Studies on the role of the phi X174 gene A protein in phi X viral strand synthesis. I. Replication of DNA containing an alteration in position 1 of the 30-nucleotide icosahedral bacteriophage origin

The influence of a C----G transversion at position 1 of the 30-base pair replication origin of bacteriophage phi X174 replicative form I DNA (phi X RFI) was examined in the RF----single-stranded circular DNA replication pathway catalyzed by the combined action of the purified phi X A protein, the Escherichia coli DNA polymerase III holoenzyme, rep helicase, and single-stranded DNA binding protein (Eisenberg, S., Scott, J.F., and Kornberg, A. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 1594-1597; Reinberg, D., Zipursky, S.L., and Hurwitz, J. (1981) J. Biol. Chem. 256, 13143-13151). RFI DNA containing this transversion was cleaved to RFII by the phi X A protein as effectively as DNA containing the wild-type origin. The altered duplex DNA, however, supported replication at a slower rate (3- to 4-fold) than the wild-type DNA due to a defect in the termination and reinitiation reactions catalyzed by the phi X A protein. This defect resulted in the accumulation of DNA products containing long single strands covalently joined to the mutant DNA. These single strands were susceptible to nuclease S1 and exonuclease VII attack. The defect in the template DNA containing C----G transversion was not corrected when this mutant origin was placed on the same strand with a wild-type origin. This double-origin DNA was also replicated poorly and led to the accumulation of large products, in contrast to the products formed with RFI DNA containing two wild-type 30-base pair replication origins on the same strand.     

DNA sequences which support activities of the bacteriophage phi X174 gene A protein

The DNA sequence of 30 nucleotides which surrounds the origin of viral strand DNA replication is highly conserved amongst the icosahedral single-stranded DNA bacteriophages. The A gene of these phages encodes a protein which is required for initiation and termination of viral strand DNA synthesis and acts as a nicking-closing activity specifically within this 30-nucleotide sequence. A system of purified Escherichia coli host proteins and phi X174 gene A protein has been developed which specifically replicates in vitro the viral strand of phi X174 from RF (replicative form) I template DNA and yields single-stranded circular DNA products (RF leads to SS(c) DNA replication system). Recombinant plasmids carrying inserts derived from phage phi X174 or G4 DNA which range in length from 49 to 1175 base pairs and contain the 30-nucleotide conserved sequence have been shown to support phi X A protein-dependent DNA synthesis in vitro in this replication system. We report here that insertion of the 30-nucleotide sequence alone into pBR322 allows the resulting recombinant plasmids to support phi X A protein-dependent in vitro DNA synthesis as efficiently as phi X174 template DNA in the RF leads to SS(c) replication system. The 30-nucleotide sequence functions as a fully wild type DNA replication origin as determined by the rate of DNA synthesis and the structure of resulting DNA products. Furthermore, the DNA sequence requirements for nicking of RF I DNA by the phi X A protein and for supporting replication origin function have been partially separated. Homology to positions 1, 29, and 30 of the 30-nucleotide conserved sequence are not required for cleavage of RF I DNA by the A protein; homology to position 1 but not 29 or 30 is required for efficient DNA replication.     

Purification of a DNA primase activity from the yeast Saccharomyces cerevisiae. Primase can be separated from DNA polymerase I

A primase activity which permits DNA synthesis by yeast DNA polymerase I on a single-stranded circular phi X174 or M13 DNA or on poly(dT)n has been extensively purified by fractionation of a yeast enzyme extract which supports in vitro replication of the yeast 2-microns plasmid DNA (Kojo, H., Greenberg, B. D., and Sugino, A. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 7261-7265). Most of this DNA primase activity was separated from DNA polymerase activity, although a small amount remained associated with DNA polymerase I. The primase, active as a monomer, has a molecular weight of about 60,000. The primase synthesizes oligoribonucleotides of discrete size, mainly eight or nine nucleotides, in the presence of single-stranded template DNA and ribonucleoside 5'-triphosphates; it utilizes deoxyribonucleoside 5'-triphosphates as substrate with 10-fold lower efficiency. Product size, chromatographic properties, alpha-amanitin resistance, and molecular weight of the primase activity distinguish it from RNA polymerases I, II, and III. The DNA products synthesized by both primase and DNA polymerase I on a single-stranded DNA template were 200-500 nucleotides long and covalently linked to oligoribonucleotides at their 5'-ends. Addition of yeast single-stranded DNA-binding protein (Arendes, J., Kim, K. C., and Sugino, A. (1983) Proc. Natl. Acad. Sci. U.S. A. 80, 673-677) stimulated the DNA synthesis 2-3-fold.     

