Nucleotide-dependent binding of the gene 4 protein of bacteriophage T7 to single-stranded DNA.

The gene 4 protein of bacteriophage T7 is a multifunctional enzyme that catalyzes (i) the hydrolysis of nucleoside 5'-triphosphates, (ii) the synthesis of tetraribonucleotide primers at specific recognition sequences on a DNA template, and (iii) the unwinding of duplex DNA. All three activities depend on binding of gene 4 protein to single-stranded DNA followed by unidirectional 5' to 3' translocation of the protein (Tabor, S., and Richardson, C. C. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 205-209). Binding of gene 4 protein to single-stranded DNA, assayed by retention of DNA-protein complexes on nitrocellulose filters, is random with regard to DNA sequence. Although gene 4 protein does not bind to duplex DNAs, the presence of a 240-nucleotide-long single-stranded tail on a 7200-base pair duplex DNA molecule is sufficient for gene 4 protein to cause retention of the DNA on a filter. The binding reaction requires, in addition to MgCl2, the presence of a nucleoside 5'-triphosphate, but binding is not dependent on hydrolysis; nucleoside 5'-diphosphate will substitute for nucleoside 5'-triphosphate. Of the eight common nucleoside triphosphates, dTTP promotes optimal binding. The half-life of the gene 4 protein-DNA complex depends on both the secondary structure of the DNA and on whether or not the nucleoside 5'-triphosphate cofactor can be hydrolyzed. Using the nonhydrolyzable nucleoside 5'-triphosphate analog, beta,gamma-methylene dTTP, the half-life of the gene 4 protein-DNA complex is greater than 80 min. In the presence of the hydrolyzable nucleoside 5'-triphosphate, dTTP, the half-life of the gene 4 protein-DNA complex using circular M13 DNA is at least 4 times longer than that observed using linear M13 DNA.

The gene 4 protein of bacteriophage T7 is a multifunctional enzyme that catalyzes (i) the hydrolysis of nucleoside 5'-triphosphates, (ii) the synthesis of tetraribonucleotide primers at specific recognition sequences on a DNA template, and (iii) the unwinding of duplex DNA. All three activities depend on binding of gene 4 protein to single-stranded DNA followed by unidirectional 5' to 3' translocation of the protein (Tabor, S . , and Richardson, C. C. (1981) Proc. Nutl.
Acad. Sei. U. S. A. 78, 205-209). Binding of gene 4 protein to single-stranded DNA, assayed by retention of DNA-protein complexes on nitrocellulose filters, is random with regard to DNA sequence. Although gene 4 protein does not bind to duplex DNAs, the presence of a 240-nucleotide-long single-stranded tail on a 7200-base pair duplex DNA molecule is sufficient for gene 4 protein to cause retention of the DNA on a filter. The binding reaction requires, in addition to MgC12, the presence of a nucleoside 5'-triphosphate, but binding is not dependent on hydrolysis; nucleoside 5'-diphosphate will substitute for nucleoside 5"triphosphate. Of the eight common nucleoside triphosphates, dTTP promotes optimal binding. The half-life of the gene 4 protein-DNA complex depends on both the secondary structure of the DNA and on whether or not the nucleoside 5'-triphosphate cofactor can be hydrolyzed. Using the nonhydrolyzable nucleoside 5'-triphosphate analog, B,y-methylene dTTP, the half-life of the gene 4 protein-DNA complex is greater than 80 min. In the presence of the hydrolyzable nucleoside 5'triphosphate, dTTP, the half-life of the gene 4 protein-DNA complex using circular M13 DNA is at least 4 times longer than that observed using linear M13 DNA.
