Role of the C-terminal Residue of the DNA Polymerase of Bacteriophage T7

The crystal structure of the DNA polymerase encoded by gene 5 of bacteriophage T7, in a complex with its processivity factor, Escherichia coli thioredoxin, a prim-er-template, and an incoming deoxynucleoside triphosphate reveals a putative hydrogen bond between the C-terminal residue, histidine 704 of gene 5 protein, and an oxygen atom on the penultimate phosphate diester of the primer strand. Elimination of this electrostatic interaction by replacing His 704 with alanine renders the phage nonviable, and no DNA synthesis is observed in vivo . Polymerase activity of the genetically altered enzyme on primed M13 DNA is only 12% of the wild-type enzyme, and its processivity is drastically reduced. Kinetic parameters for binding a primer-template ( K D app ), nucleotide binding ( K m ), and k off for dissociation of the altered polymerase from a primer-template are not significantly different from that of wild-type T7 DNA polymerase. However, the decrease in polymerase activity is concomitant with increased hydrolytic activity, judg-ing from the turnover of nucleoside triphosphate into the corresponding nucleoside monophosphate (percent-age of turnover, 65%) during DNA synthesis. Biochemical data along with structural observations imply that the terminal amino acid residue of T7 DNA polymerase plays a critical role in partitioning DNA between the polymerase and exonuclease sites. Gene 5 protein encoded by bacteriophage T7 is a replicative DNA polymerase with an associated 3 (cid:1) and (0.06–6 n M ) DNA polymerase. After incubation at 37 °C for 10 min, the reaction was quenched by the addition of 30 (cid:1) l of BSA (10 mg/ml) and 30 (cid:1) l of trichloroacetic acid (100% w/v). After incubation at 0 °C for 15 min, precipitated DNA was collected by centrifugation at 12,000 (cid:7) g for 30 min. The acid-soluble radioactivity was measured by scintillation counting in Ultra fluor (Packard). One unit of exonuclease activity catalyzes the release of 1 pmol of total nucleotides into an acid-soluble form in 1 min. Processivity Assays— The DNA used for processivity assays was a 5 (cid:1) 32 P-labeled 24-nt primer annealed M13 mGP1-2 ssDNA described The reaction mixture (22.5 (cid:1) l) contained 40 m M Tris-Cl, pH 7.5, 10 m M MgCl 2 , 5 m M DTT, 50 m M NaCl, 500 (cid:1) M each of dTTP, dATP, dGTP, and dCTP, and 8 n M primer-template. This mixture was preincubated at 37 °C for 3 min, and the reaction was started by the addition of 2.5 (cid:1) l of T7 DNA polymerase. Final concentration of the polymerase was 0.3 n M . After the indicated times, the reaction was stopped by the addition of 5 (cid:1) l of stop solution that contained 40% sucrose, 20 m M EDTA, and 0.05% bromphenol blue. The reaction mixtures were subjected to elec- trophoresis on a 0.6% agarose gel in a buffer containing 100 m M Tris borate, pH 8.3, 1 m M EDTA, and 0.06 (cid:1) g/ml ethidium bromide. The gels were dried and autoradiographed. At subsaturating concentrations of ethidium bromide, nicked M13 dsDNA has lower electrophoretic mobil-

Gene 5 protein encoded by bacteriophage T7 is a replicative DNA polymerase with an associated 3Ј-5Ј exonuclease activity (1). This 80-kDa enzyme, by itself, is distributive for DNA synthesis and can only incorporate less than 15 nucleotides before dissociating from a primer terminus (2). However, in a 1:1 complex with the host Escherichia coli protein, thioredoxin (K D ϭ 5 nM), gene 5 protein processively catalyzes the addition of thousands of nucleotides at rates approaching 300 nucleotides/s (2)(3)(4)(5)(6). The gene 5 protein-thioredoxin complex will be referred to as T7 DNA polymerase in this manuscript.
A recent crystal structure of T7 DNA polymerase in complex with a primer-template and an incoming dideoxynucleoside triphosphate solved at 2.2 Å resolution (7) reveals a bipartite structure with distinct C-terminal polymerase and N-terminal exonuclease domains (Fig. 1). Most polymerases of the polym-erase I family have a similar bipartite architecture (8 -12). A "right hand" with distinct fingers, palm, and thumb subdomains forms a distinct DNA-binding groove that leads to the polymerase active site (13). The fingers, palm, and thumb grip the primer-template by a number of direct and water-mediated contacts mainly to the phosphodiester backbone of DNA such that the 3Ј-end of the primer strand is positioned next to the nucleotide-binding site. The 3Ј-5Ј exonuclease activity on ssDNA 1 and dsDNA serves to excise mismatches during processive DNA synthesis (14,15). The polymerase and the exonuclease active sites are ϳ35 Å apart.
