Bacteriophage T7 DNA Helicase Binds dTTP, Forms Hexamers, and Binds DNA in the Absence of Mg2+

The role of Mg2+ in dTTP hydrolysis, dTTP binding, hexamer formation, and DNA binding was studied in bacteriophage T7 DNA helicase (4A′ protein). The steady state k cat for the dTTPase activity was 200–300-fold lower in the absence of MgCl2, but theK m was only slightly affected. Direct dTTP binding experiments showed that the K d of dTTP was unaffected, but the stoichiometry of dTTP binding was different in the absence of Mg2+. Two dTTPs were found to bind tightly in the absence of Mg2+ in contrast to three to four in the presence of Mg2+. In the presence of DNA there was little difference in the stoichiometry of dTTP binding to 4A′. These results indicate that Mg2+ is not necessary for dTTP binding, but Mg2+ is required for optimal hydrolysis of dTTP. Gel filtration of 4A′ in the presence of dTTP without Mg2+showed that Mg2+ was not necessary, and dTTP was sufficient for hexamer formation. The hexamers formed in the presence of dTTP without Mg2+ were capable of binding single-stranded DNA. However, the 4A′ hexamers formed in the presence of dTDP with or without Mg2+ did not bind DNA, indicating that hexamer formation itself is not sufficient for DNA binding. The hexamers need to be in the correct conformation, in this case in the dTTP-bound state, to interact with the DNA. Thus, the γ-phosphate of dTTP plays an important role in causing a conformational change in the protein that leads to stable interactions of 4A′ with the DNA.

Helicases are motor proteins that translocate on DNA and catalyze separation of the double-stranded DNA into singlestranded DNAs. The ssDNA 1 regions generated by the action of helicases serve as templates during DNA replication and repair and as intermediates in various processes of DNA metabolism (1)(2)(3)(4). The strand separation activity of helicases requires hydrolysis of NTP, and the NTPase activity in helicases, as in other motor proteins, plays a central role in the enzymatic process that leads to translocation and DNA unwinding.
Studies of T7 DNA helicase and other helicases such as Escherichia coli Rep and E. coli DnaB have shown that the type of nucleotide bound to the helicase, NTP versus NDP, influ-ences its affinity for the DNA (5)(6)(7). Therefore, NTP binding and hydrolysis steps must act as switches that allow binding and release of DNA from the protein, and these steps are necessary for translocation. In the hexameric class of helicases, the MgNTP binding is also necessary for protein oligomerization (8 -13). Bacteriophage T7 DNA helicase (4AЈ protein) forms ring-shaped hexamers in the presence of MgdTMPPCP (14,15). The T7 helicase preferentially uses dTTP as its substrate for DNA unwinding (16 -19). The hexamers that are formed in the presence of MgdTMPPCP interact tightly with ssDNA, which binds in the central hole of the ring-shaped hexamer (5,20). The MgNTP is therefore involved in multiple activities of the helicase, yet the roles of Mg 2ϩ and dTTP have not been dissected. The studies in this report are directed toward understanding the roles of Mg 2ϩ and NTP in hexamer formation and DNA binding.
To understand the role of Mg 2ϩ , we have studied dTTP hydrolysis, dTTP binding, hexamer formation, and DNA binding in 4AЈ in the absence of added Mg 2ϩ and in the presence of EDTA, which chelates contaminating metal ions. Our results showed that Mg 2ϩ was required for efficient hydrolysis of dTTP into dTDP and P i . This was not surprising, since MgNTP is believed to be the substrate of most NTPases. Interestingly, Mg 2ϩ was not necessary for dTTP binding, hexamer formation, or DNA binding. Stable 4AЈ hexamers were formed in the absence of Mg 2ϩ and in the presence of dTTP, and these hexamers were capable of binding ssDNA. The affinity of the hexamers for the DNA in the absence of Mg 2ϩ was very similar to that observed in the presence of MgdTMPPCP. These studies provide novel insights into the roles of Mg 2ϩ and dTTP in the helicase mechanism. At the same time, they have provided us with conditions to preassemble the helicase hexamer on the DNA. Such preassembly is desirable for detailed biochemical characterization of the dTTPase and unwinding kinetics, since it will allow us to measure the rate of unwinding and hydrolysis that is not limited by protein assembly and DNA binding (21).

