Deoxyribonucleic Acid Synthesis in Cell-free Extracts

DNA polymerase II requires a primer to initiate synthesis on synthetic, single stranded deoxypolynucleotides. The product of the reaction is complementary to the template and is covalently attached to the primer. The frequency of errors in incorporation is less than one nucleotide in 105. Repairtype synthesis occurs in the 5’ to 3’ direction and proceeds until a complete duplex structure is achieved. DNA polymerase II possesses exonuclease activity which is specific for single stranded DNA. The direction of attack is exclusively 3’ to 5’.

The isolation of a mutant of Escherichia co&, (Pol A;-), lacking detectable levels of DNA polymerase activity (DNA polymerase I) (l), stimulated several investigations into the nature of the DNA synthesis capacity of these cells (2)(3)(4).
Several laboratories have described the purification and initial characterization of DNA polymerase II, an enzyme distinct from DNA polymerase I (5-10).
DNA polymerase II can be distinguished from DNA polymerase I by several criteria, including its sensitivity to thiol reagents, to ionic strength, and to antiserum directed against DNA polymerase I. Compared with DNA polymerase I, the specific activity of pure DNA polymerase II is less than 10% and there are fewer than 25% as many molecules per cell (6). DNA polymerase II does, however, bear strong resemblance to DNA polymerase I with regard to its catalytic properties.
Like DNA polymerase I, DNA polymerase II catalyzes the templatedirected synthesis of DNA in the 5' to 3' direction by covalent attachment of the product to the primer.
DNA polymerase II * The 3rd paper in this series is Reference 7. This investigation was aided by Grants GMCA 18943.01 and ITOl-GM-02087-01 from the United States Public Health Service. also catalyzes the exonucleolytic degradation of single-stranded DNA from the 3' end; but unlike DNS polymerase I, hydrolysis of DNA from the 5' end is not observed (7, 10).
The physiological roles of both DNA polymerases I and II remain obscure.
The availability of mutants defective in DNA polymerase I has not helped to elucidate the role of this enzyme in DNA m&abolism (11)(12)(13). Evaluation of the biological significance of both DNA polymerases I and II must await further analysis.
During our investigations into the DNA synthetic capacity of Pol A1 cells, another DNA-synthesizing activity (DNA polymerase III) was observed (6). On the basis of its chromatographic behavior, inhibition by salt, thermal stability, and insensitivity to antiserum directed against DNA polymerase I, DNA polymerase III can be distinguished from DNA polymerases I and II. Furthermore, analysis of mutants temperature sensitive for DNA synthesis and for cell viability indicates that DNA polymerase III is essential for DNA replicat'ion (14,15). In this report we describe the purification and the general catalytic properties of DNA polymerase III, and compare these properties with those of DNA polymerases I and II. MATERIALS Nucleotides and Polymers-Unlabeled deoxynucleoside 5'triphosphates were purchased from Sigma and [3H]TTP (1'7.8 Ci per mmole) from Schwarz BioResearch.
Dr. F. Bollum kindly provided poly(dA) and poly(dC). The synthetic products, (dA-dC)4, (dG-dT)n, and (pdT)lo, were a gift from Drs. I. Molineux  E. coli DNA polymerase II (Fraction V, 270 units per mg) was prepared as previously described (6). Exonuclease III (180,000 units per mg) was isolated as previously described (6) and further purified by a method adapted from that of Richardson and Kornberg (18).
Miscellaneous-E. coli W3110 thy-, &a-, Incubations were for 5 min at 30"; nucleotide incorporation into acid-insoluble product was measured as described (5). One unit of enzyme is defined as the amount catalyzing the incorporation of 1 nmole of TTP into acid-insoluble product in 5 min at 30". Preparation of Template for DNA Polymerase III Reaction-Suitable template ("gapped" DNA) was prepared by the sequential action of DNase I and exonuclease III.