Biochemical and genetic analysis of bacteriophage T7 has revealed a requirement for the product of gene 4 of the phage for DNA replication (2)(3)(4)(5)(6). Initially, purification of gene 4 protein relied on an in vitro complementation assay (7)(8)(9). Subsequent studies of the purified protein have shown it to * This investigation was supported by United States Public Health Service Grant AI-06045 and Grant NP-1L from the American Cancer Society, Inc. This is Paper 29 in a series entitled "Replication of Bacteriophage T7 Deoxyribonucleic Acid." The previous paper is Ref. 1. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. be a multifunctional enzyme with single-stranded DNA-dependent nucleoside 5'-triphosphatase activity (10, l l ) , sitespecific primase activity (9,12-16), and DNA helicase activity (1,8,17,18). T o prime DNA synthesis the gene 4 protein synthesizes tetraribonucleotides, pppACCC and pppACCA, that are complementary to four of the nucleotides found in the specific pentanucleotide (3'-CTGGG/T-5') recognition sequence (9, 15, 16). The helicase activity of the gene 4 protein, initially inferred from its ability to stimulate DNA synthesis catalyzed by T7 DNA polymerase on duplex DNA templates (8,12,17,19), has been recently demonstrated directly using a novel DNA substrate (1). The gene 4 protein is found in two forms with molecular weights of 58,000 and 66,000 which purify together (4, 8,9,19,20). Two initiation codons located 189 base pairs apart account for the two species of gene 4 protein which have common C termini but different amino ends (21).
Much is already known about the mechanisms by which the gene 4 protein catalyzes primer synthesis and the unwinding of duplex DNA (1, 5, 9-16). Like the nucleoside 5'triphosphatase activity, both the primase and helicase activities require the presence of single-stranded DNA and a NTP.' Furthermore, the addition of a nonhydrolyzable NTP analog, P,y-methylene dTTP, inhibits both priming and helicase reactions, indicating that NTP hydrolysis is coupled to both activities (1,10,11,19). The relative rates of utilization of priming sites on circular single-stranded DNA suggests that these recognition sites on the single-stranded DNA are encountered by the gene 4 protein in a nonrandom fashion (16).
In fact, the utilization of primer sites can be best explained by a mechanism in which the gene 4 protein binds at random sites on the single-stranded DNA and remains bound while translocating 5' to 3' along the single strand of DNA (16). Studies on the nucleoside 5"triphosphate hydrolysis and the helicase activities of gene 4 protein using defined DNA substrates support such a mechanism (1,11,19). Translocation in a single direction must be coupled to some irreversible chemical process, and the obvious candidate is the hydrolysis of NTPs. A common feature of the known reactions catalyzed by the gene 4 protein is that each activity requires binding of gene 4 protein to single-stranded DNA. We have now applied a direct filter binding method to demonstrate binding of the protein to DNA. Binding of the gene 4 protein to DNA is found to require nucleotides, and the rate of dissociation of the complex on defined DNA molecules supports unidirectional translocation of the protein on single-stranded DNA.

Materials
Bacterial Strains and Bacteriophages-Escherichia coli 71.18 has been described (22). Mutants requiring a high concentration (50 gg/ ml) of exogenous thymine were isolated as described by Miller (23). Second site mutants requiring a low concentration (2 pg/ml) of exogenous thymine were isolated by plating mutants requiring a high level of exogenous thymine on M-9 minimal media plates containing 2 pg/ml added thymine. Colonies were isolated and determined to have a thymine requirement that was satisfied by the addition of 2 pg/ml thymine. All other bacterial strains and bacteriophages have been described (1,11).
Enzymes"T7 gene 4 protein, isolated as previously described ( l l ) , is greater than 95% pure as judged by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (24) and has a specific activity of 33,000 units/mg. All restriction endonucleases were purchased from New England Biolabs; reaction conditions were those suggested by the supplier. Phage T4 polynucleotide kinase was purchased from New England Nuclear; reaction conditions were those suggested by the supplier. Bovine serum albumin was from Miles Laboratories and Proteinase K was from Boehringer Mannheim. E.
coli alkaline phosphatase was obtained from Worthington and further purified as previously described (25).
