Characterization of the Deoxyribonucleic Acid Polymerase Associated with Kilham Rat Virus

Purified preparations of the parvovirus, Kilham rat virus, have associated with them a protein with DNA polymerase activity. The enzyme has been separated from the other two or three viral proteins and purified 63-fold. The viral associated enzyme was found in a single peak of DNA polymerase activity after chromatography on DEAE-cellulose, DNA-cellulose, and phosphocellulose columns. It shares some properties in common with the host cellular DNA polymerases, described in the preceding paper (Salzman, L. A.. and McKerlie, L. (1975) J. Biol. Chem. 250, 5589%5595), but also has some important distinguishing characteristics. The Kilham rat virus-associated DNA polymerase has increased enzyme activity in the presence of 0.02 M KC1 and has a strong preference for a synthetic DNA polymer containing deoxyadenylate and deoxythymidylate. The enzyme has a molecular weight of approximately 75,000 * 3,000 and appears to contain endonuclease activity.

Purified preparations of the parvovirus, Kilham rat virus, have associated with them a protein with DNA polymerase activity. The enzyme has been separated from the other two or three viral proteins and purified 63-fold. The viral associated enzyme was found in a single peak of DNA polymerase activity after chromatography on DEAE-cellulose, DNA-cellulose, and phosphocellulose columns. It shares some properties in common with the host cellular DNA polymerases, described in the preceding paper (Salzman, L. A.. and McKerlie, L. (1975) J. Biol. Chem. 250, 5589%5595), but also has some important distinguishing characteristics. The Kilham rat virus-associated DNA polymerase has increased enzyme activity in the presence of 0.02 M KC1 and has a strong preference for a synthetic DNA polymer containing deoxyadenylate and deoxythymidylate. The enzyme has a molecular weight of approximately 75,000 * 3,000 and appears to contain endonuclease activity.
Kilham rat virus (KRV) is one of a ubiquitous group of small (15 to 30 nm), DNA-containing viruses called the parvoviruses (1, 2). KRV' was first isolated from a rat sarcoma and can cause a variety of diseases and malformations in neonatal animals (3,4). All members of the parvovirus group that have been studied contain 1 molecule of single-stranded, linear DNA with a molecular weight of 1.5 to 2.2. x lo6 and three viral proteins (5). They are assembled in the nucleus of an infected cell and causes disintegration of the nucleolus and cytoplasmic structures before cell lysis (6,7). Very little information is available about the replication of a linear single-stranded viral DNA molecule in a cell. There are some reports that suggest the possibility of a double-stranded linear DNA intermediate in parvovirus replication (8)(9)(10). How this double-stranded DNA intermediate may be synthesized and how the progeny single-stranded viral DNA molecules are made is unknown.
Purified KRV has been shown to have associated with it an enzyme with DNA polymerase activity (11). We decided to purify this enzyme in the hope that knowledge about the viral-associated DNA polymerase might lead to a better understanding of the process of DNA replication or its regulation in the infected cells. In the preceding report we have described the DNA polymerases from the rat nephroma cell line in which the virus is grown. Comparison of the viral-associated DNA polymerase with the cellular polymerases have shown that the viral enzyme resembles most a cytoplasmic cellular enzyme C,, but differs from this enzyme in several characteristics. EXPERIMENTAL (12). DEAE-cellulose, DNA-cellulose, and phosphocellulose were prepared as described in the preceding report (13). Pancreatic DNase I and micrococcal nuclease were purchased from Worthington. Spectrafluor was purchased from Amersham/Searle, Arlington Heights, Ill. Munktell 410 cellulose was purchased from Bio-Rad, Richmond, Calif.

Virus Purification
and Cell Line-KRV strain 308 was originally obtained from L. Kilham. The virus was grown in a rat nephroma cell line as previously described (14). The infected cells were harvested 7 days after infection, frozen and thawed, and treated with receptor destroying enzyme. The virus-containing supernatant was then centrifuged in a CsCl gradient (average density 1.41 g/ml) in a Spinco model L centrifuge at 35,000 rpm in a type 40 rotor at 5" for 24 hours. The virus was contained in a band at a density of 1.40 to 1.41. The material from this band was purified further by one or two additional isopycnic centrifugations in CsCl. The virus was then dialyzed against 4 liters of 0.02 M Tris, pH 8.2, and lOma M EDTA and was called purified virus. The purified virus preparation was found to sediment as a single peak containing both hemagglutination and DNA polymerase activity in a sucrose gradient, as previously described (11 The enzyme (6 ml) was added to a DNA-cellulose column (9 x 1 cm) equilibrated with the same buffer and washed with this buffer until all of the unadsorbed protein was removed.