Effects of genome size on bacteriophage phi X174 DNA packaging in vitro

Effects of the size of template DNA on the DNA packaging reaction of bacteriophage phi X174 were studied using plasmids of various sizes which contain the phi X174 origin of DNA replication and the in vitro phage synthesizing system (Aoyama, A., Hamatake, R. K., and Hayashi, M. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 4195-4199). DNA between 78.5% and 101% of the length of phi X174 DNA produced infectious particles efficiently. Packaging of DNA smaller or larger than this range produced uninfectious defective particles. Although these particles contained circular single-stranded DNA, they suffered structural changes which altered the sedimentation properties or the ability to adsorb to the cells. Mutant phage were found from the packaging reaction of DNA larger than 101% of phi X174 DNA. These mutants deleted the termination region of DNA, suggesting that they were produced by early termination of the phage synthesizing reaction.     

Analysis of bacteriophage phi X174 gene A protein-mediated termination and reinitiation of phi X DNA synthesis. I. Characterization of the termination and reinitiation reactions

The phi X174 (phi X) gene A protein-mediated termination and reinitiation of single-stranded circular (SS(c] phi X viral DNA synthesis in vitro were directly and independently analyzed. Following incubation together with purified DNA replication enzymes from Escherichia coli, ATP, [alpha-32P]dNTPs, and either the phi X A protein and phi X replicative form I (RF I) DNA, or the purified RF II X A complex, the phi X A protein was detected covalently linked to newly synthesized 32P-labeled DNA. Formation of the phi X A protein-[32P]DNA covalent complex required all the factors necessary for phi X (+) SS(c) DNA synthesis in vitro. Thus, it was a product of the reinitiation reaction and an intermediate of the replication cycle. Identification of this complex provided direct evidence that reinitiation of phi X (+) strand DNA synthesis involved regeneration of the RF II X A complex. Substitution of 2',3'-dideoxyguanosine triphosphate (ddGTP) for dGTP in reaction mixtures resulted in the formation of covalent phi X A protein 32P-oligonucleotide complexes; these complexes were trapped analogues of the regenerated RF II X A complex. They could not act catalytically due to the presence of ddGMP residues at the 3'-termini of the oligonucleotide moieties. Reaction mixtures containing ddGTP also yielded nonradioactive (+) SS(c) DNA products derived from circularization of the displaced (+) strand of the input parental template DNA. The formation of the phi X A protein-32P-oligonucleotide complexes and nonradioactive (+) SS(c) DNA were used to assay both reinitiation and termination reactions, respectively. Both reactions required DNA synthesis from the 3'-hydroxyl primer at nucleotide residue 4305 which was formed by cleavage of phi X RF I DNA by the phi X A protein. Elongation of this primer by 18, but not 11 nucleotides was sufficient to support each reaction. Reinitiation reactions proceeded rapidly and were essentially complete after 90 s. In contrast, when ddGTP was replaced with dGTP in reaction mixtures, DNA synthesis proceeded with linear kinetics for up to 10 min. These results suggested that in the presence of all four dNTPs, active templates supported more than 40 rounds of DNA synthesis.     