EXPERIMENTAL PROCEDURES
Nucleotides and Other Reagents-[␣-32 P]dTTP was purchased from Amersham Pharmacia Biotech, and its purity was checked and corrected for in all of the experiments. dTTP was purchased from Sigma. The nitrocellulose (BA-S) and DEAE (NA-45) membranes were purchased from Schleicher & Schuell. The Bio-Gel P-30 resin was purchased from Bio-Rad, and the polyethyleneimine cellulose TLC plates were purchased from EM Separations Technology.
Proteins and Enzymes-The 4AЈ protein was purified and stored as described previously (19). The protein concentration was determined both by absorbance measurements at 280 nm in 8 M urea (the extinction coefficient is 76,100 M Ϫ1 cm Ϫ1 ), and by the Bradford assay using bovine serum albumin as a standard. Both methods provided similar concentrations. T4 polynucleotide kinase (10 units/l) used to 5Ј-radiolabel the oligodeoxynucleotides was purchased from Boehringer Mannheim.
Synthetic Oligodeoxynucleotides-The sequence of the 30-mer DNA used in all of the experiments was as follows: AGCTT GCATC ATAGT TGCAC CTGTT ACGTT. The 30-mer was purified on a 16% polyacryl-amide gel with 3.2 M urea in Tris-borate-EDTA. The DNA was electroeluted using an Elutrap apparatus from Schleicher & Schuell. The concentration of DNA was determined spectroscopically at 260 nm in Tris-EDTA buffer, 8 M urea, using an extinction coefficient of 305,010 M Ϫ1 cm Ϫ1 . dTTPase Assays-The dTTPase assays were carried out in an 18°C room. 4AЈ protein was mixed with dTTP and [␣-32 P]dTTP with or without 30-mer DNA in the following reaction buffer: 50 mM Tris-Cl (pH 7.5), 40 mM NaCl, 0.1 mg/ml bovine serum albumin, with 10 mM MgCl 2 or 5 mM EDTA (the concentrations of the enzyme and substrate are given in the figure legends). After various times of incubation, the reactions were quenched with 1 M HCl, treated with chloroform, and neutralized with 1 M NaOH, 0.25 M Tris base. The [␣-32 P]dTDP was separated from [␣-32 P]dTTP by polyethyleneimine cellulose TLC using 0.3 M potassium phosphate, pH 3.4, as the running buffer. The radioactivity was quantitated using a PhosphorImager (Molecular Dynamics). The molar dTDP was plotted as a function of time, and the slope of the plot provided the reaction rate. The experiment was repeated at various concentrations of dTTP, and the rate of each reaction was plotted versus dTTP concentration. The curve was fit to Equation 1 to obtain the K m and k cat values.
where v is the rate of the reaction, [E] 0 k cat is the maximal rate, [S] is the concentration of dTTP, and K m is the Michaelis constant. dTTP Binding Assay-The nitrocellulose (NC) filter binding assay was used to quantitate dTTP binding to 4AЈ protein as described previously (8,22). The assays were performed at 18°C using a constant amount of the protein and increasing dTTP. The NC membrane circles (25 mm) were washed with 0.5 M NaOH for 3 min, rinsed with doubledistilled water, and equilibrated overnight in membrane wash buffer (50 mM Tris-Cl (pH 7.5), 5 mM NaCl, with either 10 mM MgCl 2 or 2 mM EDTA). 4AЈ (12 M) was mixed with dTTP (0 -100 M) and [␣-32 P]dTTP with or without 30-mer DNA (5 M) in 50 mM Tris-Cl (pH 7.5), 40 mM NaCl, 10% glycerol, in the presence of either 10 mM MgCl 2 or 0.5 mM EDTA in a total volume of 15 l. The reactions were incubated for 30 s, and 10-l aliquots were filtered through the NC membrane assembly. The membranes were washed before and after filtration with 0.6 ml of membrane wash buffer. One-microliter aliquots were spotted onto a separate nitrocellulose membrane to measure the total radioactivity. The radioactivity on the membranes was quantitated using a Phosphor-Imager (Molecular Dynamics). The results of the nucleotide binding assays were fit to a hyperbola (Equation 2) where [ES] is the concentrations of 4AЈ-dTTP complex, [E] 0 is the concentration of 4AЈ, [S] is the concentration of dTTP, and K d is the dissociation constant. HPLC Gel Filtration-The assembly of 4AЈ into hexamers was analyzed in the presence of dTTP (in the absence of Mg 2ϩ ) using HPLC gel filtration chromatography. The elution buffer consisted of 50 mM Tris-Cl (pH 7.5), 40 mM NaCl, 10% glycerol, 0.5 mM EDTA, and the appropriate concentration of dTTP (0 -250 M). Protein samples (10 l) were prepared by mixing 4AЈ (3.6 M) with the desired concentration of dTTP (0 -250 M) in the presence or absence of 30-mer DNA (3.6 M). The 4AЈ hexamer was resolved from the monomer or dimer species at a flow rate of 0.8 ml/min using a 30-cm HPLC gel-filtration column (Bio-Silect SEC 400 -5, Bio-Rad) that was preequilibrated in the elution buffer. Chromatography was performed on a Waters model 625 LC system, and proteins were detected using a Waters 470 scanning fluorescence detector ( excitation ϭ 280 nm and emission ϭ 340 nm). The fraction of hexamers was plotted against the concentration of free dTTP, and the data were fit to the Hill equation (Equation 3) to obtain the apparent equilibrium constant for hexamer formation.
where Y L is the fractional saturation, [L] is the concentration of free dTTP, K d is the apparent dissociation constant for the hexamer, and n is the Hill coefficient.
Equilibrium DNA Binding Assays-The DNA binding assay was carried out as described previously (5). The 30-mer DNA was radiolabeled with 32 P i at the 5Ј-end using [␥-32 P]ATP and T4 polynucleotide kinase. Excess [␥-32 P]ATP was removed by Bio-Gel P-30 spin gel filtration. The NC and DEAE membranes were washed with 0.5 M NaOH for 10 min, rinsed with double-distilled water, and equilibrated with the membrane wash buffer (50 mM Tris-Cl (pH 7.5), 5 mM NaCl, and either 10 mM MgCl 2 or 2 mM EDTA) for at least 1 h at room temperature. The assays were performed at constant DNA (1 M) and increasing 4AЈ (0 -20 M) concentration in buffer containing 50 mM Tris-Cl (pH 7.5), 40 mM NaCl, 10% glycerol, and in the presence of 10 mM MgCl 2 or 0.5 mM EDTA. Experiments were carried out under the following conditions: (a) in the presence of dTMPPCP (1 mM) and Mg 2ϩ (10 mM); (b) in the presence of dTTP (1 mM) and Mg 2ϩ (10 mM); (c) in the presence of dTTP (1 mM) and EDTA (0.5 mM). The reactions were preincubated for 20 min at room temperature before the samples were filtered through the membrane assembly. The membranes were washed, before and after filtration of samples, with 60 l of membrane wash buffer (above). After the samples were filtered, radioactivity on both NC and DEAE filters was quantitated using a PhosphorImager (Molecular Dynamics).

RESULTS
To investigate the role of Mg 2ϩ and dTTP in hexamer formation, hydrolysis, and DNA binding by T7 4AЈ protein, we have characterized the activities of 4AЈ protein both in the presence and absence of Mg 2ϩ . The studies in the absence of Mg 2ϩ were carried out in the presence of EDTA, which serves to chelate the contaminating metal ions in the reaction mixture. To determine the optimal amount of EDTA for use in the Mg 2ϩ -free experiments, the dTTP hydrolysis rate was measured at increasing concentrations of EDTA in the reaction. We found that, as the concentration of EDTA was increased from 0.5 M to 20 mM, the rate of dTTP hydrolysis decreased and was saturated by 0.5-1 mM EDTA (data not shown). These results indicate that the reaction mixture without EDTA contained contaminating Mg 2ϩ or other divalent cations that were able to partially activate the dTTPase activity of 4AЈ. To chelate the free Mg 2ϩ ions in the reaction mixture, we included 0.5-5 mM EDTA in all the reactions.