The enzyme was then eluted with 160 ml of a linear gradient of 0 to 1.0 M NaCl in the above buffer. The active fractions were combined (6 ml) and dialyzed against 500 volumes of the 0.1% Triton X-100-10% glycerol buffer above.

Separation of the KRV-associated
DNA polymerase from the virion proved to be a difficult task. Treatment with various detergents and several concentrations of ammonium sulfate, and sedimentation through sucrose gradients containing detergent and salt either resulted in a failure to separate the enzyme from the rest of the virus particle or in a loss of the enzyme activity. We have found that brief sonication of the purified virions in the presence of 5 mM dithiothreitol and 4% NP-40 followed by incubation with 0.5% deoxycholate at 4" resulted routinely in a good recovery of virus polymerase activity. After dialysis the preparation was placed on a DEAE-cellulose column and eluted with a linear gradient of 0.02 to 0.5 M potassium phosphate. As seen in Fig. lA, the DNA polymerase was eluted from DEAE-cellulose in a large peak of activity at 0.04 M potassium phosphate. The active fractions (Fractions 28 to 45) were combined and accounted for 95 to 97% of the recovered enzymatic activity. The small peak at Fraction 55 was not routinely reproducible and was not examined further. After dialysis the combined fractions were added to a DNA-cellulose column, and the enzyme activity was eluted with a linear gradient of 0 to 1.0 M NaCl. A single very sharp peak of enzyme activity was eluted with 0.095 M NaCl (Fig. 1B). An aliquot of the enzymatically active fractions (Fractions 15 to 19) were combined, dialyzed, and added to a phosphocellulose column. The enzyme was eluted from the phosphocellulose with a linear gradient of 0.02 to 0.9 M potassium phosphate, pH 7.4. The DNA polymerase activity eluted in a single sharp peak with 0.02 M potassium phosphate. Table I summarizes  the  purification of the KRV-associated DNA polymerase. The enzyme was purified to a specific activity of 1125, a 63.fold purification from the disrupted virus particles which originally contained only three major proteins detectable by polyacrylamide gel electrophoresis (19). The final enzyme recovery was about 20% of the initial activity.
Properties of the KRV-associated DNA Polymerase-The KRV-associated DNA polymerase obtained after purification through the phosphocellulose step was examined for the following properties.
Effect of pH-The effect of the pH on the assay reaction was studied under the standard conditions except that the pH and buffers were varied. The pH activity curve is bell-shaped with maximum activity at pH 8.5 to 9.2. At pH 7.5 only 28% of the activity remained, and at pH 10 only 54% of the activity remained.
We routinely used a pH of 8.9 for the assay procedures.
Cation  most eukaryotic DNA polymerases. We studied it further as a function of duration of incubation as depicted in Fig. 2. The incorporation of the four deoxynucleotides proceeds almost immediately and is linear from 15 to 90 min of enzyme incubation. This is followed by a period of reduced incorporation. When only one deoxynucleoside triphosphate is used as a substrate for the enzyme, there is a lag period of 30 min, linear incorporation for 30 min, and then a leveling off of incorporation. The reason for this lag period is not known. The maximum incorporation of one deoxynucleotide is less than 30% of that found when 4 deoxynucleotides are present. Addition of fresh enzyme to the incubation mixture containing four dTNPs after a 180-min incubation reproduces the initial reaction kinetics without any lag period. The biphasic kinetics suggests that a rapid rate of initiation is followed by a slower rate of polymerization, the extent of which may be limited by some enzyme instability.

Primer-Template
Studies-The activity of the KRVassociated DNA polymerase is dependent on added primertemplate (Table III) The ability of the KRV-associated DNA polymerase to use various primer-templates is shown in Table III. The enzyme is only 30% as active in incorporating deoxynucleotides into DNA when native double-stranded salmon sperm DNA is used as a template. Single-stranded DNA does not serve as a template. Activated DNA that is heat-denatured serves as a template for the enzyme with about 60% of the activity of activated DNA. The enzyme also requires 3'.OH termini for activity, and when The KRV-associated DNA polymerase assays were performed under standard conditions using 1.2 pg of enzyme. The primer-templates used in the experiments were as indicated in the table. The activity is presented as a percentage of that obtained in the complete system in which 2.7 nmol of deoxynucleotide were incorporated per hour. b The radioactive and nonradioactive deoxynucleoside triphosphates were added to each incubation at a final total concentration of 40 nM per incubation. c [3H]NTP refers to a mixture of [3H]ATP (0.5 mCi, specific activity 6.2 mCi/pmol) and unlabeled GTP, CTP, ATP, and UTP (50 nM each per incubation). this group is not present, as after treatment of the DNA with micrococcal nuclease, enzyme activity is sharply reduced.