DNA and Nucleotides-Phage M13mp7 single-stranded DNA and RFI DNA were purified as described (19). M13mp7 [3HJDNA was prepared as follows. The low thymidine-requiring mutant of E. coli 71.18 was grown in Fraser's media containing 2 pg/ml thymine to an ASW of 0.5, and 5 pCi/ml of [3H]thymidine (New England Nuclear) were added. Aeration was continued at 37 "C for 15 min. M13 phage were added at a multiplicity of infection of 0.1 to 0.5, and aeration was continued for 6 h at 37 "C. The phage were harvested by precipitation with 6% polyethylene glycol and the phage DNA isolated as described (19). The specific activity of the DNA obtained was 34.8 cpm/pmol. Singly nicked pBR322 dimer DNA, prepared as described (26), was the generous gift of Dr. Michael J. Engler (University of Texas). The topologically stable replication fork DNA substrate (19) was the generous gift of Dr. Robert L. Lechner (Harvard Medical School). 4x174 single-stranded phage DNA was purchased from New England Biolahs. Poly(dT), all unlabeled nucleoside 5'-diphosphates, nucleoside 5'-triphosphates, P,y-methylene dTTP, and P,y-methylene rATP were from P-L Biochemicals. 3H-labeled nucleotides were from New England Nuclear; [y-"PIrATP was from ICN.

Methods
Binding Assay-The assay used to detect binding of gene 4 protein to DNA has been described (27). The standard binding reaction mixture (20 pl) contained 40 mM Tris.HCI (pH 7.5), 10 mM MgClz, 10 mM dithiothreitol, 50 pg/ml bovine serum albumin, 5% glycerol (from enzyme preparation), 10 p~ M13 single-stranded [3H]DNA, 150 PM P,y-methylene dTTP (unless otherwise indicated), and gene 4 protein. This reaction mixture is essentially the same as the one used to measure helicase activity of the gene 4 protein (1). After incubation for 10 min at 37 "C the reaction mixture was diluted to 3.0 ml with 40 mM Tris.HC1 (pH 7.5), 10 mM MgClz, 10 mM dithiothreitol, 50 pg/ml bovine serum albumin that had been warmed to 37 "C, and the solution was filtered through a nitrocellulose filter (HAWP, Millipore) at a flow rate of 4 ml/min. Filters were dried and the radioactivity bound to the filter determined in a liquid scintillation counter. Background levels of DNA binding to filters in the absence of gene 4 protein (less than 5%) have been subtracted from all reported values. In control experiments the absence of a nucleotide in the mixture used to dilute the reaction to 3.0 ml was shown to have no effect on the binding result. Nitrocellulose filters were pretreated by boiling in distilled water for 20 min and stored at room temperature in the solution used to dilute the reaction mixtures.
Preparation of 5"End-labeled HaeIIl Restriction Fragments-HaeIII restriction fragments from 4x174 single-stranded phage DNA were prepared by digesting 4x174 single-stranded DNA (10 pg) with 10 units of HaeIII restriction enzyme overnight at 37 "C. The resulting DNA fragments were dephosphorylated by incubation with E. coli alkaline phosphatase at 65 "C for 45 min. Proteinase K (20 pg) and sodium dodecyl sulfate (0.1%) were then added and the reaction mixture incubated at 45 "C for 30 min. The DNA was extracted with phenol and precipitated with ethanol. After dissolving the precipitate in a minimal volume of 5 mM Tris .HCI (pH 7.5) the DNA was incubated with T4 polynucleotide kinase (18 units) for 45 min at 37 "C in the presence of 0.1 mCi of [y-32P]rATP (7000 Ci/mmol) to label the 5'-ends of the DNA. After phenol extraction, unreacted radioactive rATP was removed from the DNA by gel filtration through a 1.2-ml Sepharose GB-CL column equilibrated with 10 mM Tris.HC1 (pH 7.5), 0.5 mM EDTA, 100 mM NaC1. The void volume of the column, which contained the DNA, was pooled and used in binding reactions.
Preparation of 5"End-labeled Duplex DNA Substrates-The topologically stable replication fork DNA substrate (0.18 pg) was 5'end labeled by incubating the DNA with T4 polynucleotide kinase (18 units) in the presence of 0.1 mCi of [y-32P]rATP (7000 Ci/mmol) for 60 min at 37 'C. After phenol extraction, unreacted radioactive rATP was removed as described above.
