Bacteriophage T7 Deoxyribonucleic Acid Replication in Vitro PURIFICATION AND PROPERTIES OF THE GENE 4 PROTEIN OF BACTERIOPHAGE T7*

The T7 gene 4 protein, a protein known from genetic analysis to participate in phage DNA replication in oiuo, has been purified approximately 500-fold with an in vitro complementation assay. The protein, purified from cells infected with a T7 gene 4 temperature-sensitive mutant, is thermolabile, establishing that the complementation activity is in the protein product of the phage gene 4. The purified protein has no detectable nuclease, DNA polymerase, or RNA polymerase activity. However, in addition to stimulating the rate of DNA replication in crude extracts of T7 gene 4 mutant-infected cells, the gene 4 protein effects a marked stimulation of DNA synthesis by the purified T7 DNA polymerase when duplex T7 DNA is used as template. This effect is not observed when denatured T7 DNA is used as template, or when phage T4 DNA polymerase or Escherichia coli DNA polymerase I, II, or III is substituted for the T7 enzyme. Analysis of the DNA synthesized by the T7 DNA polymerase in the presence of the gene 4 protein indicates that much of the product is in short DNA chains which are not covalently attached to the template. This result suggests a novel mechanism for the initiation of DNA chains in this reaction.

The T7 gene 4 protein, a protein known from genetic analysis to participate in phage DNA replication in oiuo, has been purified approximately 500-fold with an in vitro complementation assay. The protein, purified from cells infected with a T7 gene 4 temperature-sensitive mutant, is thermolabile, establishing that the complementation activity is in the protein product of the phage gene 4. The purified protein has no detectable nuclease, DNA polymerase, or RNA polymerase activity.
However, in addition to stimulating the rate of DNA replication in crude extracts of T7 gene 4 mutant-infected cells, the gene 4 protein effects a marked stimulation of DNA synthesis by the purified T7 DNA polymerase when duplex T7 DNA is used as template. This effect is not observed when denatured T7 DNA is used as template, or when phage T4 DNA polymerase or Escherichia coli DNA polymerase I, II, or III is substituted for the T7 enzyme. Analysis of the DNA synthesized by the T7 DNA polymerase in the presence of the gene 4 protein indicates that much of the product is in short DNA chains which are not covalently attached to the template. This result suggests a novel mechanism for the initiation of DNA chains in this reaction.
As summarized in a preceding paper (l), T7 DNA replication in uiuo requires at least six phage proteins and two host proteins. We have developed (2, 3). as have others (4, 5), a cell-free system in which DNA synthesis retains many of the properties of the in uiuo DNA replication reaction. In particular, DNA synthesis in vitro requires both the T7 DNA polymerase and the T7 gene 4 protein. In the accompanying papers (1, 6) the properties of the T7 DNA polymerase have been examined. The enzyme has been shown to consist of two subunits: one is the T7 gene 5 protein; the other is a host protein which is either missing or altered in the bacterial t.snC mutants. Since the T7 gene 4 protein is also required for DNA synthesis in the crude in vitro system, we have used a complementation assay to partially purify and characterize the T7 gene 4 protein.
Str?itling and Knippers (7) have also reported the partial purification of this protein.

EXPERIMENTAL PROCEDURE
Assay for Gene 4 Protein-The assay for gene 4 protein measures the stimulation of DNA synthesis in an extract prepared from cells infected with T7 carrying an amber mutation in gene 4. Reactions (final volume 0.1 ml) contained 10 mM Tris-HCl (pH 7.5), 10 mM 2-mercaptoethanol, 20 mM MgCl,, 0.3 mM each rNTP, and 0.3 mM each dNTP, with one of the dNTPs labeled with 3H or 32P, and 6 nmol of T7 DNA.
Twenty microliters of an extract (Fraction I) prepared from T7 3, 1. ,-infected Escherichia coli DllO (3) and from 0.1 to 0.5 unit of gene 4 protein, diluted in 10 rnM Tris-HCl (pH 7.5)/10 mM 2-mercaptoethanol/l mg/ml of bovine serum albumin were added to the reaction at O", and DNA synthesis was initiated by placing the reaction at 30". After incubation at 30" for 20 min, acid-insoluble radioactivity was determined as described previously (3). One unit of activity is defined as the amount of gene 4 protein which causes an increase in the rate of DNA synthesis equivalent to the incorporation of 1 nmol of radioactive nucleotide during the 20.min incubation.