The KRV-associated DNA polymerase also copies synthetic DNA molecules.
In the presence of radioactive dGTP and nonradioactive dCTP, the enzyme is only 50% as active with poly (dG-dC) as with an equal amount of activated DNA. With radioactive dCTP and unlabeled dGTP as substrates the poly(dG-dC) did not serve as a template.  Standard assay conditions were used with the additions indicated in the table. In experiments testing the effects of p-chloromercuribenzoate, P-mercaptoethanol was omitted from the incubation mix. The DNA template was activated salmon sperm DNA. The activity is presented as a percentage of that obtained in the standard assay in which 2.7 nmoles of deoxynucleotide were incorporated by 1.2 pg of enzyme protein per hour. phosphocellulose-purified enzyme was determined by sedimentation in a sucrose gradient with markers of known sedimentation coefficients (s 20J and molecular weight. When the KRV-associated DNA polymerase was sedimented in a 5 to 20% sucrose gradient (Fig. 3), the enzyme was found in a sharp peak of activity between Escherichiu coli alkaline phosphatase (A) (szO,, = 6.3, MW = 80,000) and bovine serum albumin (B) (s~,,~ = 4.4, MW = 68,000). Using the bovine serum albumin as a marker, the KRV-associated DNA polymerase has an approximate calculated s20,w = 5 and a MW = 72 to 78,000.
Other Enzyme Activities-We examined the phosphocellulose-purified KRV-associated DNA polymerase for endonuclease activity by determining its ability to "nick" the double stranded circular form of SV40 (Component I) and for exonuclease activity by the release of acid soluble radioactivity from radioactive native and denatured adenovirus DNA (see "Methods").
As seen in Fig. 4, after incubation with the enzyme (1.2 ~g), 49% of SV40 Component I is converted to a form which migrates in the area of Component II (one linear strand, one circular strand in double-stranded DNA) and Component III (two linear DNA strands in double-stranded DNA). Further analysis of the enzyme-treated SV40 DNA in an alkaline sucrose gradient to denature the DNA and separate the strands into 18 S (single-stranded circular DNA) and 16 S (single-stranded linear DNA) species showed that the KRVassociated DNA polymerase-treated DNA sedimented in fragments of 6 S and smaller. The DNA polymerase endonuclease activity appears to have cleaved both strands of SV40 Component I in several places.
We were unable to demonstrate an exonuclease activity against either native or denatured adenovirus DNA ( <I% of polymerizing activity). An exonuclease activity preferring single-stranded DNA as substrate has been found with purified preparations of a vaccinia virus-directed DNA polymerase (27). A very low exonuclease activity was recently found with the Herpes simplex virus-induced DNA polymerase (26) that could not be detected with a natural DNA such as we used We have not carried out this assay. We were also unable to demonstrate a terminal deoxynucleotidyltransferase activity when oligonucleotides such as poly(A), or d(pT), were used as primers (28). DISCUSSION KRV was purified from possible contaminating cellular components by two or three successive isopycnic centrifugations in CsCl gradients.
The purified virion sediments as a single peak during velocity centrifugation in sucrose. The peak has associated with it the virion hemagglutinin and an enzyme with DNA polymerase activity (11). We have separated the enzyme from disrupted virions and disassociated it from the other known virion proteins. From results obtained by sedimentation in sucrose gradients with known protein markers it appears to have a molecular weight of 72,000 to 78,000 and a sedimentation coefficient s,,, Iu of 5. KRV has previously been reported (19) to contain a main capsid protein (with a molecular weight of about 62,000) which comprises approximately 75% of the virion protein.
The purified virus also contains two minor protein components with molecular weights of 72,000 and 55,000 that represent approximately 15 and 11.4% of the virion protein, respectively. The DNA polymerase may account for the largest molecular weight protein associated with the virions. This protein is present in approximately 8 to 10 copies per virion (19). In a preceding paper (13), we have characterized the DNA polymerases isolated from the rat nephroma cell. The stimulation of the KRV-associated DNA polymerase by 0.02 M NaCl is a property shared by only one cellular enzyme, the cytoplasmic DNA polymerase C,,. The viral-associated polymerase and C,, FRACTION NUMBER FIG. 4. Sedimentation of SV40 DNA Component I after incubation with KRV-associated DNA polymerase. SV40 DNA was incubated with 1.2 c(g of heat-inactivated KRV-associated DNA polymerase (0---0) or with 1.2 pg of active enzyme (04) for 60 min at 37". The incubation mixtures were then layered on individual 11-ml 5 to 30% sucrose gradients containing 0.05 i Tris, pH 7.5, 0.1 M NaCl, and 0.0025 M EDTA and centrifuged at 27.000 ram for 16 hours in a SW41 T rotor. Sedimentation is-from right to 'left with Fraction 1 taken from the bottom of the tube. have a common pH optimum of 8.0 and similar K, of dNTP, K, of activated DNA, and Mg2+ optimum. Like all of the cellular enzymes with the exception of C,, the KRV-associated enzyme utilizes the synthetic polymer [d (A-T)] best of all the primer-templates tested. The viral-associated enzyme uses this primer-template with much greater efficiency than the cellular enzymes. For the KRV-associated enzyme the molecular weight is 72,000 to 78,000 and the s,,, w is 5. The cellular C,, enzyme has not been purified to the same high degree as the viral-associated enzyme, but its molecular weight is believed to be about 70,000 and the sZO, w between 4 and 5.5.