Although the error frequency during base selection by the polymerase activity occurs at the low rate of one misincorporation per 10 6 turnovers, the 3Ј-5Ј proofreading activity further reduces the error frequency by 10 -200-fold depending on the method of measurement (16,17). It is believed that the 3Ј-5Ј exonuclease activity maximizes its contribution to replication fidelity by minimal hydrolysis of correctly base paired DNA (18). The pre-steady-state kinetic pathways for the polymerization and the exonuclease reactions have been described in detail for T7 DNA polymerase (17)(18)(19) and the Klenow fragment of E. coli DNA polymerase I (20 -25). An equilibrium exists between DNA binding to the polymerase or exonuclease active site, with binding to the polymerase site being thermodynamically more favored (18). Any process that slows the rate of the elongation reaction, for example a mismatch, allows time for the transfer of DNA to the exonuclease active site (17,25). The data also suggest that DNA is transferred intramolecularly in both directions between the polymerase and exonuclease sites without dissociating from the enzyme. However, the mechanisms that underlie the occupancy of DNA in the polymerase versus the exonuclease site are not well understood.
Residues that contact DNA may be expected to govern the partitioning of DNA between the two active sites. The crystal structure of T7 DNA polymerase reveals a putative hydrogen bond between the C-terminal residue His 704 of gene 5 protein and an oxygen atom on the penultimate phosphate diester of the primer strand (Fig. 1). This C-terminal residue is conserved in T7 DNA polymerase (His 704 ) and in E. coli DNA polymerase I (His 928 ). In the DNA polymerases from Bacillus stearothermophilus (26,27) and Thermus aquaticus (28 -30), the homologous residues are Lys 876 and Lys 832 , respectively. To probe the role of this interaction with the 3Ј-OH primer terminus, His 704 of T7 DNA polymerase was substituted with alanine. The loss of this interaction could result either in reduced affinity of the polymerase for the primer-template or in reduced affinity for just the primer strand. In this report, we show that the Cterminal histidine of T7 DNA polymerase plays a critical role in orienting the primer terminus in position for catalysis and possibly in the controlled shuttling of DNA between the polymerase and exonuclease sites.

Materials
DNA-M13 mGP1-2 is a 9950-nt derivative of vector M13 mp8 containing an insert of phage T7 DNA (31). The 24-nt M13 sequencing primer (Ϫ47) (5Ј-CGCCAGGGTTTTCCCAGTCACGAC-3Ј) and oligonucleotides for in vitro mutagenesis were from Oligos Etc. Activated calf thymus DNA (Type XV) was obtained from Sigma. Poly(dA) 280 was obtained from Amersham Pharmacia Biotech, and oligo(dT) 22 was obtained from Integrated DNA Technologies. Poly(dA) 280 and oligo(dT) 22 were dissolved in a 1:1 molar ratio (20 M) in 40 mM Tris-Cl, pH 7.5, and 50 mM NaCl and annealed by heating to 95°C for 5 min, followed by slowly cooling to room temperature. The 24-nt M13 sequencing primer and M13 mGP1-2 DNA were mixed in a 1:1 molar ratio (100 nM) and annealed using the same protocol. Oligonucleotide concentrations were determined spectrophotometrically. DNA concentrations are expressed in terms of primer 3Ј-ends.
E. coli Strains, Bacteriophage, and Plasmids-E. coli strains C600 and HMS174 are from the laboratory collection. E. coli HMS174 (DE3)/ pLysS cells are from Novagen. Wild-type bacteriophage T7 and mutant T7⌬5 (gene 5 deletion) phage are from the laboratory collection. Plasmid pGP5-3 contains T7 gene 5 under the control of the T7 RNA polymerase promotor ⌽10 (2). Plasmid pGP5-5,7A contains T7 gene 5 in which two exonuclease active site residues of gene 5 (Asp 5 and Glu 7 ) have been substituted with alanine. Plasmid pT7-7 is the parent vector of pGP5-3 missing the GP5 insert. Growth and manipulation of bacteriophage T7 and E. coli were performed as described (32,33).
Mutagenesis of T7 Gene 5-Plasmid pGP5-H704A was constructed using standard polymerase chain reaction and cloning techniques. Two oligonucleotide primers, (5Ј-CGGGATCCTCAGCCGCAAATCGCCCA-ATTA-3Ј) and (5Ј-CCGTTGGTGCCGGCAAAGAGCGCGG-3Ј) were used to amplify 340 base pairs of the T7 DNA sequence containing His 704 . The codon in bold type corresponds to the amino acid alteration. Polymerase chain reaction-generated fragments were digested with BamHI and PshAI and then cloned into the PshAI and BglII sites of plasmid pGP5-3. Plasmid pGP5-H704A contains the genetically altered gene 5 that encodes gp5-H704A under control of the T7 ⌽10 promotor. Plasmid pGP5-H704A was digested with restriction enzymes HindIII and StyI, and the resulting fragment was ligated into the HindIII and StyI site of pGP5-5,7A giving rise to pGP5-H704A-5,7 A. The identity of the clones was confirmed by DNA sequencing.