Enzymes-E. coli DNA polymerase I was prepared by the procedure of Jovin et al. (8). Phage T4 DNA polymerase was the hydroxylapatite fraction purified as described by Goulian et al. (9); and phage T7 DNA polymerase was the DNA-cellulose fraction of Modrich and Richardson (6

Complementation Assay for Gene 4 Protein
Since the T7 gene 4 protein is required for extensive DNA synthesis in extracts of T7-infected Escherichia coli (3, 4), we have been able to purify the gene 4 protein using an in vitro complementation assay. The effect of the purified gene 4 protein on the rate of DNA synthesis in extracts prepared from cells infected with T7 bearing an umber mutation in gene 4 is shown in Fig. 1. The addition of the purified protein to these extracts resulted in a stimulation of the rate of DNA synthesis that was initially proportional to the amount of gene 4 protein added to the reaction. The maximum stimulation of 5-to 6-fold represents almost a full restoration of DNA synthesis activity. (Extracts prepared from cells infected with T7 bearing a wild  type gene 4 incorporate  about 1 nmol of dAMP during the  20-min incubation (3).)

Purification of Gene 4 Protein
All procedures were carried out at O-4" unless otherwise indicated.
The results of a typical purification are shown in Table I.
Growth of Phage-infected Cells-E. coli DllO was grown at 30" in a loo-liter fermentor (New Brunswick) in L-broth (10 g/liter of Bacto-tryptone, 5 g/liter of yeast extract, 10 g/liter of NaCl) supplemented with 1 g/liter of glucose and 10 mg/liter of thymine. At a cell density of log/ml (A,,, = 2.0), T7,. J. 6 phage were added at a multiplicity of 3 to 5, and 17 min after infection the culture was quickly chilled to 4" by the addition of crushed ice. The cells were harvested, and the cell paste (200 g) was resuspended in 800 ml of 50 mM Tris-HCl (pH 7.5)/10% sucrose, and 200-ml aliquots in 250.ml polycarbonate bottles were frozen in liquid nitrogen.
Preparation of Cell Extract-Frozen cells were thawed overnight at 4", and 20 ml each of 5 M NaCl and 10 mg/ml of lysozyme were added. After 45 min at 0" the solution was placed in a 37" water bath, stirred gently for 10 min to bring the temperature to 20", and then transferred to an ice bath and stirred until the temperature reached 5". The lysate was then centrifuged for 1 I% hours at 19,000 rpm in an International A54 rotor. The supernatant fluid was recovered and adjusted to A,,, = 200 by addition of 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 10% sucrose (Fraction I).
Streptom,vcin Sulfate and Ammonium Sulfate Fractionation-To 900 ml of Fraction I were added, 90 ml of freshly dissolved 30% (w/v) streptomycin sulfate. The solution was stirred for 30 min, and the precipitate was removed by centrifugation.
Ammonium sulfate (288 g) was added to 920 ml of the supernatant fluid. The solution was stirred for 45 min, and the precipitate was collected by centrifugation and redissolved in 1 liter of 20 mM Tris-HCl (pH 7.5)/0.1 mM EDTA/lO mM 2-mercaptoethanol/lO% glycerol (Buffer A) (Fraction III).

DEAE-cellulose
Chromatography-A column of Whatman DE-52 DEAE-cellulose (28 cm2 x 36 cm) was prepared and washed with 5 liters of Buffer A. The concentration of (NH,),SO, in Fraction III was determined by measuring the conductivity.