The reaction mixture (10 ml) contained 5.1 M Tris-acetate (pH 8.2); 2 mM MgC&; 3 mM 2-mercaptoethanol; 40 pmoles of calf thymus DNA; and 0.3 pg of DNase I. After a 20-min incubation at 30", DNase was inactivated by heating at 65" for 10 min. An appropriate amount of exonuclease III, titrated to give maximal DNA polymerase activity, was added and allowed to incubate for 5 min at 30". The reaction was terminated by heating at 65" for 10 min.
Preparation of Nuclease-treated T-7 DNA-Digestion of T-7 DNA with DNase I was carried out in a reaction mixture (0.55 ml) containing 0.1 M Tris-acetate (pH 8.2); 2 mM MgC&; 4 mu 2-mercaptoethanol; 175 nmoles of T-7 DNA; and 2.5 ng of DNase I. Incubation was for 5 min at 30" followed by heating at 65" for 10 min. Digestion with micrococcal nuclease (2.5 ng) was in an identical reaction mixture except that 2 mu CaC12 replaced MgC12. Subsequent treatment with exonuclease III was performed in the reaction mixture as described for the assay of DNA polymerase III except that calf thymus DNA and ethanol were omitted.
Exonuclease III (10 units) and 16 nmoles of "activated" T-7 DNA were added, and after 5 min at 30", the reaction was terminated by heating for 10 min at 65".
Preparation of [32P]pdT(pdT)9-Unlabeled, chemically synthesized (pdT)lo was dephosphorylated with 0.2 unit of alkaline phosphatase in a reaction mixture (0.15 ml) containing 60 mM Tris-acetate (pH 8.2); and 50 nmoles of (pdT)l,,. After 30 min at 65" the incubation mixture was cooled to 0" and made 10% in trichloroacetic acid by the addition of cold 50 y0 trichloroacetic acid. After 30 min, the precipitate was removed by centrifugation and the supernatant extracted three times with ether. Phosphorylation was carried out in a reaction mixture (0.6 ml) containing 85 mM Tris-acetate (pH 7.5); 17 mM MgClz: 17 mM 2.mercaptoethanol; 8.5 mu potassium phosphate (pH 7.5); 17 mM [y-32P]ATP (5 x lo3 cpm per pmole) ; and 5.7 units of polynucleotide kinase. After 30 min at 37", an additional 5.7 units of polynucleotide kinase were added. After an additional 30 min, t,lle incubation mixture was heated at 100" for 5 min and applied to a column of Sephadex G-50 (1 X 110 cm) at 65". [""PI-pdT(pdT)s was eluted from the column after collecting 25% of the bed volume.
General-Protein was determined by the method of Bucher (19), with bovine serum albumin as the standard.
Salt concentration was measured using a conductivity bridge, with potassium phosphate as the standard.
Deoxynucleoside triphosphates were neutralized with M Tris buffer.  Mutants defective in DNA polymerase III with normal amounts of DNA polymerase II (14) have no measurable polymerase activity in the SlOO fraction.

PurQication
All steps were performed at 4" and all buffers contained 50 mM 2-mercaptoethanol and 20% glycerol (v/v). The purification is designed for 100 g of cell paste. A summary of the purific&tion is given in Table I. R&-Preparation of the SlOO cell-free extract was described (6). The SlOO (200 ml) was brought to 20% glycerol by the addition of glycerol (50 ml) and brought to 400 ml by the addition of 0.01 M potassium phosphate buffer, pH 6.5. DEAE-cellulose I-This procedure was performed as previously described (6). The diluted SlOO was brought to 0.2 M (NH&SO4 by the dropwise addition of saturated ammonium sulfate, previously neutralized with NH40H. The sample was applied to a column of DEAE-cellulose (7.1 x 10 cm) previously equilibrated with 0.01 M potassium phosphate buffer, pH 7.5, containing 0.2 M (NH&S04. The protein not adhering to the column was collected in a single fraction (420 ml).