Enzymes-Gene 5 proteins were overproduced from E. coli HMS174 (DE3)/pLysS cells carrying plasmids. The 1:1 complex of polymerase and thioredoxin was purified to apparent homogeneity as described (2). Protein concentrations were determined by the method of Bradford (34) and were confirmed by amino acid analysis. Restriction enzymes were from New England Biolabs.
Nucleotides-Unlabeled nucleotides (high pressure liquid chromatography grade) were obtained from Amersham Pharmacia Biotech. [
[ 3 H] Thymidine Incorporation Assays-Thymidine incorporation assays were carried out at 30°C (35). E. coli C600 cells harboring plasmid pGP5-3 or pGP5-H704A were grown to a density of 3 ϫ 10 8 cells/ml in Davis medium (0.7% potassium diphosphate, 2% potassium monophosphate, 0.05% sodium citrate, 0.01% magnesium sulfate, 0.1% ammonium sulfate) supplemented with glucose, thiamine, casamino acids, and ampicillin (80 g/ml). The cells were infected with either wild-type T7 phage or T7 phage containing a deletion of gene 5 (⌬5 phage) at a multiplicity of infection of ϳ5 (35). At indicated time intervals, 200 l of the samples were removed, and [ 3 H]thymidine was added to a final concentration of 50 Ci/ml. Radioactive labeling was terminated after 90 s by the addition of 3 ml of ice-cold 0.3 N trichloroacetic acid. Acid-insoluble radioactivity was collected via filtration on glass microfiber filters and washed three times with 1 M HCl (3 ml) and twice with ethanol (3 ml). The acid insoluble radioactivity was measured using a scintillation counter.
To measure a single round of DNA synthesis, (dA) 280 ⅐(dT) 22 was used as a primer-template, and the reactions were carried out at 22°C as described (6). The DNA polymerase was preincubated with the primertemplate in the absence of Mg 2ϩ and dTTP for 5 min. The reactions were initiated by the addition of Mg 2ϩ , dTTP, and challenger DNA to trap any free polymerase. The aliquots were withdrawn at the indicated times and quenched with a final concentration of 100 mM EDTA. The control reactions to measure background incorporation by polymerase not trapped by challenger DNA were carried out by adding challenger DNA to the preincubation mix. This background reaction was subtracted wherever applicable. The preincubation mixture (30 l) contained 270 nM poly(dA) 280 ⅐oligo(dT) 22  A ternary complex of T7 gene 5 protein with its processivity factor, thioredoxin, a primed DNA template and an incoming 2Ј,3Ј-dideoxynucleotide (ddGTP) determined at 2.2 Å resolution (7). Inset, the C-terminal residue, His 704 in T7 DNA polymerase contacts the phosphate backbone at the primer 3Ј-end (2.7 Å). 25 nM DNA polymerase, 10 mM MgCl 2 , and 200 g/ml calf thymus DNA. All reaction mixtures also contained 40 mM Tris-Cl, pH 7.5, 5 mM DTT, 50 mM NaCl, and 50 g/ml BSA.
The reaction mixtures for exonuclease assays (100 l) contained 40 mM Tris-Cl, pH 7.5, 10 mM MgCl 2 , 5 mM DTT, 50 mM NaCl, 1 nmol (in terms of total nucleotides) [ 3 H]-labeled M13 mGP1-2 ssDNA or dsDNA, and (0.06 -6 nM) DNA polymerase. After incubation at 37°C for 10 min, the reaction was quenched by the addition of 30 l of BSA (10 mg/ml) and 30 l of trichloroacetic acid (100% w/v). After incubation at 0°C for 15 min, precipitated DNA was collected by centrifugation at 12,000 ϫ g for 30 min. The acid-soluble radioactivity was measured by scintillation counting in Ultra fluor (Packard). One unit of exonuclease activity catalyzes the release of 1 pmol of total nucleotides into an acid-soluble form in 1 min.
Processivity Assays-The DNA used for processivity assays was a 5Ј 32 P-labeled 24-nt primer annealed to M13 mGP1-2 ssDNA as described (2). The reaction mixture (22.5 l) contained 40 mM Tris-Cl, pH 7.5, 10 mM MgCl 2 , 5 mM DTT, 50 mM NaCl, 500 M each of dTTP, dATP, dGTP, and dCTP, and 8 nM primer-template. This mixture was preincubated at 37°C for 3 min, and the reaction was started by the addition of 2.5 l of T7 DNA polymerase. Final concentration of the polymerase was 0.3 nM. After the indicated times, the reaction was stopped by the addition of 5 l of stop solution that contained 40% sucrose, 20 mM EDTA, and 0.05% bromphenol blue. The reaction mixtures were subjected to electrophoresis on a 0.6% agarose gel in a buffer containing 100 mM Tris borate, pH 8.3, 1 mM EDTA, and 0.06 g/ml ethidium bromide. The gels were dried and autoradiographed. At subsaturating concentrations of ethidium bromide, nicked M13 dsDNA has lower electrophoretic mobility than M13 ssDNA annealed to a 24-nt oligonucleotide.
