Replication of phi X174 dna with purified enzymes. I. Conversion of viral DNA to a supercoiled, biologically active duplex.

Conversion of phi X174 viral, single-stranded circular DNA to the duplex replicative form (RF), previously observed with partially purified enzymes, has now been demonstrated with the participation of 12 nearly pure Escherichia coli proteins containing approximately 30 polypeptides. To complete the synthesis of a full length complementary strand, E. coli DNA polymerase I was needed to fill the short gap left by DNA polymerase III holoenzyme, and to remove the primer and replace it with DNA. Production of supercoiled RF required the further actions of E. coli DNA ligase and gyrase. Net synthesis of viral circles was obtained by coupling the formation of RF supercoils to the actions of the phi X174-encoded gene A protein and E. coli rep protein. Viral DNA circles produced from enzymatically synthesized supercoiled RF, serving as template-substrate, were indistinguishable from those produced from RF isolated from infected cells; synthetic RF and the viral circles generated from it by replication were as biologically active in transfection of spheroplasts as the forms obtained from infected cells and virions. The conversion of single-stranded circular DNA to RF is suggested here as a model for discontinuous synthesis of the lagging strand of the E. coli chromosome. The primosome, a complex of some of the replication proteins responsible for initiations of DNA chains, will be described elsewhere. Multiplication of RF supercoils, described in the succeeding paper, proceeds by a rolling-circle mechanism in which the synthesis of viral strands may have analogies to the continuous synthesis of the leading strand of the E. coli chromosome.

Conversion of 4X174 viral, single-stranded circular DNA to the duplex replicative form (RF), previously observed with partially purified enzymes, has now been demonstrated with the participation of 12 nearly pure Escherichia coli proteins containing -30 polypeptides. To complete the synthesis of a full length complementary strand, E. coli DNA polymerase I was needed to fill the short gap left by DNA polymerase Ill holoenzyme, and to remove the primer and replace it with DNA. Production of supercoiled RF required the further actions of E. coli DNA ligase and gyrase. Net synthesis of viral DNA circles was obtained by coupling the formation of RF supercoils to the actions of the 4X174-encoded gene A protein and E. coli rep protein. Viral DNA circles produced from enzymatically synthesized supercoiled RF, serving as template-substrate, were indistinguishable from those produced from RF isolated from infected cells; synthetic RF and the viral circles generated from it by replication were as biologically active in transfection of spheroplasts as the forms obtained from infected cells and virions. The conversion of single-stranded circular DNA to RF is suggested here as a model for discontinuous synthesis of the lagging strand of the E. coli chromosome. The primosome, a complex of some of the replication proteins responsible for initiation of DNA chains, will be described elsewhere. Multiplication of RF supercoils, described in the succeeding paper, proceeds by a rollingcircle mechanism in which the synthesis of viral strands may have analogies to the continuous synthesis of the leading strand of the E. coli chromosome.
Mechanisms for replication of OX174 DNA, suggested from in vivo and in vitro studies (1-7), had not been firmly established because the many enzymes that participate in the several stages of the complex process had not been isolated in homogeneous form. Whether all the enzymes and factors needed for the process in vitro were serving essential functions and whether the major components of the physiological process in vivo were included among the components of the in * This work was supported by grants from the National Institutes of Health and the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. vitro reactions, remained uncertain. Furthermore, the coupling of complementary strand and viral strand synthesis to obtain multiplication of the duplex replicative form could not be achieved with partially purified proteins. Excision of uracil, misincorporated in place of thymine, resulted in fragmentation of the synthetic DNA products (8,9); dUTPase was required to avert this pitfall (9). This measure, and the removal of inhibitory factors by further purifications, made it possible to link efficiently the synthesis of viral strands from RF' with their direct use as templates for the synthesis of complementary strands to generate RF in net amounts (10).
Inasmuch as initial conversion of the infecting viral singlestranded circle produced a duplex form (RF II) with an incomplete synthetic strand, several questions regarding the effective reconstitution of the pathways of OX DNA replication from purifled enzymes remained unanswered. (i) Could the discontinuity be fEiled by DNA polymerase I and sealed by E. coli DNA ligase and could the relaxed, sealed duplex be supercoiled by E. coli gyrase? (ii) Would the supercoiled RF synthesized by the action of these pure enzymes prove to be as active as natural RF as a template-substrate in the generation of single-stranded viral circles? (iii) Are the synthetic products (RF and single-stranded viral circles) as biologically active in transfection of spheroplasts and are they replicated in vivo with the same high fidelity as the natural forms? Experiments that supply affirmative answers to these questions are reported here. Considered in the succeeding paper (11) are mechanisms for effective coupling of complementary to viral strand synthesis in the multiplication of RF.