Nicked pBR322 dimer DNA (1.25 pg) was 5'-end labeled after incubation with E. coli alkaline phosphatase at 65 "C and proteinase K as described above. The dephosphorylated DNA was extracted with phenol and precipitated with ethanol. The DNA was resuspended in a minimal volume of 5 mM Tris. HCl (pH 7.5) and incubated with T4 polynucleotide kinase as described above. The 5'-end-labeled DNA was phenol extracted and the unreated radioactive rATP removed as described.
Preparation of Linear Single-stranded M13 rHlDNA"M13 [3H] DNA (16 pg) was incubated with 10 ng of pancreatic DNase (Boehringer Mannheim) in 40 mM Tris.HC1 (pH 7.5), 10 mM MgCL, 10 mM dithiothreitol, 50 pg/ml bovine sreum albumin for 1 min at 23 "C. The reaction was terminated by the addition of 10 mM EDTA and then heated at 70 "C for 20 min. Greater than 90% of the circular DNA molecules were converted to linear molecules as judged by electrophoresis in a 1.0% agarose gel at 20 V/cm.
Other Methods-Polyacrylamide gel electrophoresis in the presence of 8 M urea was carried out by the method of Maxam and Gilbert (28). DNA concentrations were determined by directly measuring absorbance at 260 nm and are expressed as nucleotide equivalents except as indicated in Fig. 4.

RESULTS
The three activities of the phage T7 gene 4 protein, helicase, primase, and DNA-dependent NTPase, all apparently require binding of the protein to single-stranded DNA (see the introduction). However, binding has never been demonstrated directly. The retention of nucleoprotein complexes by nitrocellulose filters has provided a useful method for characterizing interactions between protein and nucleic acids (see Ref. 27). We have applied this method to study the binding of gene 4 protein to various DNAs. Fig. 1, in the presence of the nonhydrolyzable nucleoside 5'triphosphate analog P,y-methylene dTTP, the addition of gene 4 protein to single-stranded M13 DNA leads to the retention of more than 70% of the DNA on a nitrocelluose filter. The addition of more gene 4 protein to the binding reaction mixture does not significantly increase the amount of DNA retained. The reason for this is unknown, but it may reflect the efficiency with which the DNA-protein complexes are retained on the nitrocellulose filter under the conditions of the binding assay. When 70% of the single-stranded DNA is retained on the filter there are approximately 100 gene 4 protein molecules/DNA molecule in the reaction mixture.* The binding of a single gene 4 protein molecule to a molecule of M13 DNA should be sufficient to cause retention of the DNA-protein complex on the filter. The requirement for excess gene 4 protein may reflect a low percentage of active enzyme molecules or a low efficiency of retention of the complex on the filter due to conditions of the assay. It is interesting to note that the gene 4 protein stimulation of DNA This estimation of the number of gene 4 protein molecules/DNA substrate molecule is based on the following: (i) all gene 4 protein molecules are active, (ii) a molecular weight for gene 4 protein of 60,000, and (iii) a monomeric active species of gene 4 protein.  The retention of duplex M13 RFI DNA under the same conditions is considerably lower at low concentrations of gene 4 protein (Fig. 1). The binding to duplex DNA is approximately 7-fold lower than binding to single-stranded DNA. The binding interaction observed with M13 RFI DNA may, in fact, reflect the presence of regions of single-stranded DNA in the supercoiled RFI DNA. This interpretation is supported by the finding that nicked circular duplex DNA is far less effective in binding to gene 4 protein. The binding of gene 4 protein to nicked pBR322 DNA is 10-fold less than binding to a circular duplex DNA bearing a single-stranded tail (see Fig. 4).

Gene 4 Protein Binds Single-stranded DNA-As shown in
Requirements for Binding to Single-stranded DNA-The requirements for binding of gene 4 protein to single-stranded DNA are shown in Table I. The binding reaction is dependent on the presence of nucleotide and MgCl,. In the absence of added MgClz or in the presence of excess EDTA, the amount of DNA retained on the filter is dramatically reduced. The addition of 50 or 100 mM NaCl had little or no effect on the binding reaction ( Table I). The MgC12 and NaCl requirements for binding to single-stranded DNA are essentially identical to the requirements for helicase activity and for DNA-dependent NTP hydrolysis (1,ll).