Nucleotide Turnover Assays-The reaction mixture for turnover assays was essentially the same as for polymerase assays with primed M13 mGP1-2 DNA. The concentrations of wild-type T7 DNA polymerase and gp5-H704A were 2 and 4 nM, respectively. The incorporation of [␣-32 P]dTMP into the primer strand was measured as described for polymerase assays. The amount of [␣-32 P]dTMP obtained upon exonucleolytic hydrolysis by the enzyme was quantified by TLC. An aliquot of the reaction mixture was applied to a polyethyleneimine plate (JT Baker) prewashed with distilled water. The thin layer plate was developed with 0.6 M LiCl in 1 M formic acid. The amount of [␣-32 P]dTMP formed was measured using phosphorus imaging analysis with a Fuji BAS 1000 bio-imaging analyzer.
Dissociation Kinetics-The half-life of a preformed polymerase-DNA complex was determined by preincubating the polymerase with poly(dA) 280 ⅐oligo(dT) 22 and adding challenger DNA at time 0. After varying periods of time (5-60 s), the polymerase reaction was started by the addition of Mg 2ϩ and dTTP and stopped after 30 s by the addition of EDTA to a final concentration of 100 mM. Final concentrations of all of the reagents were identical to those used in the polymerase assay with poly(dA) 280 ⅐oligo(dT) 22 .
Determination of K m and K D -Assays to determine K m were carried out at 22°C with poly(dA) 280 ⅐oligo(dT) 22 . A series of 10 concentrations of dTTP were used to bracket the K m value. All reaction mixtures contained 40 mM Tris-Cl, pH 7.5, 10 mM MgCl 2 , 5 mM DTT, 50 mM NaCl, 160 nM poly(dA) 280 ⅐oligo(dT) 22

RESULTS
Gp5-H704A Cannot Support the Growth of T7 Phage-To ascertain the role of the C-terminal histidine in T7 DNA polymerase, a gene 5 mutant was constructed in which histidine 704 was replaced with alanine (gp5-H704A). The effect of the genetically altered gene 5 protein on the growth of T7 phage was tested (Table I). When gp5-H704A is produced from a plasmid, it is unable to support the growth of T7 phage in which the wild-type gene 5 has been deleted (T7⌬5). T7⌬5 phage are dependent on the expression of plasmid encoded gene 5 protein for viability. Furthermore, gp5-H704A inhibits the growth of wild-type T7 phage that is expressing the wildtype gene 5. In contrast, production of wild-type gene 5 protein from a plasmid supports the growth of T7⌬5 phage and has no effect on the growth of on wild-type T7 phage. Thus, the histidine to alanine substitution is dominant lethal for T7 phage growth.
DNA Synthesis in Vivo-The rates of in vivo DNA synthesis were measured by monitoring the incorporation of [ 3 H]thymidine into DNA in phage-infected cells (Fig. 2). E. coli C600 cells harboring a plasmid (pGP5-3 or pGP5-H704A) were infected with either T7⌬5 phage ( Fig. 2A) or wild-type T7 phage (Fig.  2B). T7 DNA synthesis in infected cells starts ϳ10 min after infection, presumably after the shut down of host DNA synthesis (33). Fig. 2A shows a plot of the incorporation of [ 3 H]thymidine as a function of time upon infection with T7⌬5 phage. When wild-type gene 5 protein is produced from the plasmid, DNA synthesis starts to increase 10 min after infection and continues to increase up to 30 min after infection. In contrast, when gp5-H704A is produced from the plasmid, DNA synthesis is strongly inhibited 10 min after infection with T7⌬5 phage.
Even more striking is the kinetics of DNA synthesis upon infection with wild-type T7 phage (Fig. 2B). With wild-type gene 5 protein, T7 DNA synthesis starts 10 min after infection, reaching a maximum at 40 min after infection. In contrast, with gp5-H704A, DNA synthesis starts to decrease immediately upon infection. These data indicate that the altered protein, gp5-H704A, cannot restore DNA synthesis in T7⌬5 phageinfected cells, and furthermore, gp5-H704A inhibits DNA synthesis in wild-type T7 phage-infected cells.