The fraction was diluted to 2 liters with Buffer A to reduce the (NH,),SO, concentration to 30 mM and then was applied to the column. The resin was washed with 1 liter of Buffer A containing 0.1 M NaCl, and proteins were then eluted with an S-liter linear gradient from 0.  mM 2-mercaptoethanol/lOO/o glycerol and was applied to the column. The resin was washed with 120 ml of Buffer B, and the proteins were eluted with a ,500.ml linear gradient from 0.0 to 0.5 M KC1 in Buffer B. Gene 4 complementation activity eluted at about 0.18 M KCl. Fractions containing the activity were pooled (123 ml), concentrated 5-fold by dialysis against dry polyethylene glycol for 6 hours. and then dialyzed overnight with 500 ml of Buffer B containing 0.15 M NaCl (Fraction V).

DEAE-Sephadex
Chromatographv-A column of DEAE-Sephadex A-50 (0.8 cmZ x 11 cm) was prepared and washed with 100 ml of Buffer B containing 0.15 M NaCl. Fraction V was applied to the column, and the resin was washed with 10 ml of Buffer B containing 0.2 M NaCl. Protein was then eluted with a loo-ml linear gradient from 0.2 to 0.5 M NaCl in Buffer B. Gene 4 complementation activity eluted at 0.3 M NaCl at the leading edge of the major protein peak (Fig. 2). Fractions containing greater than 1,000 units/mg of gene 4 activity were pooled. concentrated approximately 2-fold by dialysis against dry polyethylene glycol, and then dialyzed overnight against 10 mM sodium phosphate buffer (pH 7.1), 10 mM 2-mercaptoethanol/ 0.1 mM EDTA/60% (w/v) glycerol (Fraction VI). The concen- trated protein was stored at -15" for several months without significant loss of activity.
Evidence That Complementation Activity Resides in Product of T7 Gene 4 The assay employed for the purification of the gene 4 protein is not completely specific for the product of the phage gene 4. The addition of highly purified T7 DNA polymerase to these extracts can also result in significant stimulation of DNA synthesis. Moreover, it seemed likely that a number of nucleases might also bring about this effect. However, two lines of evidence argue that the complementation activity that we have purified is indeed the product of the T7 gene 4. First, when a "mock" purification was carried out through the DEAE-cellulose step using cells infected with a T7 gene 4 amber mutant, no complementation activity corresponding to the gene 4 protein could be identified.
A small amount of stimulatory activity was observed in fractions from the DEAE-cellulose column, but this activity was eluted coincident with the T7 DNA polymerase at a lower salt concentration than the gene 4 protein. Second, a thermolabile activity has been purified through DEAE-cellulose from cells infected with T74tS,0,, a T7 gene 4 temperature-sensitive mutant. The thermolability of this protein is compared with that purified from a T7+ infection in Table II. When the complementation assay was carried out a 42" instead of the standard 30", the protein purified from the T71t8,01 infection appeared to be only slightly more thermolabile than the wild type protein. However, if the gene 4 protein was first incubated in buffer at 42" and then assayed at 300, the wild type protein lost only 15% of its complementation activity, while the mutant protein suffered a greater than 95% loss of activity. An attempt to purify the temperature-sensitive protein further resulted in a rapid loss of activity, presumably because of its increased lability.

Physical Properties
Purity-Analysis of Fraction VI of the gene 4 protein by acrylamide gel electrophoresis in the presence of sodium dodecyl sulfate revealed several protein bands (Fig. 3). However, analysis of separate fractions from the DEAE-Sephadex column revealed that only the 57,000-and the 66,000-dalton proteins were eluted from the column coincident with gene 4 complementation activity.
The other major proteins were found to elute from the column just after the gene 4 activity along with the major protein peak. Studier (12) has reported that gel electrophoresis of radioactive proteins synthesized by cells infected with T7+ and T7 gene 4 amber mutants indicates that two polypeptides are altered in the mutant infections. We have confirmed this observation (data not shown) and find that the 57,000-and the 66,000-dalton proteins present in the purified gene 4 protein run coincident with the radioactive polypeptides which are absent in T7 gene 4 amber mutant infections during electrophoresis in the slab gel system described by Studier (13). Assuming, therefore, that these two proteins are the product of the T7 gene 4, one perhaps being a proteolytic cleavage product of the other, then Fraction VI is approximately 25% pure.