Ammonium Sulfate-The DEAE-cellulose fraction was brought to 35% saturation by the addition of solid ammonium sulfate (73.5 g) over a 30-min period.
After an additional 30 min, the precipitate was removed by centrifugation at 17,000 X g. TO the supernatant, solid ammonium sulfate (38.7 g) was added over a 30-min period.
After an additional 30 min, the precipitate was collected by centrifugation.
The precipitate was washed once with 50% saturated ammonium sulfate and then dissolved in 0.02 M potassium phosphate buffer, pH 6.5 (Buffer A). The ammonium sulfate fraction (21.4 ml) was dialyzed for 10 hours against 1 liter of Buffer A and the dialysis buffer was changed once.
DEAE-cellulose II-The dialyzed ammonium sulfate fraction was applied to a column of DEAE-cellulose (4.1 x 30 cm) previously equilibrated with Buffer A. A linear gradient, 0.02 to 0.3 M potassium phosphate, pH 6.5 with a total volume of 2 liters, was applied immediately and 20 ml fractions were collected. DNA polymerase activity eluted in two peaks. The first peak, DNA polymerase II, eluted at a salt concentration of 0.  The reaction mixture and the details of the assay are described under "Methods." Reactants Incorporation prides Complete system.. cm) previously equilibrated with 13uffer A. A linear gradient 0.02 to 0.2 M potassium phosphate, pH 6.5, with a total volume of 2 liters was applied immediately and 20-ml fractions were collected. DNA polymerase activity was eluted in a single peak at 0.1 M potassium phosphate.3 Fractions (about one-third of the activity) were pooled, dialyzed against 0.04 M potassium phosphate (Buffer B), and concentrated by adsorption to a column of phosphocellulose (1 x 4 cm) previously equilibrated with Buffer B. DNA polymerase III was eluted with 0.2 M potassium phosphate, pH 6.5. Fractions of 0.5 ml were collected and fractions representing about half of the activity applied to the column were pooled (1.5 ml). This fraction, Fraction V, after dialysis for 10 hours against 100 ml of Buffer B, was used for all experiments to be described.
DNA polymerase III purified by this procedure yielded preparations purified greater than lO,OOO-fold with respect to the SlOO fraction.
The yield of enzyme activity varied from 3 to 10%. Preparations stored for three months at 0" lost 50% of their original act.ivity.

General Properties of DNA Polymerase III
Fraction V is devoid of detectable deoxyribonuclease activity directed against native DNA capable of generating acid-soluble DNA.
Incubation of native, E. coli [3H]DNA with an amount of Fraction V sufficient to catalyze the incorporation of 60 nmoles of nucleotide resulted in the appearance of less than 1 pmole of acid-soluble material. Incubation of 5' terminally labeled (pdT)lo with enzyme (1.2 units) did not result in the appearance of inorganic phosphate.
Thus Fraction V does not appear to contain phosphatase activity.
The general requirements for DNA polymerase III-catalyzed synthesis are summarized in Table II. pH Dependence of Reaction Rate---Maximal activity was observed at pH 7.0 in morpholinopropane sulfonic acid buffer. The rate of polymerization was 85% of the optimal value at pH 6.5 in morpholinopropane sulfonic acid buffer or at pH 7.5 in Tris buffer.
Divalent Metal Requirement-DNA polymerase III requires Mg+f for optimal activity.
In the absence of Mg++ and in the presence of EDT.4, (3 mM), no detectable activity was observed. Under the conditions of the standard assay, 6 to 16 mM MgClz was optimal.

Suljhydryl
Requirement-DNA polymerase III requires 2mercaptoethanol or dithiothreitol for maximal activity. The optimum concentration of either compound was 50 mM; in their absence and with N-ethylmaleimide (10 mu) present, all activity was abolished (Table II).