DNA Synthesis in Vitro-To determine how the mutation in gene 5 leads to decreased DNA synthesis in vivo, gp5-H704A was overproduced, purified, and characterized bio- a Plasmid pGP5-3 contains T7 gene 5 under the control of the T7 RNA polymerase promotor 10, pGP5-H704A contains T7 gene 5 with the H704A codon substitution, and pT7-7 is the parent vector of pGP5-3 missing the gene 5 insert.
chemically. Initially, gp5-H704A was compared with wildtype T7 DNA polymerase by measuring the rate of polymerization of nucleotides. Initial rates of DNA synthesis with purified enzymes were measured by following the incorporation of [ 3 H]dTMP into DNA using primed M13 ssDNA (Fig. 3) or poly(dA) 280 ⅐oligo(dT) 22 (Fig. 4) as primertemplates. The reactions were carried out under conditions of excess nucleotides and DNA. The steady-state rate constant for polymerization (k pol ) was determined by dividing the rate of incorporation of nucleotides by the polymerase concentration. The polymerase activity has been expressed in terms of k pol . Gp5-H704A has 8-fold lower polymerase activity (k pol ϭ 30 s Ϫ1 ) than the wild-type polymerase (k pol ϭ 240 s Ϫ1 ) on M13 ssDNA (Fig. 3). With poly(dA) 280 ⅐oligo(dT) 22 , the polymerase activity of gp5-H704A (k pol ϭ 60 s Ϫ1 ) is 3-fold lower than of the wild-type polymerase (k pol ϭ ϳ200 s Ϫ1 ) (Fig. 4A). One possible explanation of the lower rate of DNA synthesis on the long M13 DNA template relative to that observed on the shorter poly(dA) 280 ⅐oligo(dT) 22 template is that gp5-H704A is less processive.
DNA synthesis for a single processive cycle was measured by preincubating the polymerase with poly(dA) 280 ⅐oligo(dT) 22 and initiating the polymerase reaction by the addition of Mg⅐dTTP along with activated calf thymus DNA added to trap any polymerase dissociating from the primer-template (6). The back-ground incorporation from polymerase not trapped by challenger DNA was subtracted (Fig. 4C). Under these conditions, where polymerase activity is measured for a single primertemplate binding, polymerization, and dissociation reaction, gp5-H704A retains 50% of the activity observed with the wildtype polymerase (Fig. 4B). It has been suggested that with short templates, the rate of dissociation and reassociation of T7 DNA polymerase with DNA becomes rate-limiting, and processivity is unimportant (2). Taken together, these results are indicative of a lower processivity of DNA synthesis for gp5-H704A. The processivity of DNA synthesis catalyzed by gp5-H704A has also been measured directly by measuring the length of products formed by gel electrophoresis.
Effect of Ionic Strength on Polymerase Activity-Rates of DNA synthesis were measured as a function of NaCl concentration by following the incorporation of [ 3 H]dTMP into primed M13 ssDNA. 100 mM NaCl is known to reduce the processivity of T7 DNA polymerase (2) because the higher ionic strength may stabilize secondary structures in the template. The rate of DNA synthesis is reduced at the higher ionic strength (Table  II) for both the wild-type and gp5-H704A polymerases. Although at 50 mM NaCl, gp5-H704A retains 12% of the activity of the wild-type polymerase, at 100 mM NaCl concentration, gp5-H704A retains only 3% of the activity of the wild-type polymerase. Since the alanine substitution removes a putative electrostatic interaction with DNA, it is plausible that higher salt concentrations preferentially destabilize the binding of gp5-H704A to DNA.
Processivity Assays-The processivity of DNA synthesis catalyzed by the polymerase was measured by the method of von Hippel and co-workers (36,37). M13 ssDNA annealed to a 5Ј 32 P-labeled 24-nt oligonucleotide was used as the primer-template. To measure the length of products formed from a single polymerase-DNA binding event, the concentration of the primer-template was maintained in ϳ25-fold molar excess over the polymerase. Aliquots of the reaction mixture were removed at different times and were subjected to electrophoresis in an agarose gel (Fig. 5). The processivity of T7 wild-type DNA polymerase is in the order of thousands of nucleotides as observed from the accumulation of the full-length 9950-nt product in less than 5 min (2). Although a rate of incorporation of 240 nt/s is observed in steady-state assays for short periods of incubation (Table II), this value may not be extrapolated for longer incubations on long templates that contain secondary structure. In direct contrast, the length of products synthesized by gp5-H704A is drastically shorter with no full-length products observed even after a reaction time of 60 min. Furthermore, the decrease in amount of unreplicated primer-template with time indicates the cycling of the altered polymerase. The extent of elongation with gp5-H704A is dependent on the enzyme concentration (Fig. 5B). To obtain the longer products that are formed with the wild-type enzyme, 80 times more gp5-H704A as compared with wild-type T7 DNA polymerase is required. This last experiment is not a processivity assay per se because the polymerase concentration is higher than the concentration of the primer-template. Nevertheless, the results establish that the limited extensions are a result of dissociation from the primer-template and, thus, reduced processivity, rather than a lower rate of elongation.
