viral origin of the enzymes. IIovvevcr, the large size of the vaccinia genome, 160 x 106, makes this approach estremely difficult.
Synthesis of the enzymesin a DNA-directed, cell-free protein synthesizing system is a possibility that remains to be explored.