Binding to Single-stranded DNA Requires a Nucleotide Cofactor-The binding of gene 4 protein to single-stranded M13 DNA is dependent on the presence of a nucleoside 5"diphosphate, nucleoside 5'-triphosphate, or a NTP analog (Table I, Fig. 2 ) . In the absence of added nucleotide there is essentially no retention of single-stranded DNA on the filter (less than 3%) in the presence of gene 4 protein. The binding that does occur in the absence of added NTP at high concentrations of gene 4 protein may be due to the presence of trace amounts of NTP in the preparation of gene 4 protein. Although the binding reaction requires the presence of a NTP, it is not dependent on NTP hydrolysis. In fact, the nonhydrolyzable nucleoside triphosphate analog, @,?-methylene dTTP, is the most effective nucleotide cofactor we have found (Fig. 2). At low concentrations of gene 4 protein more DNA is retained on the filter in the presence of P,y-methylene dTTP than in the presence of dTTP. The concentration of &y-methylene dTTP required to obtian maximal binding is less than 100 pM ( d a t a not shown). Increasing concentrations of P,y-methylene dTTP up to 1 mM have no effect on the amount of DNA bound by gene 4 protein.
Gene 4 protein is known to hydrolyze seven of the eight commonly occurring nucleoside 5'-triphosphates in the presence of single-stranded DNA (rCTP is not hydrolyzed) (11). All eight NTPs were tested separately in binding reactions as well as two nonhydrolyzable nucleoside 5'-triphosphate analogs (Table 11). Interestingly, of the hydrolyzable NTPs tested only dTTP can effectively serve as the NTP cofactor required by gene 4 protein for binding to single-stranded DNA. Either nonhydrolyzable NTP analog, &y-methylene dTTP or &ymethylene rATP, may serve as cofactors for the binding reaction (Table 11). Since the nonhydrolyzable analog of ATP, P,y-methylene rATP, is active we tested the ability of ATP -+ P,Y-rnethylene dTTP, The data presented for binding in the presence of P,y-methylene dTTP is the same as that presented in Fig. 1. to function as a cofactor at concentrations up to 5 mM (data not shown). At the highest concentration of ATP tested, this nucleoside 5"triphosphate failed to promote binding. Surprisingly, dTDP serves effectively as a cofactor in the gene 4 protein binding reaction (Table 11). The binding curve obtained when dTDP is substituted for d T T P in the reaction mixture is essentially the same as the binding curve obtained when d T T P is used as the NTP cofactor (data not shown). In addition, the concentration of nucleoside 5"diphosphate required to obtain maximal binding is similar to that required of @,y-methylene d T T P (less than 100 p~) . These results indicate that the affinity of gene 4 protein for DNA is unchanged in the presence of dTDP, a product of d T T P hydrolysis. It is unlikely that a contaminant of nucleoside 5'-triphosphate in the nucleoside 5"diphosphate is responsible for the binding as the concentration of nucleoside 5'-diphosphate required to obtain maximal binding is too low to permit significant levels of nucleoside 5"triphosphate contamination. Analysis of dTDP by thin layer chromatography indicates contamination by d T T P of less than 2%. 100 p~ dTDP is sufficient to promote maximal binding, but 2 PM d T T P is insufficient to observe any retention of DNA on the filter. We conclude that both dTTP and dTDP will serve as cofactors in the gene 4 protein binding reaction; dTMP is not active (Table 11).

Reaction mixtures were as described under "Experimental Procedures'' using 240 ng of gene 4 protein and 1 mM of the indicated nucleoside triphosphate. Reaction mixtures containing dTDP and dTMP were altered to contain nucleotide at a concentration of 150 PM.