Requirements for Deoxynucleoside Triphosphates-All four deoxynucleoside 5'-triphosphates are required for the enzyme to utilize calf thymus DNA as template.
A reciprocal plot of substrate concentration versus velocity indicates an apparent K, for deoxynucleoside triphosphates of 2 X 10m5 M (Fig. 1). A concentration of 1.3 x 10e4 M of triphosphates was sufficient to support the maximal rate of synthesis by DNA polymerase III.
Addition of ATP to the reaction mixture had no effect. Effect of Ionic Strength on Rate of Reaction-DNA polymerase III activity is sensitive to the ionic strength of the reaction mixture.
Activity was maximal at low ionic strength in a reaction mixture including the minimum amount of buffer required for pH maintenance, stranded '(gap" in native DNA. "Gapped" T-7 DNA effectively supports synthesis by DNA polymerase III (Table III). Gaps initiated by DNase I yield 5'-phosphoryl termini. Gaps with 5'-hydroxyl termini result from sequential digestion with micrococcal nuclease and exonuclease III.
Both templates are equally active, indicating that neither a 5'-phosphoryl nor a 5'-hydroxyl is mandatory.
Thus DNA polymerase III is capable of "repair" type synthesis using templates with single-stranded regions and 3'-hydroxyl-terminated primer strands. Fig. 3 gives a schematic representation of these results and compares the template-primer requirements of DNA polymerase III to the known properties of DNA polymerases I (20) and II (6, 10).
The sequential digestion of DNA by DNase I and exonuclease III yields the most active template-primer for synthesis studied to date.4 DNA polymerase III performs the same "repair" reaction on DNA extensively degraded by DNase I, although this template is less than 40%) active relative to the best "gapped" templates studied.

Kinetics oj Synthesis
Under the conditions of the stranded assay, DNA polymerase III catalyzes a "repair" type reaction. Both the initial rate and the final extent of the reaction are determined by the template.
When an excess of "gapped" DNA was used, the initial rate of incorporation of TMP residues was directly proportional to enzyme concentration (0.006 to 0.24 units).
However, under standard assay conditions (30"), synthesis was linear with time for only 10 mm and continued at a decreasing rate thereafter. At 37" all synthesis ceased after 10 min of incubation.
Addition of more DNA is without effect, while addition of more enzyme resulted in the resumption of synthesis.
These results indicate that curtailment of synthesis can be independent of DNA concentration, and is not due to the saturation of available substrate. These observations suggest that the enzyme is labile in the reaction mixture.
Sensitivity of DNA polymerase III activity to incubation was not overcome by the addition of bovine serum albumin (1 mg per ml) or by the addition of a sulfhydryl-reducing agent. The proposed lability of DNA polymerase III is further substantiated by the observation that at 23", in the presence of 20% glycerol and excess substrate, incorporation was linear for at least 235 hours (Fig. 4).
As shown in the inset to Fig. 4, the initial rate of the reaction increases with temperature between 15" and 37". A 1.5.fold increase in the rate of incorporation at 37" relative to 30" was observed.

Primer Requirement
As shown in Table III, DNA polymerase III is most effective in repairing single-stranded regions generated by exonuclease III digestion.
Can DNA polymerase III initiate synthesis de 720~0, in the absence of a primer strand?
The failure of DNA polymerase III to utilize the polymers as template suggests that DNA polymerase III is incapable of de no~o chain initiation with any of the four common deoxynucleoside triphosphates.
That single-stranded circles are inactive as well indicates that the failure of DNA polymerase III to utilize single-stranded polymers as template is not a consequence of exonucleolytic degradation.   (pdT)lo labeled at the 5' terminus with 32P was used to prime poly(dA)-directed synthesis of (pdT),.