Gel Electrophoresis-Electrophoresis of DNA was in 0.8-1% agarose gels at constant voltage (1 V/cm) at room temperature in either 90 mm Tris-borate (pH 8.3) and 2.5 mM EDTA or in 40 mM Tris. acetate (pH 7.8), 5 mM sodium acetate, and 1 mM EDTA. Ethidium bromide, when present, was at 1 gg/ml. Isolation of DNA from gels was as described elsewhere (28).

Conversion of OX Viral DNA to the Duplex Replicative
Form I The RF II Product Has a Gap in the Synthetic Strand-With a cruder enzyme system prepared from cells deficient in DNA polymerase I (29), the RF II product contained a nearly full length but incomplete complementary strand. It was not clear whether the discontinuity in the synthetic strand was a gap or a nick. The DNA product synthesized by the purified enzyme system reconstituted from proteins dnaB, dnaC, i, n, n', n", primase, DNA polymerase III holoenzyme, and SSB has the following properties: (i) insusceptibility tc ci-jercoiling by ethidium bromide (Fig. 1B) and therefore is not a covalently closed form (RF I), (ii) nonconvertibility to RF I by E. coli DNA ligase (Fig. 1C) and therefore possesses either a gap or the RNA primer at the 5' end, or both, and (iii) nonconvertibility to RF I by T4 DNA ligase (Fig. ID) and thus likely possesses a gap, inasmuch as T4 DNA ligase can join DNA and RNA termini. However, other anomalies at the 3'-hydroxyl and 5'-phosphate termini that might interfere with T4 DNA ligase action or even the lack of a 5'-phosphoryl group are not excluded.
RF II is Converted to RF I by DNA Polymerase I Gapfilling and Primer Removal and by DNA Ligase Sealing-When E. coli DNA polymerase I and E. coli DNA ligase were present during the conversion of SS to RF II or added to the incubation mixture after RF II was formed, the product was a covalently closed duplex circle (RF I) (Fig. 1E). Approximately 80% of the input viral DNA was converted to RF I. T4 DNA ligase served as well as E. coli DNA ligase (data not shown).
Complex Products are Linked to DNA Polymerase I Action-The products of the SS to RF reaction obtained in the absence of DNA polymerase I and DNA ligase were almost exclusively RF II as judged from ethidium bromide staining in the gel electrophoretic analyses (Fig. 2); however, radioactivity measurements indicated that 2 to 4% of the labeled products were located between the origin and the RF II position. In the presence of polymerase I and ligase, this FI(G. 3 (ri.ght). Electron micrograph of complex DNA product linked to INA polymerase I action. SS to RF reactioin conmpoients w ere as in Fig. 1. wkith the adldlitioni of A. coli D'\-A polynmerase I as descritbed uinder '-Materials and Metheds." Preparation of .samplles for electron mict 05(op was as described in Ref. 9. The bar re)resents 1 am. the a;iou p)oints to the jLnction 1etween the polx-genomic-lenlgth tail aniic the coX-lenigth circle. fraction of slow migrating forms increased -3-folcl and could be detectable by ethidium bromide staining of gels. In the absence of ligase and with onlx polymerase I preseiit, these fornms predominated, re)resenting 60'(% or more of the radioacti itv. Electron microscopfic analsis showed these forms to be duplex, oX-unit-length, relaxed circles with long duplex tails (Fig. 3). Of 43 molecules examined, the linear tail was up to 4 genomes in length in 18, between 5 and 10 genome lengths in 21, and up to 13 to 19 genonme units long in the remainder. The significance of these complex forms will be considered in the "L)iscussion." S'upercotling of RF I Product by D.NA Gvrase A preparation of purified E. colt gy rase did not act directlv in the SS to RF reaction mixture, a failure likelv due to an inhibitory factor. In this instance, the problemi was overcome by using a cruder enzvnme fraction as the source of gvrase. A portion of 'In sub)seuen-eiit studies (32), another purified gyrase preparation prioxved to he nulb active in supercoiling the REF I product. Fraction II (10 jg of protein) prepared as previouslx described (14) (channels 1 and31). Novot)iocini (100 otg/nml) and oxollicic acid (30 ,ug nil) were )resent in the reaction mixtures showin in channels 2 and 4. respectixelx AlaUers aire RF I and HRF II I)NAs extractedc from infected cells. Agarose gel analxsis was as described under "Materials aind MNethods---  Fig. 1), in 50 mM Tris Cl (pH .6), 20 mm KCI, a> mM dithiothreitol, 7 mM MgCI. t-Imi spernmidine Cl, 1.5 miN ATP, and 50 tpg of boxvine serum albumini5was treated as indicated.