Rep protein as a helicase in an active, isolatable replication fork of duplex phi X174 DNA

Rep protein as a helicase combines its actions with those of gene A protein and single-stranded DNA binding protein to separate the strands of phi X174 duplex DNA and thereby can generate and advance a replication fork (Scott, J. F., Eisenberg, S., Bertsch, L. L., and Kornberg, A. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 193-197). Tritium-labeled rep protein is bound in an active gene A protein. phi X174 closed circular duplex supercoiled DNA complex in a 1:1 ratio. Catalytic separation of the strands of the duplex by rep protein, as measured by incorporation of tritium-labeled single-stranded DNA binding protein, requires ATP at a Km value of 8 microM, and hydrolyzes two molecules of ATP for every base pair melted. When coupled to replication in the synthesis of single-strand viral circles, a "looped" rolling-circle intermediate is formed that can be isolated in an active form containing gene A protein, rep protein, single-stranded DNA binding protein, and DNA polymerase III holoenzyme. Unlike the binding of rep protein to single-stranded DNA, where its ATPase activity is distributive, binding to the replicating fork is not affected by ATP, further suggesting a processive action linked to gene A protein. Limited tryptic hydrolysis of rep protein abolishes its replicative activity without affecting significantly its binding of ATP and its ATPase action on single-stranded DNA. These results augment earlier findings by describing the larger role of rep proteins as a helicase, linked in a complex ith other proteins, at the replication fork of a duplex DNA.     

The bacteriophage lambda O and P protein initiators promote the replication of single-stranded DNA.

A soluble enzyme system that specifically initiates lambda dv plasmid DNA replication at a bacteriophage lambda replication origin [Wold et al. (1982) Proc. Natl. Acad. Sci. USA 79, 6176-6180] is also capable of replicating the single-stranded circular chromosomes of phages M13 and phi X174 to a duplex form. This chain initiation on single-stranded templates is novel in that it is absolutely dependent on the lambda O and P protein chromosomal initiators and on several Escherichia coli proteins that are known to function in the replication of the lambda chromosome in vivo, including the host dnaB, dnaG (primase), dnaJ and dnaK replication proteins. Strand initiation occurs at multiple sites following an O and P protein-dependent pre-priming step in which the DNA is converted into an activated nucleoprotein complex containing the bacterial dnaB protein. We propose a scheme for the initiation of DNA synthesis on single-stranded templates in this enzyme system that may be relevant to strand initiation events that occur during replication of phage lambda in vivo.     

The J gene of bacteriophage phi X174: in vitro analysis of J protein function.

The J protein of phi X174 is a small, highly basic protein and is a component of the phage capsid. We have investigated the role of J protein during single-stranded viral DNA synthesis and phage morphogenesis by using an in vitro system composed of purified viral and host components (Aoyama et al., Proc. Natl. Acad. Sci. U.S.A. 80:4195-4199, 1983). The characterization of the products made in the presence and absence of J protein shows that J protein is not required for viral DNA synthesis, but is required for the packaging of DNA into infectious phage. The ability of J protein to bind to double-stranded DNA as well as single-stranded DNA and other interactions with DNA suggest a model in which J protein binds to double-stranded, replicative form DNA and enters the phage prohead by remaining bound to viral DNA as it is encapsidated.     

Presynapsis and synapsis of DNA promoted by the STP alpha and single-stranded DNA-binding proteins from Saccharomyces cerevisiae

We previously purified an activity from meiotic cell extracts of Saccharomyces cerevisiae that promotes the transfer of a strand from a duplex linear DNA molecule to complementary circular single-stranded DNA, naming it Strand Transfer Protein alpha (STP alpha) (Sugino, A., Nitiss, J., and Resnick, M. A. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 3683-3687). This activity requires no nucleotide cofactor but is stimulated more than 10-fold by the addition of yeast single-stranded DNA-binding proteins (ySSBs). In this paper, we describe the aggregation and strand transfer of double-stranded and single-stranded DNA promoted by STP alpha and ySSB. There is a good correlation between the aggregation induced by various DNA-binding proteins (ySSBs, DBPs and histone proteins) and the stimulation of STP alpha-mediated DNA strand transfer. This implies that the stimulation by ySSBs and other binding proteins is probably due to the condensation of single-stranded and double-stranded DNA substrates into coaggregates. Within these coaggregates there is a higher probability of pairing between homologous double-stranded and single-stranded DNA, favoring the initiation of strand transfer. The aggregation reaction is rapid and precedes any reactions related to DNA strand transfer. We propose that condensation into coaggregates is a presynaptic step in DNA strand transfer promoted by STP alpha and that pairing between homologous double- and single-stranded DNA (synapsis) occurs in these coaggregates. Synapsis promoted by STP alpha and ySSBs also occurs between covalently closed double-stranded DNA and single-stranded linear DNA as well as linear double-stranded and linear single-stranded DNAs in the absence of any nucleotide cofactors.     