Sedimentation Coefficient of Gene 4 Protein-Fraction V of gene 4 protein (20 units) was analyzed by zonal sedimentation through a 10 to 30% glycerol gradient containing 20 mM Tris-HCl buffer (pH 7.5)/0.1 M NaCl. A single peak containing 70% of the applied activity was recovered from the gradient. The szo.w was calculated to be 4.7 S using bovine serum albumin (4.4 S) as a standard. Assuming that the gene 4 protein is globular and has a v = 0.73, this sedimentation coefficient would correspond to a protein with a molecular weight of 50,000 to 60,000 (14). This molecular weight is in reasonable agreement with that of either of the two polypeptides that have been identified as the products of gene 4, and suggest that, at least under the conditions of the analyses, the gene 4 protein is not in a complex with other polypeptides.
Absence of Nuclease Activities-The purified gene 4 protein (Fraction VI) contains no detectable exonuclease activity. Incubation of 4 pg of the purified protein with 1 nmol of either native or denatured T7 [3H]DNA for 20 min at 30" under the conditions used for the complementation assay produced less than 1 pmol of acid-soluble radioactive nucleotides.
In addition, analysis of the treated DNA by zonal sedimentation in an alkaline sucrose gradient (data not shown) shows no detectable increase in the number of internal breaks in the DNA, indicating that the purified protein is also free of endonuclease contamination.
Absence protein. However, in addition to stimulating the rate of DNA synthesis in extracts of T7 gene 4 amber mutant-infected E. coli, the purified protein effects a marked stimulation of DNA synthesis by the T7 DNA polymerase when duplex T7 DNA is used as template (Fig. 4). As previously reported (19,20), T7 DNA polymerase is essentially inert when given duplex T7 DNA as template. The addition of purified gene 4 protein to these reactions results in a 10. to 20-fold increase in the rate of DNA synthesis. A maximum rate of DNA synthesis is obtained when the weight ratio of gene 4 protein (Fraction VI) to DNA polymerase is about 10. Since this preparation of gene 4 protein is estimated to be only about 25% pure, this suggests that the two proteins may be required in approximately stoichiometric amounts.
DNA synthesis in the presence of these two protein fractions proceeds at a constant rate of at least 30 min and the total amount of DNA synthesized in the reaction can represent at least 20% the amount of template DNA added to the reaction. The maximum rate of synthesis in this purified system is at most only 10% that observed in the crude in vitro reaction, suggesting that additional stimulatory factors are present in the crude extract.
While we have not yet established that the activity responsible for the stimulation of DNA synthesis by the T7 DNA polymerase resides in the gene 4 protein, this activity purifies together with gene 4 complementation activity through phosphocellulose and DEAE-Sephadex chromatography.
Furthermore, as will be discussed below, DNA synthesis catalyzed by the T7 DNA polymerase in the presence of Fraction VI of gene 4 protein resembles in several ways the reaction carried out in the crude in vitro system. In particular, our preliminary amber mutant can be restored only by the T7 DNA polymerase and not by a variety of other DNA polymerases tested (6). DNA synthesis with the purified gene 4 protein exhibits the same specificity for the T7 DNA polymerase (Table III). The rate of DNA synthesis on duplex T7 DNA by phage T4 DNA polymerase and by the E. coli DNA polymerases I, II, and III was not detectably increased by the addition of up to 0.4 pg of gene 4 protein.
As shown in Table III absence of the four rNTPs. Under alkaline conditions most of the 32P-labeled product bands at the density expected for fully light (L) DNA and is separated from the fully heavy (H) template DNA. Comparison of the gradients indicates no marked difference between the distribution of DNA synthesized in the presence and in the absence of the four rNTPs, although it does appear that the rNTPs cause a specific increase in the synthesis of DNA which bands at the L position of the gradient. These results indicate that a major portion of the DNA synthesized in this in vitro reaction is not covalently attached to the template DNA, suggesting that de novo initiation of DNA chains may occur. Sucrose Gradient Analysis of Product-We have further characterized the product of the in vitro reaction by zone sedimentation through both neutral and alkaline sucrose density gradients (Fig. 6). It is apparent that the 3H-labeled template DNA is not significantly degraded during incubation with the purified gene 4 and 5 proteins. While the analysis in alkaline sucrose density gradients indicates that a fraction of the template DNA molecules contained single strand breaks, the sedimentation profiles shown in Fig. 5 are not detectably different from that obtained when the template DNA was analyzed before addition to the DNA synthesis reaction.