F. coll I)NA gxrase, when used as indicated in the figure. was added as described under "Materials and Methods," either 45 nmimi prior to or simultaneouislv with additioni of HF SS reactioll com-iponenits the product of the SS to RF reaction (containing DNA polymerase I and ligase) was converted to the supercoiled HF I (Fig.  4, channels 1 and 3). Inhibition of this conversion by novobiocin (100 [ig/ml) or oxolinic acid (30 pg/ml) is indicatixe of gyrase action (30, 31) (Fig. 4, channels 2 and 4). Purified E. coli DNA gyrase converted the relaxed RF 1, isolated from the SS to RF incubation mixture bv agarose gel electrophoresis, to the supercoiled form and thus rendered it competent as a template and substrate for svnthesis of single-stranded xiral circles (see below).
Synthetic RF I DNA a,s a Temnplate-Siubstrate for Sy,nthesis of Single -stranded Viral Circles Inertness of Sxynthetic Relaxed RF I is Reliev,ed bx, DNA Gvrase-When the purified sv,nthetic relaxed RF 1 DNA was used in an RF o SS reaction (reconstituted with rep protein, OX gene A protein, SSB, and DNA polvmerase III holoenzyme), synthesis of single-stranded viral circles was not detected (Fig. 5, A and B, channel 6). Supplementing the reaction mixture with pure E. cali DNA gyrase resulted in the production of single-str-anded circles (Fig. 5, A and B, chann2els 2 and 5). The kinetics of the RF SS reaction using synthetic RF I DNA supercoiled prior to its addition were not distinguishable from those of a reaction in which the RF I DNA had been obtained from OX-infected cells (Fig. 5, A and B,   channel 7). Based on nucleotide incorporation, an average of three single-stranded circles were formied/molecule of input RF I. Gel electrophoretic analvsis (Figs. 5A and 6) and electron microscopy (not shown) demonstrated the production of single-stranded DNA circles; autoradiographv of gels showed that more than 95% of the radloactivity was incorporated into single-stranded DNA circles the size of OX DNA (Fig. SB).
Single-sttanded Circle Produiction Depends on Rep Protein and Gene A Protein-In the absence of rep protein, the synthetic RF I DNA, supercoiled by DNA gyrase, was relaxed by the nicking activity of gene A protein (Fig. 5A, channel 3);  Fig. IE) (2,50 pmol as niucleotide) was ilncubated with DNA gvrase ("Materials and Methods"-) in a 20-pl reaction mixture containinig 50 mmi Tris CI (IpH 7.6). 20 imnM KCl, 5 om (lithiothreitol, i mM MgCl, 5 nM spermidine Cl, 1.5 mM ATP. aincd 50 uig/ml of bovine serum albumin. After 45 min at 30°C components of the HF to SS reaction and dltTPase were added (see "Materials and Methods"). The 37-pl reaction mixture was then incubated at 3))`C anid at itndicated tinmes, aliquots were examined for DNA sxnthesis Photo insert. a sample after a 45-mni iiicubation was analxzed bx agariseethidium gel electrophoresis as under "Materials and Metholds.-" D.NA markers are HF I anid RF ii extractecd from oX-infected cells and single-stranded circular DNA extracted fromi phage (12).
in the absence of gene A protein, the supercoiled portion of synthetic RF I DNA renmained supercoiled (Fig. 5A, channel  4). Incorporation of radioactixvitx into a single-stranded, circular product was not observed in the absence of either rep oi'  I Transfecting activity of synthetic .X DNA Synthetic OX DNA circles were prepared as follows. Phage-extracted DNA was replicated in an SS --RF reaction with addition of DNA polymerase I and E. coli DNA ligase. Reactions were stopped by addition of EDTA, sodium dodecyl sulfate, glycerol, and bromphenol blue to final concentrations of 50 mM, 1, 10, and 0.01%, respectively; DNA products were electrophoresed in 0.8% agarose gels containing 0.1% (w/v) ethidium bromide. Covalently sealed doublestranded molecules, supercoiled by ethidium bromide, were isolated from the gel (28). E. coli DNA gyrase was added (as under "Materials and Methods") to 300 pmol (as nucleotide) of the synthetic, relaxed RF I DNA in a reaction mixture containing 50 mM Tris. Cl (pH 7.6), 20 mM KCI, 5 mM dithiothreitol, 7 mM MgC12, 5 mm spermidine-Cl, 1.5 mM ATP, and 50 gg/ml of bovine serum albumin; the reaction was incubated at 30 'C for 45 min. Components of the RF -> SS reaction were added and the mixture incubated at 30 'C for an additional 30 min. Reactions were stopped by EDTA and sodium dodecyl sulfate as above, and the DNA isolated either by agarose gel electrophoresis or by extraction with hot phenol (12). Spheroplasts from E. coli W3350 were used except where E. coli BD1154 is indicated. E. coli CR (Su+) was the indicator strain for determining specific transfectivity; E. coli C was used to determine am3 revertants. Additional details are found under "Materials and Methods." gene A proteins. Transfecting Activity of Synthetic RF I and Singlestranded Circles Synthesized from It-Single-stranded circles, synthesized from RF I DNA isolated from OX-infected cells, and using the RF -s SS enzyme system, are biologically active in transfection; specific transfectivity values were about two orders of magnitude higher than those for RF I (33) ( Table I). Synthetic RF I obtained by in vitro conversion of viral circles was as transfective as RF I isolated from @Xinfected cells. When used as a template-substrate, this synthetic RF I directed the synthesis of transfective singlestranded circles. Specific transfectivities of single-stranded viral DNA synthesized on synthetic, complementary strand templates were comparable to those of DNA synthesized on a OX RF I template isolated from infected cells (see above).