The details of the reaction are given in the legend to Table  V. After 80 min of incubation the reaction mixture (0.6 ml) was heated to 100" for 5 min and subjected to filtration at 65" on a column of Sephadex G-50 (1 X 110 cm). A portion (0.1 ml) was analyzed for 3H (+-) and for 32P (---

Mechanism of Chain Elongation; Direction of Synthesis
The role of the complementary oligonucleotide primer in polymer-directed synthesis was assessed through the use of selectively labeled oligonucleotides. The oligonucleotide, ["*PI-pdT(pdT)g, was prepared by sequential treatmen' of (pdT)Io with alkaline phosphatase, and, using [T-~~P]ATP, with polynucleotide kinase.
Acidinsoluble radioactivity was determined after the addition of 40 nmoles of calf thymus DNA and 5% trichloroacetic acid. Radioactivity adsorbable to Norit was removed by the addition of 2 drops of 25y0 Norit, 20 ~1 of 5% trichloroacetic acid, and 10 ~1 of saturated sodium pyrophosphate.
After 5 min at 0", the Norit was collected on a Millipore filter, and washed with 1 ml, 1% of trichloroacetic acid. Norit nonadsorbable material was collected and counted in Bray's solution. incubation 50 7; of the added primer was rendered acid insoluble; each acid-insoluble oligonucleotide was extended by approximately 12 residues. The product of this reaction was subjected to filtration at 65" on a column of Sephadex G-50. The oligonucleotide primer, which prior to incubation with DNA polymerase III was eluted at 25yc of the bed volume, was now shown to be excluded from the gel as was a fraction of the 3H (Fig. 5).
Treatment of the product with alkaline phosphatase at 65" rendered the 32P both acid soluble and Norit nonadsorbable.
The 3H remained insoluble in acid and adsorbable by Norit.
Thus, 32P initially present on the 5' end of the (P~T)~~ primer remained susceptible to phosphomonoesterase after the reaction, indicating that addition of TMP residues was to the 3' end of the primer.
Thus, the product of synthesis was covalently attached to the primer; the 5' end of the primer remained intact and the primer was extended by the addition of TMP residues to the 3' end.
The ability to quantitate the amount of primer active in chain elongation permits the examination of the mechanism of chain elongation by DNA polymerase III. In the experiment described above, 80 pmoles of oligonucleotide primer molecules were used in chain extension.
Yet the 0.8 unit of enzyme present should not exceed 1 pmole of enzyme molecules.
These results indicate that one enzyme molecule is capable of extending more than one DNA primer molecule under these conditions. We cannot, however, conclude from these results with what frequency the enzyme remains attached to the chain which it has just extended. DISCUSSION We have previously shown that DNA polymerase III performs an essential function in DNA replication (14). E. coli strains with thermosensitive mutations at the dnaE locus are temperature sensitive for DNA synthesis and for viability.
DNA polymerase III is mutationally altered in these strains; both polymerases I and II are normal.
We have concluded from these results that both DNA polymerase I and II are unrelated to DNA polymerase III and that DNA polymerase III is essential for viability, independent of the presence of DNA polymerases I and II. The present communication represents an attempt to assess the distinctive functions of DNA polgmerase III through an examination of both its general properties and its catalyt'ic capabilities.
Perhaps the most distinctive feature of DNA polymerase III is the rate at which it can synthesize DNA.
Contrary to previous reports that extracts from the Pol Ai-mutant retain less than 1% the DNA-polymerizing capacity of wild type E. coli (I, 5-lo), we now find that cell-free extracts of the Pol Almutant, assayed under conditions optimal for DNA polymerase III activity, possess DNA-synthesizing activity approximately equal to that amount measured in Pal+ cells. Purification of this activity permits the evaluation of the mechanism of its catalysis.
Our most highly purified preparations of DNA polymerase III have a specific activity in excess of 60,000 units per mg of protein.
