The DNA binding properties of the Escherichia coli RecQ helicase.

The RecQ helicase family is highly conserved from bacteria to men and plays a conserved role in the preservation of genome integrity. Its deficiency in human cells leads to a marked genomic instability that is associated with premature aging and cancer. To determine the thermodynamic parameters for the interaction of Escherichia coli RecQ helicase with DNA, equilibrium binding studies have been performed using the thermodynamic rigorous fluorescence titration technique. Steady-state fluorescence anisotropy measurements of fluorescein-labeled oligonucleotides revealed that RecQ helicase bound to DNA with an apparent binding stoichiometry of 1 protein monomer/10 nucleotides. This stoichiometry was not altered in the presence of AMPPNP (adenosine 5'-(beta,gamma-imido) triphosphate) or ADP. Analyses of RecQ helicase interactions with oligonucleotides of different lengths over a wide range of pH, NaCl, and nucleic acid concentrations indicate that the RecQ helicase has a single strong DNA binding site with an association constant at 25 degrees C of K=6.7 +/- 0.95 x 10(6) M(-1) and a cooperativity parameter of omega=25.5 +/- 1.2. Both single-stranded DNA and double-stranded DNA bind competitively to the same site. The intrinsic affinities are salt-dependent, and the formation of DNA-helicase complex is accompanied by a net release of 3-4 ions. Allosteric effects of nucleotide cofactors on RecQ binding to DNA were observed only for single-stranded DNA in the presence of 1.5 mM AMPPNP, whereas both AMPPNP and ADP had no detectable effect on double-stranded DNA binding over a large range of nucleotide cofactor concentrations.

DNA helicases catalyze the unwinding of duplex DNA (dsDNA) 1 to provide the single-stranded DNA (ssDNA) intermediate that is required for diverse DNA metabolic processes such as DNA replication, recombination, transcription, and DNA repair (1). DNA helicases function as molecular motors and use the chemical energy derived from nucleoside 5Јtriphosphate binding or hydrolysis to mechanically disrupt the hydrogen bonds between the two strands in dsDNA and to translocate along DNA for processive unwinding (2).
The RecQ helicase family, which is widespread in diverse or-ganisms from Escherichia coli to humans, has become more interesting since the discovery that several human diseases, such as Bloom's, Werner's, and Rothmund-Thompson syndromes, are linked with these enzymes (3)(4)(5). Analyses of these and related diseases at the cellular and molecular levels led to the conclusion that RecQ helicase plays a conserved role in the preservation of genomic integrity (6 -8). An increasing amount of data has shown that the RecQ helicase family acts in concert with other proteins to perform functions in multiple processes in vivo. A number of proteins, including topoisomerases (9, 10), primase (11), polymerases (12,13), proliferating cell nuclear antigen (10), replication protein A (14), P53 protein (15), and FLAG endonuclease 1 (16), have been identified to interact physically and functionally with Werner and Bloom syndrome proteins and RecQ helicases. The prototypical member of the RecQ helicase family, the E. coli RecQ protein, plays an essential role both in DNA recombination and in suppression of illegitimate recombination (17)(18)(19). E. coli RecQ helicase (610 amino acids, molecular mass, 68,920 Da) displays a 3Ј-5Ј polarity in DNA unwinding and can unwind diverse DNA substrates including DNA with blunt ends, with 5Ј or 3Ј overhangs, nicked or forked DNA, and threeor four-way junctions as well as G4 DNA (a guanine-rich parallel four-stranded DNA structure) (20,21). Elucidation of the interactions between helicase and its DNA substrate is a very important step in understanding the mechanism by which helicases bind and unwind dsDNA. Currently little is known about the thermodynamic characteristics of the binding reaction between RecQ and DNA. Additionally, the effect of solution conditions, salt, and type of salt on the intrinsic affinities and the roles of nucleotide cofactors remain unclear. Thus, more knowledge concerning the properties and kinetic mechanism of RecQ helicase binding to DNA is needed to understand the molecular mechanism underlying helicase function. Toward this goal, we have investigated the DNA binding properties of E. coli RecQ helicase under equilibrium conditions using fluorescein-labeled DNA substrates. Free fluorescein-labeled oligonucleotide (ssDNA or dsDNA) displays low fluorescence anisotropy. Its anisotropy is enhanced upon protein binding due to the reduced rotational mobility of the formed protein-DNA complex. According to the Perrin equation (22) the anisotropy of a fluorescent complex increases with an increase of its volume. Thus, the change in anisotropy of fluorescently labeled oligonucleotides can be used to measure the extent of protein binding. This approach has been used successfully in a number of studies (23)(24)(25)(26).