Purification of the rep protein of Escherichia coli. An ATPase which separates duplex DNA strands in advance of replication.

The product of the rep gene of Escherichia coli catalytically separates phiX174 duplex DNA strands in advance of their replication, utilizing ATP in the process (Scott, J. F., Eisenberg, S., Bertsch, L. L., and Kornberg, A. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 193-197). The enzyme has now been purified to near-homogeneity. Relatively large quantities were obtained from ColE1-plasmid-containing cells in which the enzyme level was 7 to 10 times above wild type. The assay for rep protein was based on its essential role, with phage-induced cistron A protein, in enzymatic synthesis of phage phiX174 (+) strands, using duplex circular DNA as template. The protein exhibits a molecular weight of 65,000 under denaturing and reducing conditions. The turnover number of the enzyme is approximately 6800 ATP molecules/min in strand separation as measured by extent of replication, or in an uncoupled reaction using single-stranded DNA effector.     

On the fidelity of DNA replication. The accuracy of Escherichia coli DNA polymerase I in copying natural DNA in vitro

The accuracy with which Escherichia coli DNA polymerase I (Pol I) copies natural DNA in vitro has been determined. When phi X174 viral DNA containing an amber mutation (am3) is primed with a single restriction endonuclease fragment, copied in vitro with Pol I and then expressed in E. coli spheroplasts (Weymout, L. A., and Loeb, L. A. (1978) Proc. Natl. Acad. Sci. U. S. A. 75, 1924), the reversion frequency of this DNA is greater than that of uncopied DNA. This change in reversion frequency can be increased by selectively increasing the concentration of either dATP or dCTP relative to the other deoxyribonucleotide substrates. DNA sequence analyses of revertants obtained from substrate pool bias experiments demonstrates that the revertants contain the selectively biased nucleotide as an incorrect substitution at position 587 of the am3 codon. We have analyzed the product of the in vitro Pol I reaction using neutral and alkaline sucrose gradients. Fifty per cent of the input phi X174 DNA template molecules are copied past the am3 site. The phenotypic expression of the product (revertant) strand in the spheroplast assay was estimated using a model heteroduplex molecule similar in structure to the product of the reaction and containing a single base mismatch (A:A or A:C) at position 587. Using these data, and by extrapolation from pool bias experiments, we estimate the error rate of Pol I in Mg2+-activated reactions using equimolar concentrations of the four deoxynucleotide substrates is 1/680,000 for an A:C mispair and < 1/6,300,000 for an A:A mispair at position 587 of the am3 codon in phi X174 DNA.     

A prepriming DNA replication enzyme of Escherichia coli. II. Actions of protein n': a sequence-specific, DNA-dependent ATPase

Protein n' of Escherichia coli is required for formation of the prepriming complex in replication of the single-stranded circle of phiX174 DNA. The protein, purified to near homogeneity, possesses ATPase (dATPase) activity in the presence of single-stranded, but not duplex, DNAs. Except for phiX174 DNA, ATPase activity is completely suppressed by coating the DNA with single strand binding protein (SSB). phiX174 DNA possesses a unique sequence with a potential hairpin structure that is recognized as an effector (Shlomai, J., and Kornberg, A. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 799-803). Sequences with secondary structure in SSB-coated M13 DNA which are recognized by RNA polymerase, and in coated G4 DNA by primase, are inert for protein n'. Approximately 30 of the 180 molecules of SSB bound to phiX DNA are destabilized by protein n' in an ATP-dependent reaction. These actions by protein n' may be important in recognizing an origin for forming the prepriming complex that leads to initiation of phiX complementary strand synthesis.     

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.     