From polyacrylamide gel analysis, we judge this preparation to be approximately 33% pure. Assuming that the purity is 33% and that the molecular weight is approximately 140,000,6 an approximate estimate can be made both of the rate of nucleotide incorporation per molecule of enzyme and of the number of DNA polymerase III molecules per E. coli cell. There are approximately 10 molecules of DNA polymerase III per bacterial cell; the rate of nucleotide incorporation at 30" is in excess of 15,000 nucleotides per min, per molecule of DNA polymerase III.
(This measurement is limited to templates studied to date.) Relative to DNA polymerases I and II, of which there are approximately 400 (21) and 100 (6) molecules, respectively, per bacterial cell, the rate of synthesis by DNA polymerase III is greater than the rate of synthesis catalyzed by DNA polymerases I and II by a factor of 15 and 300, respectively. The total activiby of DNA polymerase III is sufficient to account for the in vivo rate of replication.
Several groups of investigators have reported that a nonionic detergent was necessary to obtain DNA polymerase II from extracts of E. coli (8, 9). We have not observed such a require ment.
We have previously reported that both DNA polymerases II and III can be obtained from French pressure cell lysates, and now report that this method yields DNA polymerase III activity in amounts sufficient to account for the in viva rate of replication. Recently, Wickner et al. (10) have isolated DNA polymerase II from cells lysed by alumina grinding, confirming the observation that detergent treatment is not required for the isolation of DNA polymerase II. Thus, DNA polymerases II and III, like DNA polymerase I (21), can be obtained in soluble form without sonication or detergent treatment, suggesting that procedures disruptive to membranes are not required to yield soluble enzyme.
DNA polymerase III can be distinguished from DNA polymerases I and II by virtue of a low pH optimum, a requirement for high concentrations of sulfhydryl reagent, sensitivity to salt, and stimulation by ethanol.
The apparent K, for deoxynucleoside triphosphates is greater than the levels required for the saturation of DNA polymerases I and II.
DNA polymerase III closely resembles DNA polymerases I and II with regard to its catalytic properties.
DNA polymerase III is not capable of de novo chain initiation using either singlestranded synthetic polynucleotides or single-stranded circular DNA as templates.
Both can be rendered active if a primer (ribo or deoxyribo) with a free 3'-hydrosyl is provided. Synthesis can proceed in the 5' to 3' direction by covalent linkage of the product to the 3'.hydroxyl end of the primer. Although elongation of primers in the 5' direction has not been observed, we cannot conclude that DNA polymerase III is incapable of carrying out synthesis in the 3' to 5' direction.
Such an event may require a triphosphate moiety at the 5' terminus (22), and this possibility is currently being explored.
As has been observed with DNA polymerases I and II, preparations of DNA polymerase III possess an associated nuclease which catalyzes the degradation of single-stranded DNA exonucleolytically from the 3' end. In studies which are not reported here, the nuclease activity was found to be inactive with doublestranded DNA; its catalytic requirements resembled those described for the polymerizing capacity of DNA polymerase III (i.e. sensitivity to sulfhydryl reagents and to salt). The rate of nucleotide removal is in excess of 5000 nucleotides per min per enzyme molecule.
The template requirements of DNA polymerase III suggest a strong similarity to DNA polymerase II. Neither enzyme, in contrast to DNA polymerase I, can utilize single-stranded natural DNA templates or DNA with single strand scissions. None of the polymerases can achieve the replication of native, duplex DNA.
All perform a "repair" function with greatest efficiency. Thus, the catalytic properties of DNA polymerase III do not establish its identity as a polymerase distinctly different from either DNA polymerase I or II, nor do they suggest the role of DNA polymerase III in DNA metabolism.
The only properties which clearly differentiate DNA polymerase III from DNA polymerases I and II are its requirement for low ionic strength and ethanol as well as a rapid rate of nucleotide incorporation. Although DNA polymerase III has not been observed to initiate strands de novo, to replicate double-stranded DNA, or to carry out synthesis in the 5' direction, it may be naive to expect a single enzyme to do so. The distinctive features of this DNA polymerase may lie in its ability to cooperate with other proteins