Binding to Single-stranded DNA Is Random with Respect to DNA Sequence-A comparison of the rates of utilization of primase recognition sites on 4 x 1 7 4 DNA (16) suggested that binding of gene 4 protein to single-stranded DNA is random with regard to DNA sequence. In order to directly demonstrate this we have incubated gene 4 protein with single-stranded 5"end-labeled HaeIII restriction fragments from 4 x 1 7 4 DNA and determined which fragments are retained by gene 4 protein on a nitrocellulose filter (Fig. 3). Wit,h all concentrations of gene 4 protein used (Fig. 3, lanes  2-4) all the restriction fragments are represented on the polyacrylamide gel. This is a good indication that gene 4 protein binds single-stranded DNA of any sequence. It is important to note that the predominant gene 4 protein primase recognition sequences, 3'-CTGGG/T-5' (16), are not present in all of the restriction fragments, while several of the restriction fragments have more than one primase recognition sequence. Therefore, the recognition sequence is not required for a stable complex between gene 4 protein and singlestranded DNA. In addition, to eliminate the possibility of some oligonucleotide recognition sequence, which does occur in all of the HaeIII restriction fragments, being responsible for the binding we have used poly(dT) as a competitor for the binding of gene 4 protein in this reaction (data not shown). Poly(dT) effectively competes with the labeled 6x174 DNA fragments for gene 4 protein in the binding reaction. We conclude that binding of gene 4 protein to single-stranded DNA has little or no sequence dependence. protein using both a nicked, duplex circular DNA molecule and the preformed replication fork DNA substrate (Fig. 4). Gene 4 protein causes the retention of the DNA molecule containing the replication fork on the nitrocellulose filter. Retention of the nicked duplex DNA molecule on the filter by gene 4 protein is nearly an order of magnitude lower. These data suggest that the presence of a single-stranded segment of DNA is sufficient to allow binding of gene 4 protein to a DNA molecule that is otherwise completely duplex. We conclude that gene 4 protein does not stimulate DNA synthesis catalyzed by Form I1 of T7 DNA polymerase on a nicked duplex DNA template because it is unable to bind the DNA substrate and unwind the duplex for the DNA polymerase. Dissociation Rate of Gene 4 Protein-DNA Complex-In order to assess the stability of gene 4 protein-DNA complexes, the half-time for dissociation of gene 4 protein from singlestranded DNA when challenged with a high concentration of an unlabeled competing DNA was determined using both linear and circular single-stranded labeled DNA (Fig. 5). In the presence of P,y-methylene dTTP, the half-time for dissociation of gene 4 protein from circular single-stranded DNA  After incubation for 5 min at 37 "C an aliquot (20 pl) was removed to determine the fraction of DNA bound as described under "Experimental Procedures," and competing unlabeled single-stranded M13 DNA was added to a final concentration of 92 p~. Aliquots (20 pl) were removed at the indicated times, and the fraction of labeled DNA bound was determined. Nucleotide concentrations were 150 PM for P,y-methylene dTTP and 1 mM for dTTP. A control experiment in which gene 4 protein was added to a mixture of labeled and unlabeled M13 DNA showed retention of less than 10% of the labeled DNA on the filter. . " -. , single-stranded circular M13 DNA and &y-methylene dTTP; 0---0, singlestranded circular M13 DNA and dTTP; U , single-stranded linear M13 DNA and dTTP. is greater than 20 min. Even after 80 min no significant reduction in the amount of labeled DNA retained on the filter was observed (data not shown). When dTTP is used, allowing NTP hydrolysis and translocation of gene 4 protein along the DNA molecule, the half-life of the DNA-gene 4 protein complex is approximately 3 min at 37 "C (Fig. 5). Thus, the gene 4 protein forms a relatively stable complex with singlestranded DNA even under conditions of NTP hydrolysis. If, on the other hand, the same measurements are made using linear single-stranded M13 DNA we obtain a different set of results. The half-time for dissociation of gene 4 protein from an equal-sized linear DNA molecule is still greater than 20 min in the presence of &y-methylene dTTP (data not shown). In addition, there is virtually no difference between the binding curves obtained using either linear or circular singlestranded DNA in the presence of &y-methylene dTTP. However, using dTTP the half-time for dissociation from linear molecules is about 45 s at 37 "C. This result indicates that, under conditions that permit NTP hydrolysis, the gene 4 protein dissociates rapidly from linear single-stranded DNA. Since NTP hydrolysis is coupled to unidirectional translocation of gene 4 protein along single-stranded DNA, this may be interpreted to mean that translocation to the end of the DNA results in dissociation of the gene 4 protein-DNA complex. On a circular DNA molecule gene 4 protein remains bound to the DNA, even under conditions of NTP hydrolysis, for a relatively long period of time as it processively translocates along the single-stranded DNA since ends are not encountered.