The Role of Mg 2ϩ in dTTP Hydrolysis Reaction Catalyzed by 4AЈ-To investigate the role of Mg 2ϩ in the 4AЈ-catalyzed dTTP hydrolysis reaction, we measured the dTTPase activity in the absence of Mg 2ϩ . Our results showed that 4AЈ was able to hydrolyze dTTP at a reduced rate in the absence of Mg 2ϩ . To further investigate this dTTPase activity, we determined the steady-state K m and k cat values of dTTPase in the absence and presence of DNA, with and without Mg 2ϩ (see Fig. 1A-D). The measured k cat and K m values are reported in Table I. In the absence of DNA and Mg 2ϩ , the k cat for dTTP hydrolysis was reduced 300-fold relative to the rate in the presence of Mg 2ϩ . The difference in the K m values of dTTPase in the absence and presence of Mg 2ϩ was relatively small. The K m increased by only 2-fold in the absence of Mg 2ϩ . In the presence of 30-mer ssDNA, we also observed a large change in the k cat and a small change in the K m value in reactions without Mg 2ϩ . The dT-TPase k cat decreased about 200-fold from the DNA-stimulated rate in the presence of Mg 2ϩ and the K m value decreased about 2-fold in the absence of Mg 2ϩ . dTTP Binding to 4AЈ with and without Mg 2ϩ -The equilibrium binding of dTTP to 4AЈ protein was measured both to determine the number of dTTPs that bind 4AЈ and their affinity in the absence of Mg 2ϩ . Fig. 2A shows that in the presence of 10 mM Mg 2ϩ , 4AЈ protein bound 3-4 dTTPs per hexamer, consistent with our previous results (8,22). In the absence of Mg 2ϩ , 4AЈ bound only 2 dTTPs per hexamer. The K d for the bound nucleotides was very close to that in the presence of Mg 2ϩ (11 Ϯ 2 M with Mg 2ϩ and 9 Ϯ 1 M without Mg 2ϩ ). The binding of dTTP was also investigated in the presence of 30-mer ssDNA.
In the presence of 10 mM Mg 2ϩ , 4AЈ bound 3-4 dTTPs per hexamer (Fig. 2B), and the same number of dTTPs were bound in the presence of DNA without Mg 2ϩ . The affinity of the dTTPs was about 2-fold weaker when Mg 2ϩ was removed (35 Ϯ 6 M without Mg 2ϩ and 13 Ϯ 2 M with Mg 2ϩ in the presence of DNA). Together, the above results indicate that Mg 2ϩ is necessary for dTTP hydrolysis, but the absence of Mg 2ϩ does not impair dTTP binding to 4AЈ to a significant extent.
The Formation of 4AЈ Hexamer in the Absence of Mg 2ϩ -The 4AЈ protein assembles into stable hexamers in the presence of MgdTMPPCP and MgdTTP (14,15). Since the above dTTP binding results show that Mg 2ϩ is not necessary for dTTP binding, we asked whether dTTP binding was sufficient, and hence Mg 2ϩ was not necessary for hexamer formation. We used HPLC gel filtration to monitor hexamer formation in 4AЈ protein. The gel filtration experiments were performed by incubating 4AЈ protein with increasing concentrations of dTTP in the presence of EDTA. The samples were applied to the gel filtra-tion column eluted with a buffer containing the same amount of dTTP and EDTA. Fig. 3A shows the gel filtration profiles of 4AЈ protein in the presence of various concentrations of dTTP. The gel filtration profiles show that 4AЈ was capable of forming stable hexamers in the absence of Mg 2ϩ , and the formation of hexamers was dependent on the concentration of dTTP. Thus, 4AЈ monomer or dimer species were converted to the hexamer species as the concentration of dTTP was increased. The dTTP dependence of the hexamer peak area was sigmoidal as expected for a ligand-linked oligomerization process (Fig. 3B). The data were fit to the Hill equation with an apparent K d of 38 M for dTTP-dependent hexamer formation and a Hill coefficient of 1.5. The gel filtration experiments were also carried out in the presence of 30-mer ssDNA (Fig. 3C). The dTTP-dependent hexamer formation occurred with a tighter K d in the presence of DNA. The apparent K d for dTTP-dependent hexamer formation in the presence of DNA was 7 M and the Hill coefficient was 2.5 (Fig. 3D). These results show that Mg 2ϩ is not necessary for hexamer formation in 4AЈ helicase, and dTTP binding is sufficient for formation and stabilization of the hexamer.