In this paper we have used a steady-state fluorescence anisotropy assay to characterize the DNA binding properties of RecQ helicase from E. coli. Our results indicate that the binding site size of 10 nucleotides per monomeric enzyme does not change as oligonucleotide length increases, implying that RecQ helicase binds to DNA as a monomer. Interestingly, RecQ helicase appears to bind to DNA with high affinity and low cooperativity. In addition, only the nucleotide cofactor AMPPNP increases the affinity of the enzyme for ssDNA in a narrow concentration range, whereas ADP has no detectable effect. The formation of a helicase-DNA complex is accompanied by net release of ions. These properties offer insight into the mechanism by which RecQ helicase exercises its functions in cells.

EXPERIMENTAL PROCEDURES
Reagents and Buffers-All buffers were made with reagent grade chemicals using distilled water purified through a Milli-Q water purification system (Millipore Corp., France). ADP, ATP, and AMPPNP were purchased from Sigma. Standard titration buffer is 20 mM Tris-HCl, pH 7.4, at 25°C, 20 mM NaCl, 3 mM MgCl 2 , and 0.1 mM dithiothreitol.
pH Measurements-pH measurements were performed with a Knick 654 pH meter and an Ingold microelectrode. For calibration three standard buffers were used, and the precision was about Ϯ 0.01 pH unit at 25°C. To obtain a pH range from 5 to 10, 30 mM sodium citrate, MES, and Tris acetate were employed for pH intervals 5.0 -6.0, 5.5-7.2, and 6.8 -10, respectively. The possible effect of pH on anisotropy of DNA substrate was taken into account.
Agarose Gel Mobility Shift Assay-RecQ helicase-DNA complexes were formed in standard titration buffer in a total volume of 35 l at room temperature. The EcoRI-restricted 3-kilobase plasmid DNA (60 M, in nucleotides) was mixed with various concentrations of protein in the range of 0 -20 g. After incubation for 30 min, 5 l of loading buffer (40 mM Tris acetate, pH 7.5, 50% glycerol, and 0.25% (w/v) bromphenol blue) was added to each sample. The complexes were separated by electrophoresis through 0.8% agarose gel in TAE buffer (40 mM Tris acetate, 1 mM EDTA, pH 8.3) at 100 V for 1.5 h and were visualized by ethidium bromide staining.
RecQ Helicase-His 6 -tagged E. coli RecQ helicase was expressed from pET 15b expression plasmid in E. coli strain BL21 (DE3) as described (27). Briefly, the overexpressed protein was purified under native conditions by chromatography on Ni 2ϩ -nitrilotriacetic acid columns (Qiagen) followed by fast protein liquid chromatography size exclusion chromatography columns (Superdex 200, Amersham Biosciences) and an ion-exchange chromatography (DEAE Sephadex A-50). The His tag was cleaved using biotinylated thrombin, and removal of the biotinylated thrombin was accomplished using streptavidin-agarose magnetic beads (Novagen, Madison, WI). Based on Sypro Orangestained SDS-PAGE and electrospray mass spectrometry analyses, the purity of the RecQ preparation was Ͼ95%.
Oligonucleotides-The PAGE-purified unlabeled and fluorescein-labeled synthetic oligonucleotides were purchased from Proligo (France). Their structures, sequences, and extinction coefficients are shown in Table I. The concentrations of oligonucleotides were determined by measuring the extinction coefficients. Double-stranded oligonucleotides were made in a 20 mM Tris-HCl buffer, pH 7.2, containing 100 mM NaCl. The mixture was heated to 85°C for 5 min, and annealing was allowed by slow cooling to room temperature.