DISCUSSION
The gene 4 protein of bacteriophage T7 catalyzes three distinct reactions: (i) DNA-dependent hydrolysis of NTPs, (ii) site-specific template-directed synthesis of tetraribonucleotides, and (iii) unwinding of duplex DNA. All three of these reactions require binding of gene 4 protein to DNA (1, 10, 11, 16). The primase (9, 14, 16) and single-stranded DNAdependent NTPase (10, 11) activities of gene 4 protein are manifest only on single-stranded DNA, and gene 4 protein requires single-stranded DNA adjacent to duplex DNA for unwinding of the duplex to occur (1,19).
The gene 4 protein requires the presence of single-stranded DNA for nucleoside 5'-triphosphatase activity (10) suggesting that the gene 4 protein must bind DNA during the catalytic cycle of nucleotide hydrolysis. In this paper we have shown that the gene 4 protein can form a stable complex with singlestranded DNA in the presence of a nucleotide cofactor. Taken together, these observations support a model (Fig. 6) in which the gene 4 protein interacts with a nucleoside 5'-triphosphate, perhaps in the absence of DNA (Fig. 6, I), but is unable to catalyze the hydrolysis of the NTP unless bound to singIestranded DNA. The gene 4 protein-NTP complex binds to single-stranded DNA (Fig. 6, 11), perhaps leading to a conformational change in the gene 4 protein, such that it will now catalyze the hydrolysis of the bound NTP. It is interesting that not all NTPs are capable of promoting the formation of a stable complex between gene 4 protein and single-stranded DNA. In fact, of the ones tested, the only hydrolyzable nucleoside 5"triphosphate capable of serving as a cofactor for binding is dTTP. Previous results indicate that although seven of the eight NTPs tested are hydrolyzed by gene 4 protein and, therefore, presumably interact with gene 4 protein, only four of these are hydrolyzed to an appreciable extent (11). It should be noted that of these four NTPs dTTP also has the lowest K , value for both the NTP hydrolysis reaction and for unwinding DNA. It is likely that dTTP promotes the strongest binding interaction between gene 4 protein and single-stranded DNA. The assay used to detect this interaction is relatively insensitive when compared to the DNAdependent NTP hydrolysis assay. Interactions between gene 4 protein and NTPs detected in the NTP hydrolysis assay may not be observed with the less sensitive DNA-binding assay (1, 11). The nonhydrolyzable analog of rATP, P,ymethylene rATP, does serve as a cofactor for the binding reaction, albeit not as well as @,y-methylene dTTP. However, even at high concentrations, rATP does not promote binding of gene 4 protein to single-stranded DNA. Apparently the gene 4 protein-rATP-DNA complex is weak. The reason for this result is unknown; however, it is clear that the interactions between gene 4 protein and rATP must be different from those of the other nucleotides since rATP is also the initiating nucleotide of the tetraribonucleotide primer synthesized by gene 4 protein (13, 15).
It is equally interesting that dTDP promotes stable binding, although dTMP does not. Even in the presence of one of the I II m products of hydrolysis, the gene 4 protein remains bound to single-stranded DNA. This is consistent with the ability of the gene 4 protein to translocate along single-stranded DNA. The enzyme would be expected to remain bound to the DNA throughout the hydrolytic cycle (11,16).