Progeny phages produced in spheroplasts transfected with OX DNA carrying the am3 mutation were plated on a nonsuppressor strain to detect revertants. The rate of reversion of the am3 mutation in viral circles synthesized from synthetic RF I was similar to that observed in circles synthesized from cell-extracted RF I (Table I). DISCUSSION Initiation of DNA replication of the single-stranded circle of infecting pX174 DNA by E. coli enzymes has been taken as a model for initiation of Okazaki fragments in the discontinuous replication of the lagging strand in replication of the E. coli chromosome (34). Isolation and studies of the 12 proteins required for the apparently simple conversion of a singlestranded circle to the duplex replicative form have made it possible to examine their functions and to group their actions into stages of prepriming, priming, elongation, and termination.
In the absence of DNA polymerase I, the terminal stage of forming a supercoiled RF does not take place. The combined actions of single-stranded DNA binding protein, the prepriming proteins (i, n, n', n", dnaB, and dnaC), primase, and DNA polymerase III holoenzyme produce an incomplete complementary strand. Annealed to the circular viral DNA template, the complementary strand in this replicative form (RF II) is nearly full length as judged by its electrophoretic mobility. The size of the gap in the RF II has not been determined. It is judged to be too short to sustain the processive action of DNA polymerase III holoenzyme, but appropriate for the gapfilling action of DNA polymerase I. The location of the gap, inferred from the distribution of primers, is at many places around the circle.
The actions of DNA polymerase I in completing the replication of the viral template and in filling the gap created by its 5' -* 3' exonucleolytic removal of the 5' terminal RNA primer must precede the sealing action of E. coli DNA ligase. The covalently complete duplex circle becomes the substrate for supercoiling by E. coli DNA gyrase and conversion to a form active in RF replication.
The studies reported here have demonstrated a possible pathway for the terminal stages of forming supercoiled RF and have shown that the synthetic product and the synthetic single-stranded DNA circles synthesized using this synthetic product as a template are indistinguishable from the forms isolated from phages and infected cells. The behavior of the synthetic RF I as a template-substrate in the RF -* SS reaction and the biological activity of both the synthetic RF I and the single-stranded product in transfection of spheroplasts prove that the enzymatic actions in vitro produce the same, high fidelity products as in vivo. These conclusions are based on measurements of the specific transfectivity values and the reversion rate of the am3 mutation (reversal of a G > A transition at position 587 (35) that converts the TGG codon for trytophan to a TAG terminator) of the synthetic complementary strand and the viral strand synthesized on this complementary strand as template.
Complex molecular forms are produced in the absence of DNA ligase at the terminal stage. Duplex circles with tails many genomes in length develop apparently as the result of DNA polymerase I action unchecked by ligase; presumably, extensive displacement of the complementary strands from its 5' end supervenes over exonucleolytic digestion during processive polymerization. Should DNA polymerase switch templates from the viral circle to the displaced complementary strand, a linear duplex tail would be formed. Another possibility is that initiations take place on the displaced single strand despite its supposed lack of precisely the same recognition region.
In this and the succeeding paper (11), we report that the stages of SS -> RF and RF --SS as discrete or linked reactions can be reconstituted from highly purified proteins. Aberrations are observed when impurities are present or when any of the reaction components are omitted. It had been anticipated that when more exacting criteria of optimal rates and efficient coupling were imposed, the importance of new factors and conditions would be disclosed. Recent studies (36,37) have in fact revealed that isolation of the synthetic, covalently closed duplex (synthetic parental RF) under native conditions conserves a bound complex of priming proteins (primosome) that obviates the need for gyrase, potentiates the RF for gene A protein cleavage, and is nearly adequate to sustain complementary strand synthesis in the formation of progeny RF.