3Ј-5Ј Exonuclease Activity-Gene 5 protein has a 3Ј-5Ј exonuclease activity on both ssDNA and dsDNA. Thioredoxin greatly stimulates the 3Ј-5Ј exonuclease activity on dsDNA (15). The 3Ј-5Ј exonuclease activity of the two enzymes was measured on both ssDNA and dsDNA (Table III) in the absence of added nucleotides. The intrinsic exonuclease activity of the two enzymes is comparable on both ssDNA and dsDNA.
Nucleotide Turnover-Nucleotide turnover is defined as the DNA-dependent conversion of deoxynucleoside triphosphates into their corresponding monophosphates under conditions of DNA synthesis. The polymerization of nucleotides into a primer followed by exonucleolytic degradation by the 3Ј-5Ј exonuclease activity associated with the polymerase results in nucleotide turnover. Thus, it reflects the levels of "editing" (38,39) that take place in a reaction. To determine whether the substitution of His 704 with alanine that removes an interaction with the primer 3Ј-terminus in the polymerase domain results in an increase in exonucleolytic editing, the nucleotide turnover associated with gp5-H704A and wildtype T7 DNA polymerase was measured. The rate of hydrolysis of deoxynucleoside triphosphates to the corresponding monophosphate (turnover) during DNA synthesis was measured using primed M13 ssDNA and [␣-32 P]dTTP. Enzyme concentrations were normalized to yield similar amounts of DNA synthesis, and the rates of DNA synthesis were measured by the incorporation of [␣-32 P]dTMP. To measure the exonucleolytic release of [␣-32 P]dTMP, an aliquot of the reaction mixture was removed at the indicated time and analyzed by TLC (Fig. 6). For similar amounts of DNA synthesis, the amount of dTMP formed with gp5-H704A was 11 times higher than that with wild-type polymerase (Table IV). Because the exonuclease activity on ssDNA of gp5-H704A is comparable with that of the wild-type polymerase, the increased nucleotide turnover observed with gp5-H704A must reflect an increased rate of switching of DNA from the polymerase to exonuclease active site rather than an intrinsically higher exonuclease activity.
Dissociation Kinetics of Preformed Complexes-DNA is held in position by the fingers, palm, and thumb subdomains of T7 DNA polymerase by a number of contacts, both direct and water-mediated, primarily to the phosphodiester backbone (7). To determine whether the removal of a single hydrogen bond to the primer terminus perturbs the overall binding of the polym- FIG. 5. Processivity of wild-type T7 DNA polymerase and gp5-H704A. A 5Ј 32 P-labeled 24-nt primer was annealed to a M13 ssDNA template. The primer was extended by the incorporation of four deoxyribonucleoside monophosphates in a standard polymerase assay at 37°C with either gp5 or gp5-H704A. DNA synthesis conditions are described under "Experimental Procedures." A, the molar ratio of polymerase to primer-template was 1:25. For each reaction, the aliquots were removed at the indicated times and mixed with an equal volume of 0.2% SDS and 25 mM EDTA. The reaction products were separated by electrophoresis in a 0.6% agarose gel containing 0.06 g/ml ethidium bromide and visualized by autoradiography. B, effect of increasing polymerase concentration on the length of products. The reaction time was 1 h. bp, base pairs. erase to the primer-template, the rate of dissociation of gp5-H704A and wild-type polymerase from a primer-template terminus was compared. T7 DNA polymerase was incubated with poly(dA) 280 ⅐oligo(dT) 22 to form a polymerase-DNA complex, and the half-life of the complex was measured (6). The complex was incubated for varying periods of time in the presence of calf thymus DNA to trap any polymerase that dissociates. The polymerase reaction was then initiated by the addition of Mg⅐dTTP. Fig. 7 is a plot of the amount of dTMP incorporated as a function of the time interval between the addition of challenger DNA and Mg⅐dTTP. The line was calculated for a first order decay of the polymerase-DNA complex. While the complex of wild-type T7 polymerase with DNA decays with a half-life of 28 s, the complex of gp5-H704A and DNA decays slightly more rapidly with a half-life of 16 s. The half-life for the complex of wild-type enzyme and DNA is in agreement with previously reported values (6). Dissociation Constants for Polymerase-DNA and Polymerase-Nucleotide Complexes-To investigate whether the increased nucleotide turnover associated with gp5-H704A is due to a change in the intrinsic binding of nucleotides or DNA, the dissociation constants for polymerase-nucleotide and polymerase-DNA complexes were determined. The apparent dissociation constant for dTTP binding was determined by measuring the rate of incorporation of dTMP into poly(dA) 280 ⅐oligo(dT) 22 as a function of dTTP concentration (5-800 M). The data were fit by nonlinear regression and gave a K m of 90 M with a maximum rate of incorporation of 65 s Ϫ1 at 22°C for gp5-H704A. The wild-type polymerase has a K m of 40 M with maximum rate of incorporation of 170 s Ϫ1 . The approximately two times higher value of K m for gp5-H704A suggests a slightly weaker binding of nucleotides in the polymerase active site.