The binding between gene 4 protein and single-stranded DNA is not sequence specific (Fig 6,111). The gene 4 protein binds all the single-stranded HaeIII restriction fragments of 4x174, indicating no sequence specificity in the binding interaction. In addition, poly(dT) effectively competes with the fragments for binding of gene 4 protein. The original studies which demonstrated the gene 4 protein to prime DNA synthesis at specific sites also suggested that the protein binds randomly to single-stranded DNA and translocates in the 5' to 3' direction (16). This behavior is in contrast to that of some of the prepriming and priming proteins of E. coli. Both the dnaG protein and the n' protein recognize and interact with specific sequences on the DNA template (30,31). Binding to single-stranded DNA without regard for sequence is, however, expected of an enzyme that is responsible for opening the helix at a replication fork. Nonspecific binding to singlestranded DNA is consistent with results presented earlier (11) suggesting that gene 4 protein dose not remain bound for determinable periods of time at primase recognition sequences under the conditions of primer synthesis.
The complex formed between gene 4 protein and singlestranded DNA is extremely stable in the presence of a nonhydrolyzable nucleotide cofactor. Under the conditions we have used, the half-life of the complex is greater than 80 min at 37 "C. When dTTP is substituted for &y-methylene dTTP the half-life of the complex is reduced to approximately 3 min at 37 "C. This, however, is still a remarkably stable complex considering the fact that the protein is actively hydrolyzing NTP and translocating along the single strand of DNA (Fig.   6, ZV). The stability of the complex formed in the presence of dTTP depends on the secondary structure of the DNA. The complex with circular single-stranded DNA has a halflife of more than 150 s; the half-life of the complex with linear single-stranded DNA is about 45 s. These data suggest that the gene 4 protein is translocating along the single-stranded DNA effector as it hydrolyzes NTP. The circular DNA is infinitely long and, therefore, the gene 4 protein remains bound for a considerable period of time. When linear DNA is substituted for circular DNA, the gene 4 protein reaches a terminus and is forced to dissociate and rebind a new DNA molecule.
The binding interaction between gene 4 protein and singlestranded DNA is more complex than originally envisioned in that it requires the presence of a nucleotide cofactor. The model presented in Fig. 6 is consistent with data currently available. The gene 4 protein (I) binds NTP (11)  tively, free gene 4 protein may bind weakly to DNA but this complex is stabilized by subsequent binding of NTP. Once bound to the DNA, the gene 4 protein translocates unidirectionally 5' to 3' (16) along the single-stranded DNA template (IV). The unidirectional translocation is coupled to NTP hydrolysis and is essential for both the primase and helicase activities of gene 4 protein (1, 11,16,18,19). Since both dTDP and dTTP promote binding of gene 4 protein to singlestranded DNA, both complexes 111 and IV (Fig. 6) are relatively stable. Displacement of dTDP by dTTP must occur in such a way that the gene 4 protein remains bound to the DNA since additional cycles of nucleotide hydrolysis and translocation occur without release of the DNA.
Recently, two enzymatically distinct forms of T7 DNA polymerase have been described (20, 26, 29). Form I of T7 DNA polymerase catalyzes strand displacement synthesis on a nicked DNA primer-template; its rate of DNA synthesis is stimulated by gene 4 protein severalfold. Form I1 of T7 DNA polymerase does not catalyze a strand displacement synthesis reaction and, in addition, cannot be stimulated by T7 gene 4 protein on a nicked duplex primer-template. If, however, a stable replication fork containing a single-stranded 5'-tail is used, the rate of synthesis catalyzed by Form I1 of T7 DNA polymerase can be stimulated by gene 4 protein (19). In view of the results presented here, we conclude that gene 4 protein is unable to bind and initiate unwinding of a nicked duplex DNA molecule. When a single-stranded region is present, either created by strand displacement synthesis (by Form I of T7 DNA polymerase) or as a pre-existing part of the DNA substrate, the gene 4 protein binds to it and translocates 5' to 3' to the fork to unwind the DNA in concert with the advancing DNA polymerase. The result is an increased rate of DNA synthesis catalyzed by the combined action of gene 4 protein and T7 DNA polymerase. The enzymatic activities of gene 4 protein coupled with those of T7 DNA polymerase are sufficient to accomplish synthesis on both strands of template DNA.