4AЈ Protein Binds DNA in the Absence of Mg 2ϩ -We have shown earlier that 4AЈ hexamer binds DNA with a high affinity only in the presence of MgdTMPPCP, the nonhydrolyzable analog of dTTP, and the interactions with the DNA are weak in the presence of MgdTDP (5). These results suggested that the dTTP-bound form of 4AЈ may interact more tightly with the DNA. However, dTTP hydrolysis in the presence of Mg 2ϩ is fast, thus we were unable to test this idea. Now we have conditions under which dTTP can be bound to 4AЈ without extensive hydrolysis, and hence we are able to test DNA binding in the presence of dTTP instead of dTMPPCP. The binding of DNA was measured in a quantitative manner using the NC-DEAE filter binding assay. The experiment was carried out using a constant concentration of radiolabeled 30-mer DNA, which was titrated with increasing concentration of the 4AЈ protein. Experiments were carried out in the presence of various ligands such as MgdTMPPCP, MgdTTP, and dTTP with EDTA. The positive control (dTMPPCP and Mg 2ϩ ) showed tight binding of DNA with a stoichiometry of one 30-mer DNA per hexamer (Fig. 4). Very little DNA binding was observed in the presence of dTTP and Mg 2ϩ . This is because the protein translocates and falls off from the 30-mer DNA when dTTP is being continuously hydrolyzed to dTDP in the presence of Mg 2ϩ . Interestingly, tight binding of DNA to the 4AЈ hexamer was observed in the presence of dTTP without Mg 2ϩ . The stoichiometry of DNA binding was very similar to that observed in the presence of MgdTMPPCP, one DNA per 4AЈ hexamer.

DISCUSSION
The MgdTTP serves as the energy source for unidirectional translocation of 4AЈ helicase on the DNA during DNA unwinding. Here we have studied the properties of 4AЈ protein in the presence and absence of Mg 2ϩ to dissect the roles of Mg 2ϩ and dTTP. The studies described in this report show that the presence of Mg 2ϩ is necessary for maximally activating the intrinsic as well as the DNA-stimulated dTTPase activity of 4AЈ protein.
The presence of Mg 2ϩ is also necessary for the helicase activity (21). We have shown that dTTP binding is not greatly affected in the absence of Mg 2ϩ , thus the lack of helicase activity in the absence of Mg 2ϩ must be due to the 200-fold slower dTTPase rate of 4AЈ in the absence of Mg 2ϩ . The reduced k cat but an almost unchanged K m of dTTPase suggests that the major role of Mg 2ϩ is in stabilization of the transition state of the dTTPase reaction. Mg 2ϩ plays a minor role in dTTP binding in the ground state. The exact mechanism of dTTP hydrolysis in helicases is not known, but it is likely to be similar to the mechanism of small GTPases that have been extensively studied. In these enzymes, it is believed that the water molecule that serves as the nucleophile is either activated by protein active site residues (23,24), or the ␥-phosphate of the NTP may itself serve to activate the water molecule (25,26). Thus, the role of Mg 2ϩ is most likely in neutralizing the negative charges that develop during the reaction. Our results also show that DNA binding by itself can activate the dTTPase activity about 10-fold in the absence of Mg 2ϩ . The structures of E. coli Rep helicase and hepatitis C virus helicase show that the DNA binding site is close to the NTP binding site (27,28). Therefore, it is possible that the DNA activates the dTTPase by a direct mechanism. Alternatively, the DNA may cause changes in the conformation of the protein active site residues that lead to stabilization of the dTTPase transition state.