Dissociation constants for DNA binding were determined for a single processive cycle of DNA synthesis. The enzyme was preincubated with varying concentrations of poly(dA) 280 ⅐oligo(dT) 22 , and the reaction was initiated by the addition of Mg⅐dTTP and challenger DNA. Challenger DNA was used to trap any dissociating polymerase. The apparent dissociation constants of gp5-H704A (K D ϭ 40 nM) and wild-type polymerase (K D ϭ 30 nM) in complex with poly(dA) 280 ⅐oligo(dT) 22 are similar. The kinetic and dissociation constants have been tabulated in Table V. DISCUSSION Consequent to the addition of a base to a growing DNA chain in the polymerase active site, there are three competitive sinks for the polymerase: (i) further polymerization with catalysis of phosphodiester bond formation, (ii) dissociation from the primer terminus, or (iii) exonucleolytic cleavage of the phosphodiester bond. An equilibrium exists between DNA binding to the polymerase or exonuclease active site with binding to the polymerase site being thermodynamically more favored (18). It is the balance between these two opposing reactions that governs high fidelity of replication with minimal hydrolysis of nucleotides. Although the polymerization and exonuclease re-  a Specific activity has been expressed as the percentage of the activity observed for wild-type T7 DNA polymerase. One unit of exonuclease activity catalyzes the release of 1 pmol of total nucleotides into an acid soluble form in 1 min. 100% double-stranded activity corresponds to 9100 units/mg, and 100% single-stranded activity corresponds to 6600 units/mg. actions have been the subject of detailed kinetic study, the mechanisms that underlie the occupancy of DNA in the polymerase versus exonuclease active site are not well understood. In this report, we describe a mutant of T7 DNA polymerase, T7 gp5-H704A, that has an increased turnover of nucleotides during DNA polymerization. We hypothesize that this increase represents an increase in shuttling of the primer strand to the exonuclease active site. T7 DNA polymerase makes a number of contacts with base pairs near the 3Ј terminus of the primer (7). Substitution of a number of analogous residues in E. coli polymerase I results in altered affinity of the enzyme for DNA (40,41). These residues have been implicated in scanning for errors during polymerizaton. It has been postulated that such residues could be involved in the transfer of mismatched nucleotides to the 3Ј-5Ј exonucleolytic active site in T7 DNA polymerase (11).
In a ternary crystal structure of T7 DNA polymerase in complex with a primer-template and the incoming dideoxynucleoside triphosphate (7), the C-terminal amino acid residue of T7 DNA polymerase, His 704 , makes a putative hydrogen bond with the penultimate phosphate diester of the 3Ј-OH terminus of the primer strand. From the structure, this residue may orient the 3Ј-OH terminus of the primer strand in position for the next phosphoryl transfer reaction. Removal of this contact could either promote the local denaturation of the 3Ј-end that would result in the movement of the primer terminus to the exonuclease site or perturb the overall binding to DNA. To probe the contribution of this single electrostatic interaction in DNA binding and catalysis, histidine was substituted with an alanine.
The single amino acid substitution in T7 DNA polymerase results in a polymerase that not only cannot support growth of T7 phage but that is also inhibitory to the T7 wild-type allele. DNA synthesis cannot be detected in vivo upon infection by T7 phage of cells harboring a plasmid encoding for the genetically altered protein. To understand the biochemical basis for this phenotype, the genetically altered protein, gp5-H704A, was overproduced from E. coli overexpressing the gene and purified to apparent homogeneity. Biochemical experiments with purified protein show that gp5-H704A retains only 12% of the polymerase activity of the wild-type enzyme using primed M13 ssDNA as a substrate. Furthermore, although wild-type T7 DNA polymerase is extremely processive for DNA synthesis and can catalyze the addition of hundreds of nucleotides in a single encounter with DNA, the processivity for nucleotide polymerization of gp5-H704A is severely reduced. Thus, removal of this canonical electrostatic interaction with the phosphate backbone of DNA lowers the efficiency of translocation of T7 DNA polymerase on DNA. Processivity measurements are dependent on a number of parameters that affect both the polymerase and the template. Consequently, we do not know the precise molecular mechanism for the decrease in processiv-ity. Certainly one parameter that decreases processivity may be a change in the exonuclease:polymerase activity ratio. The higher the ratio, with stalling of the polymerase at secondary structures, the shorter the products.