The DNA polymerase-primase from drosophila melanogaster embryos. Rate and fidelity of polymerization on single-stranded DNA templates

The DNA polymerase activity of the near homogeneous, multisubunit DNA polymerase-primase from Drosophila melanogaster embryos has been compared to Escherichia coli DNA polymerase III core, DNA polymerase III, and DNA polymerase III holoenzyme. The rate of deoxynucleotide incorporation by the Drosophila polymerase on singly primed phi X174 DNA is similar to that observed with equivalent levels of DNA polymerase III holoenzyme in the absence of E. coli single-stranded DNA binding protein. However, analysis of the DNA products indicates that the Drosophila polymerase is less processive than DNA polymerase III holoenzyme, and closely resembles DNA polymerase III. The Drosophila polymerase-primase contains neither 3'-5' exonuclease nor RNase H-like activities, and catalyzes no significant pyrophosphate exchange. There is a low level of DNA-dependent ATPase activity which can be eliminated by a second glycerol gradient sedimentation (Kaguni, L.S., Rossignol, J.-M., Conaway, R.C., and Lehman, I.R. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 2221-2225). Although lacking a 3'-5' exonuclease, the replication fidelity of the D. melanogaster polymerase is similar to that of E. coli DNA polymerase III holoenzyme which possesses such an activity.     

Signal sequence of alkaline phosphatase of Escherichia coli.

The amino acid sequence of the signal sequence of phoA was determined by DNA sequencing by using the dideoxy chain termination technique (Sanger et al., Proc. Natl. Acad. Sci. U.S.A. 74:5463-5467, 1977). The template used was single-stranded DNA obtained from M13 on f1 phage derivatives carrying phoA, constructed by in vitro recombination. The results confirm the sequence of the first five amino acids determined by Sarthy et al. (J. Bacteriol. 139:932-939, 1979) and extend the sequence in the same reading frame into the amino terminal region of the mature alkaline phosphatase (Bradshaw et al., Proc. Natl. Acad. Sci. U.S.A., 78:3473-3477, 1981). As was predicted (Inouye and Beckwith, Proc. Natl. Acad. Sci. U.S.A. 74:1440-1444, 1977), the signal sequence was highly hydrophobic. The alteration of DNA sequence was identified for a promoter mutation that results in the expression of phoA independent of the positive control gene phoB and in insensitivity to high phosphate.     

In vitro repair of psoralen-DNA cross-links by RecA, UvrABC, and the 5'-exonuclease of DNA polymerase I

Psoralens produce DNA interstrand cross-links which are thought to be repaired via a sequential excision and recombination mechanism in Escherichia coli. The first round of incision by UvrABC has been characterized: it results in 11-base oligonucleotide cross-linked to an intact DNA strand (Van Houten, B., Gamper, B., Holbrook, S.R., Hearst, J.E., and Sancar, A. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 8077-8081). In the present work, DNA substrates containing 4'-hydroxymethyl-4,5',8-trimethylpsoralen (HMT) cross-links in defined positions are constructed and used to analyze the other steps in repair. It is shown that RecA protein mediates strand transfer past an oligonucleotide cross-linked to a single-stranded DNA circle and that the resulting heteroduplex is a substrate for the UvrABC complex: it excises a double-stranded oligonucleotide which contains the HMT cross-link. It is also found that the first round of UvrABC incision does not lead directly to strand exchange but that an intervening step is needed. That step is carried out in vitro by the 5'-exonuclease activity of DNA polymerase I (pol I) which creates a single-stranded DNA region (a gap) at an incised cross-link such that RecA can initiate strand exchange. Studies using cross-linked oligonucleotides showed that the gap produced by pol I results from the inability of the polymerase to add nucleotides to a 3'-OH end two to three nucleotides away from the furan side of an HMT cross-link. Pol I can, however, extend a 3'-OH end next to the pyrone side of the cross-link. Since UvrABC incises predominantly the furan side of psoralen cross-links in duplex DNA, this discrepancy has important consequences for repair.     