The 32P-labeled product shows a broad distribution of sedimentation coefficients under neutral conditions. The sedimentation profile is very similar to that we have obtained in After incubation at 30" for 20 min, reactions were diluted with 1 ml of 10 mM Tris-HCl buffer (pH 8.0)/10 mM NaCl/5 mM EDTA, and were divided into equal aliquots for pycnographic analysis under neutral and alkaline conditions. Eight grams of CsCl, and for alkaline gradients 0.7 ml of 1 M potassium phosphate (pH 12.5), were added to each sample in a tared polyalomer tube, and each gradient was then brought to 14.0 g by the addition of 10 mM Tris-HCl (pH 8.0), 10 mM NaCl, and 1 mM EDTA. Centrifugation was for 90 hours at 33,000 rpm in a Spinco 40 angle rotor at 25". Fractions were collected, and acid-insoluble radioactivity was determined as described previously In studies using a temperature-sensitive gene 4 mutant, Wolfson and Dressler (23,24) have shown that, after shifting to the nonpermissive temperature.
large gaps appear on one side of each growing fork, further supporting the idea that the gene 4 protein functions in the initiation of "Okazaki fragments." Thus, studies with the purified gene 4 protein should elucidate the mechanism of DNA chain initiation during DNA replication.
After a 500-fold purification, the gene 4 protein is estimated to be at most 25% pure. The preparation has no detectable nuclease activities, but contains T7 DNA ligase as a major contaminant.
We have not identified catalytic activity for the purified gene 4 protein. However, the protein appears to stimulate DNA synthesis significantly by the T7 DNA polymerase when native T7 DNA is used as template. While we have not firmly established that this stimulatory activity resides in the gene 4 protein, it purifies with the gene 4 complementation activity. Furthermore, DNA synthesis by the purified gene 4 and 5 proteins shares many features in common with the reaction catalyzed in our crude in vitro system (3). Stimulation by the gene 4 protein requires the T7 DNA polymerase specifically, and also stimulation is not observed when denatured T7 DNA is used as a template-primer.
The rate of synthesis is increased by the addition of the four rNTPs to the reaction; rATP alone has no effect. The ratio of gene 4 protein to DNA polymerase required for maximal activity is approximately the same in the purified system as is observed in the complementation of crude extracts. Finally, pycnographic analysis of the product synthesized using T7 [13C,1jN]DNA as template suggests that a major portion of the product DNA is not covalently attached to the template DNA. While we cannot rule out the possibility that short pieces of the template-primer are attached to this product, it is possible that some novel mechanism for the initiation of DNA chains functions in this reaction.
The addition of the four rNTPs to the reaction appears to stimulate preferentially the production of short DNA chains that are not attached to the template DNA. Okazaki and his co-workers (25-27) have presented evidence that in E. coli the synthesis of each "Okazaki fragment" is initiated with an RNA primer. This mechanism for initiation might also function during T7 DNA replication.
We have been unable to detect any synthesis of RNA during the reaction catalyzed by the gene 4 and 5 proteins. Under conditions where 2 nmol of dTMP were incorporated into an acid-insoluble product, no detectable UMP (<lo pmol) was incorporated.
A small amount of ATP (10 pmol) was incorporated, but this may be caused by the adenylation of T7 DNA ligase which contaminates the gene 4 protein preparation.
However, in this in vitro reaction the rate of DNA synthesis is only reduced 50% by the omission of the rNTPs. Treatment of the dNTPs with periodate to destroy any ribonucleotides, which might be present as contaminants, did not further reduce the rate of DNA synthesis in the absence of ribonucleotides.
The rate of DNA synthesis in a reaction containing the purified gene 4 and 5 proteins is at most 10% that obtained when the same amount of these two proteins is added as a crude extract prepared from T7-infected cells. Apparently the extract contains a factor or factors which increase the rate of DNA synthesis by these proteins. The E. coli and T7 DNAbinding proteins both significantly increase the rate of DNA