There are, however, several important differences between the viral-associated enzyme and the cellular enzyme, C,,. The elution pattern of the two enzymes from DEAE-cellulose and DNA-cellulose is different. C,, enzyme elutes from DEAE-cellulose with 0.1 M KPO, and from DNA-cellulose at 0.55 M NaCl. The viral-associated enzyme eluted from DEAE-cellulose with 0.04 M KPO, and from DNA-cellulose with 0.095 M NaCl. The viral-associated enzyme activity is stimulated 2.5fold by Mn2+ in comparison to Mg2+ as the divalent cation. Mn2+ in the presence or absence of KC1 is as good a source of divalent cation as Mg2+ plus KC1 for this enzyme. The cellular enzyme, CI1, in the presence of Mn2+ with or without KCl, has only 32% of the activity found in the presence of Mg2+. A difference in DNA template specificity appears to exist between these two enzymes. The rat nephroma enzyme C,, can utilize the ribostrand of d lT19.
[A], to polymerize thymidylic acid, thus exhibiting R-DNA polymerase activity. The KRVassociated enzyme does not have R-DNA polymerase activity. The cytoplasmic cellular enzyme C,, can utilize native DNA as a primer-template just as efficiently as activated DNA. It cannot, however, effectively copy the synthetic heteropolymer d(G),,.d(C), and exhibits only 10 to 20% of the optimal activity with this template. The viral-associated enzyme does not utilize native DNA as an optimal DNA template and the enzyme activity falls to 30 to 40% of that found with an equal weight of activated DNA. This enzyme can copy effectively the synthetic heteropolymer d(G) 12 .d(C),. The KRV enzyme can also utilize denatured, activated DNA as a template, whereas the C,, enzyme has virtually no activity with this template. We do not know the conformation of this denatured, activated DNA. Since salmon sperm has A-T-rich regions, there may be regions of hairpin turns or folding back of the ends of the DNA. This finding is especially interesting if the viral-associated enzyme is involved in copying the single-stranded KRV DNA which contains 56% A-T.
The viral-associated enzyme shares some properties and has some characteristic differences from the other rat nephroma cell polymerases. The enzymes all have similar K, values for dNTP and activated DNA. The pH optimum is similar for the KRV-DNA polymerase, the cytoplasmic enzyme C,, and nuclear enzyme N, and N,. The nuclear enzymes have a similar Mg2+ optimum, although none of the cellular enzymes can substitute Mn2+ for Mg'+. All of the enzymes are inhibited to varying degrees by pyrophosphate, ethidium bromide, and p-chloromercuribenzoate.
As reported in the preceding report none of the other cellular enzymes except C,, appear to be stimulated by salt. In fact, the cytoplasmic enzyme C, and the major nuclear enzyme N, are strongly inhibited by KC1 (60 to 70% inhibition in 0.1 M KU). The cytoplasmic enzyme C, does not utilize the synthetic templates poly [d(A-T) ] and poly [d(G-C)]. The elution properties of the cellular enzymes from the three cellulose columns used in purification differ from the elution properties of the viral-associated enzyme. The nuclear enzyme N, and the cytoplasmic enzyme C,, are the only two cellular enzymes with calculated values for sZO, w and molecular weights close to the values found for the viral associated enzyme.
The KRV-associated enzyme appears to be a single DNA polymerase activity made in an infected cell containing multiple DNA polymerases. The viral-associated enzyme does not appear to be identical in properties to any of the cellular enzymes. These facts tend to support the idea that the viral-associated enzyme may play a role in viral DNA synthesis and may be a specific enzyme for viral replication.
However, these experiments do not rule out the possibility that the KRV-associated enzyme may represent a modification of a cellular enzyme such as C,, or be synthesized following derepression of a host specified DNA polymerase not normally detected in the cell.