DNA strand transfer protein beta from yeast mitotic cells differs from strand transfer protein alpha from meiotic cells

We have purified to homogeneity an activity from mitotic cell extracts of the yeast Saccharomyces cerevisiae, which promotes the transfer of a strand from a duplex linear DNA molecule to a complementary circular single strand. This activity does not require any nucleotide cofactor and is greatly stimulated by yeast single-stranded DNA-binding protein. It consists of a single polypeptide of an apparent molecular mass of 180 kDa as determined by SDS-polyacrylamide gel electrophoresis. This activity, which we call DNA strand transfer protein beta (STP beta), has reaction properties similar to those of DNA strand transfer protein alpha (STP alpha) purified from crude extracts of yeast meiotic cells (Sugino, A., Nitiss, J., and Resnick, M. A. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 3683-3687). However, STP beta differs from STP alpha in its molecular weight and column chromatographic behavior as well as by immunological comparison. Furthermore, the STP beta polypeptide remains in cells in which the STP alpha gene has been disrupted. Thus, we conclude the STP beta activity is encoded by a gene different from that for STP alpha. Although STP beta was isolated from mitotic cells, the amount of STP beta increases severalfold during meiosis. STP beta also appears to differ in molecular weight from similar activities described by other groups and may be an intact form of their activities.     

Deletion analysis of the DNA sequence required for the in vitro initiation of replication of bacteriophage lambda

Supercoiled DNA containing the replication origin of bacteriophage lambda can be replicated in vitro. This reaction requires purified lambda O and P replication proteins and a partially purified mixture of Escherichia coli proteins (Tsurimoto, T., and Matsubara, K. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 7639-7643; Wold, M. S., Mallory, J.B., Roberts, J. D., LeBowitz, J. H., and McMacken, R. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 6176-6180). The lambda origin region has four repeats of a 19-base pair sequence to which O protein binds. To the right of these sites on the lambda map is a 40-base pair region that is rich in adenine and thymine, followed by a 28-base pair palindromic sequence. To define more precisely the boundaries of the lambda origin, we cloned a 358-base pair piece of lambda DNA containing the origin region into M13mp8 in both orientations. In vitro replication of RF I DNAs prepared from cells infected with these two M13 ori lambda phage was dependent on lambda O and P proteins and a crude protein fraction from uninfected E. coli; with these conditions there was no replication of M13mp8 RF I DNA. We made deletions from the left and the right ends of the lambda origin DNA and determined the deletion end points by DNA sequencing. We have tested RF I DNAs prepared from cells infected with phage carrying ori lambda deletions for their ability to function as templates for O- and P-dependent replication in vitro. Our results show that lambda DNA between nucleotide positions 39072 and 39160 is required for efficient O- and P-dependent replication. This 89-base pair piece of DNA includes only two of the four 19-base pair O protein-binding sites (the two right-most) and the adjoining adenine- and thymine-rich region to the right of the O-binding sites.     

Fidelity of replication of bacteriophage phi X174 DNA in vitro and in vivo

Seven different revertants of bacteriophage phi X174am16 (AB5276G leads to T) have been isolated and the nature of the reversions determined by sequencing their DNA. The revertants each differ from am16 by just a single base substitution. These may be distinguished with varying degrees of ease by characteristic temperature sensitivities of growth. This has facilitated the determination of the frequency at which DNA polymerase III catalyses different types of substitution mutations in copying phi X174 DNA in vitro and in vivo. During the replicative form (RF) leads to single-stranded (SS) stage of replication in vitro, four different revertants may be readily produced according to well-defined rate laws on biasing the concentrations of dNTPs. Transversion mutations are found to be formed predominantly by purine x purine mismatching, whilst transitions are formed predominantly by G x T mismatching. The substitutions via G x T and G x A mismatches are estimated to occur at similar frequencies in vivo. The two most common revertants isolated in vivo, however, are not those readily produced during the RF leads to SS stage in vitro but are those produced on purine x purine mismatching in the SS leads to RF stage. The accuracy of the DNA polymerase in vitro appears to be similar to that in this stage in vivo. However, the overall accuracy of the RF leads to SS replication in vivo is more accurate than predicted from the measurements of the accuracy in vitro.