dTTP binding in the absence of Mg 2ϩ showed that only two dTTPs were bound to 4AЈ. Since the nitrocellulose filter-binding assay detects sites that have high affinity, we cannot rule out the possibility that other sites are occupied in the absence of Mg 2ϩ . We can, however, conclude that only two sites are tightly bound in the absence of Mg 2ϩ as compared with three to four tight binding sites in the presence of Mg 2ϩ . The presence of DNA restored the binding of one or two dTTPs to 4AЈ. The dTTP binding in the presence of DNA was therefore the same with or without Mg 2ϩ . We have shown previously that 4AЈ has two kinds of dTTP binding sites (29). Approximately half of the dTTP binding sites on the hexamer appear to be noncatalytic. The dTTPs bound at the noncatalytic sites do not turn over very rapidly and hence these dTTPs are most likely not involved in providing the energy for translocation and DNA unwinding. Future studies are directed toward determining whether the dTTPs that are bound in the absence of Mg 2ϩ are at the catalytic sites or the noncatalytic sites.
The 4AЈ protein was found to assemble into stable hexamers even in the absence of Mg 2ϩ . Thus, Mg 2ϩ is not necessary, and dTTP or dTDP binding is sufficient, for hexamer formation. Since no high resolution structural information of hexameric helicases is available at the present time, it is not known exactly where dTTP binds with respect to the hexamer subunits and how it promotes hexamer formation. There is, however, considerable evidence that the structure of hexameric helicases may be similar to the F1-ATPase enzyme, which is hexameric (27, 30 -33). In the F1-ATPase, it is known that ATP molecules bind near the ␣, ␤ subunit interface (30), which may be the case with the hexameric helicases as well. That is, NTP may be bound at the subunit interface. The binding of dTTP at the interface provides a simple mechanism by which it can stabilize hexamer formation. We show here that Mg 2ϩ is not necessary for hexamer formation. This behavior of 4AЈ is different from that of E. coli DnaB helicase that was shown to form hexamers in the presence of Mg 2ϩ alone (34). Since we have not been able to detect stable hexamers of 4AЈ protein in the presence of Mg 2ϩ alone (data not shown), we believe that nucleotide binding, either dTTP or dTDP, is necessary for stable hexamer formation.
Our studies show that the ␥-phosphate plays a very important role in regulating the interactions of the helicase with the DNA. Previous studies have shown that 4AЈ helicase interacts tightly with the DNA in the presence of MgdTMPPCP, and very little DNA binding was observed in the presence of MgdTDP (5). We were not sure if dTMPPCP was a good analog of dTTP because the bond lengths and the bond angle of P-C-P are different from those of P-O-P. It is possible that the structure of the dTMPPCP-bound 4AЈ is different from the dTTP-bound form, and the dTMPPCP-bound form resembles an intermediate in the dTTPase reaction pathway that has strong interactions with the DNA. Part of the reason for believing this was because weak DNA binding was observed in the presence of MgdTTP. Now we are able to observe tight binding of DNA in the presence of dTTP without Mg 2ϩ , conditions under which dTTP hydrolysis is very slow and hence translocation and dissociation of 4AЈ from the DNA is slow. These results indicate that the dTTP-bound form of 4AЈ does interact tightly with the DNA. We can also conclude that hexamer formation is not sufficient for DNA binding, since no DNA binding was observed in the presence of dTDP, although stable hexamers were formed in the presence of MgdTDP (8) and dTDP (data not shown). It appears that the ␥-phosphate of dTTP may be involved in causing a critical conformational change in the protein that "opens up" the DNA binding site on a 4AЈ subunit. We can infer that the hydrolysis of dTTP or the dissociation of P i leads to dissociation of the DNA. These steps provide the switch mechanism that allows the helicase subunit to switch from a tight DNA binding state to a weak one. Since the DNA predominately interacts with one or two subunits (20), the resulting DNA binding and release steps can occur in a sequential manner among the hexamer subunits, providing a mechanism for processive translocation of the helicase on the DNA.
Finally, the findings from this study are likely to be general for helicases that assemble into hexamers in the presence of MgNTP (8 -13). This study has provided us with conditions where a complex of hexamer bound to the DNA can be preformed without extensive dTTP hydrolysis occurring in the complex. This is the first time that we have been able to form such a complex. The ability to preform this complex will allow us to examine the presteady state kinetics of the DNA-stimu-lated dTTP hydrolysis reaction, and the kinetics of DNA unwinding at a replication fork, without complications from oligomerization and/or DNA binding processes.