To examine the contribution of His 704 in binding DNA, the rate of dissociation of preformed polymerase-DNA complexes and the apparent binding affinity of the polymerase for a primer-template were determined. The off rate of gp5-H704A (k off ϭ 0.04 s Ϫ1 ) from poly(dA) 280 ⅐oligo(dT) 22 is not significantly different from that of the wild-type polymerase (k off ϭ 0.02 s Ϫ1 ). The K D app of a gp5-H704A-polyd(A) 280 ⅐oligod(T) 22 complex is also not significantly different relative to the wild-type polymerase. These results are not surprising given that the primer-template is held in position by the polymerase via a number of direct and water mediated interactions with the phosphodiester backbone. The substitution of histidine with alanine removes only a single electrostatic contact with DNA. Hence, a large perturbation in the overall binding constant of DNA to enzyme is not expected.
The position of His 704 from the crystal structure does not suggest a functional role for this residue in catalysis via interaction with the incoming nucleotide or the metal ions in the polymerase active site. However, His 704 contacts Glu 655 , one of the three highly conserved carboxylates in the polymerase I family of polymerases. Glu 655 is located near the metal-ligating residues Asp 475 and Asp 654 but does not contact the metal ions itself. Substitution of the analogous residue in Klenow (Glu 883 ) with alanine results in a modest increase in the K m for dNTP binding (40). To investigate the role of the interaction of Glu 655 with His 704 in nucleotide binding and catalysis, K m (dNTP) was measured using steady-state kinetics. The K m for dNTP binding is 2-fold higher than that of the wild-type enzyme. A more drastic change in K m (dNTP) is not expected from the information from the crystal structure. Because the measured apparent dissociation constant is the overall dissociation constant of the ternary polymerase-dNTP-DNA complex, the increase in K m for gp5-H704A may reflect a perturbation in maintaining the DNA primer in position for catalysis.
Thus, removal of a single hydrogen bond between His 704 and the primer terminus does not significantly alter the binding affinity of the polymerase for DNA or for nucleotides. However, gp5-H704A has a much higher percentage of nucleotide turnover than the wild-type polymerase during DNA synthesis than what would be expected from the level of its exonuclease activity. When the hydrolytic function of the enzyme was monitored by the turnover rate of dTTP during DNA polymerization, the amount of free dTMP formation with gp5-H704A was 11 times higher than that with wild-type enzyme at 37°C. While wildtype T7 DNA polymerase removes 13 of every 100 nucleotides it inserts, for every 100 nucleotides that are inserted by gp5-H704A, 65 are hydrolyzed to the monophosphate. Because the exonucleolytic rate of gp5-H704A on ssDNA is comparable with that of the wild-type polymerase, the higher nucleotide turnover is not a result of a higher intrinsic exonuclease activity. It is possible that the local denaturation of the primer-template upon removal of this interaction with the polymerase increases the single-stranded nature of the 3Ј-end, thus enhancing the propensity for exonucleolytic editing.
Interactions with the primase-helicase of bacteriophage T7 encoded by gene 4 (gp4) were also examined in assays with purified proteins. Gp5-H704A catalyzes strand displacement DNA synthesis in complex with gp4, 2 implying that it interacts with gp4 in a functional mode comparable to wild-type DNA polymerase (42,43). The ability of gp5-H704A to interact with a primase-primer complex was also examined. The primase domain of gene 4 catalyzes the synthesis of tetraribonucleotides at primase recognition sites in the presence of ATP and CTP (44). These tetraribonucleotide primers are stabilized on the DNA template by gp4 until used by T7 DNA polymerase (45,46). The gp4-dependent DNA synthesis rate of T7 DNA polymerase was measured on M13 ssDNA. In this experiment, gp5-H704A has ϳ4-fold lower activity than the wild-type polymerase. 2 This difference in activity is not substantial given that gp5-H704A has ϳ8-fold lower rate of DNA synthesis on DNA primed M13 ssDNA by itself.
In experiments where bacteriophage T7 was used to infect cells that harbored the plasmid pGP5-H704A, gp5-H704A inhibited the growth of wild-type T7 phage. Biochemical experiments have not clarified the basis for this strong in vivo dominant lethal phenotype of the H704A substitution. DNA synthesis catalyzed by wild-type T7 DNA polymerase in vitro on primed M13 ssDNA is not inhibited by the addition of gp5-H704A. However, in the complementation assay, gp5-H704A, overexpressed from plasmid pGP5-H704A, is present at levels many times greater than wild-type gene 5 protein encoded by T7 phage. Thus, this gene dosage effect may account for the dominant lethal phenotype.
In summary, His 704 is an essential residue of T7 DNA polymerase. A systematic dissection of the reaction pathway supports a critical role of this residue in positioning the 3Ј-OH terminus of the primer strand in the polymerase domain in place for phosphodiester bond catalysis. Removal of this residue modifies the partitioning of DNA between the polymerase and exonuclease active sites. The higher turnover of nucleotides most likely is a manifestation of the polymerase spending a higher proportion of time with DNA in its exonuclease site. Because high DNA replication fidelity with minimal turnover of nucleotides is incumbent upon a balance between the competing polymerase and exonuclease activities, disruption of this balance may become untenable for T7 phage growth.