Anisotropy Assays-The binding assay was performed by using a Beacon 2000 polarization instrument (PanVera Corp.) (28). An appropriate quantity of fluorescein-labeled ssDNA or dsDNA was added to a standard titration buffer (150 l total volume) in a temperature-controlled cuvette at 25°C. The anisotropy of the fluorescein-labeled DNA was measured successively until it stabilized. Then an appropriate quantity of RecQ helicase was added. After the addition of each titrant the anisotropy was measured continuously until it reached a stable plateau. To determine the concentration of the helicase-DNA complex, fluorescence signals observed in these RecQ helicase titrations were subtracted from that observed in the absence of enzyme. The increase in sample volume during the titration was taken into account in the analysis of the data. The reported values are from the average of two to three measurements.
Determination of the Binding Site Size-The binding site size was determined from the intersection of lines drawn through the ascending and plateau regions of the anisotropy data by visual inspection. Stoichiometry (N) and dissociation constant (K D , in units of molarity) parameters were further determined by globally fitting the anisotropy data according to Equation 1 (29,30), where D T and P T are the total molar concentrations of DNA and helicase protein used, respectively, and where A is the fluorescence anisotropy at a given concentration of helicase protein, A max is the anisotropy at saturation, and A min the initial anisotropy. Determination of Rigorous Thermodynamic Binding Isotherms and Calculation of Binding Parameters-In this work we followed the binding of RecQ helicase to ssDNA or dsDNA by monitoring the anisotropy change upon complex formation. The method used to obtain rigorous estimates of the average binding density v (number of protein molecules bound per nucleotide) and free protein concentration L F has been previously described in the literature (31,32).
where represents the fractional saturation, n is the binding site size, M is the DNA lattice length, k is the average number of ligands binding to the DNA substrate, g is the maximal number of k, and L T is the total ligand concentration (M). The intrinsic binding constant (K, in units of M Ϫ1 ) and the binding cooperativity parameter (, unit-less) were deter-  where ϭ KL T , and the value of n used is 10 for RecQ (see "Results").

Validity of Fluorescence Anisotropy Measurements to
Determine the Binding Isotherms-We used two separate means to show that the increase in anisotropy is directly proportional to binding. First, we directly analyzed the relationship between the fractional saturation and the fluorescence anisotropy during the titration. Fig. 1A shows that, at a given concentration of fluorescein-labeled 21-mer oligonucleotide, the observed fluorescence anisotropy increases during the protein titration. If we assume that at saturation the moles of RecQ helicase added are equal to the moles of fluorescein-labeled DNA and all fluorescein-labeled DNAs are bound without precipitation of the RecQ-DNA complexes, then the fractional saturation of RecQ helicase can be derived from the anisotropy determined at each RecQ helicase concentration during titration. If the fluorescence anisotropy signal change really reflects the degree of saturation of the ligand, a linear relationship should exist between the fluorescence anisotropy and the fractional saturation of the ligand. As expected, Fig. 1B (circle) shows that the observed anisotropy increases linearly with an increase of the fractional saturation of RecQ helicase.
Secondly, to obtain thermodynamically rigorous binding parameters independent of any assumption about the relationship between the observed signal and the degree of binding, we analyzed the measured fluorescence anisotropy titration curves using the macromolecular binding density function method of Lohman and Bujalowski (31). At any given particular anisotropy value in Fig. 1A all titrations have the same binding density v ϭ L B /M T , where L B is the bound ligand concentration (moles of RecQ) and M T the total macromolecule concentration (moles of nucleotides). The relation between the total ligand concentration and free ligand concentration (L F ) should obey the law of mass conservation,  Fig. 1B (square) shows the linear dependence of the observed anisotropy on the fractional saturation of RecQ helicase. The value of could be determined experimentally up to 0.85, and its maximum value was derived from the extrapolation of the anisotropy data to its maximum (when A max ϭ 0.23, ϭ 1). Both of the above analyses clearly show that the fluorescence anisotropy assay is a reliable method for binding studies.
DNA Binding Site Size and Effect of DNA Length on Helicase Binding-An agarose gel mobility shift assay was used to determine the binding site size. Under our experimental conditions a shift of the 3-kilobase DNA was observed, and its magnitude was dependent on the RecQ helicase concentration (Fig. 2, inset). This shift was stabilized when the stoichiometric ratio between DNA and RecQ helicase corresponded to 10 -12 nucleotides of DNA per RecQ helicase monomer (Fig. 2). This indicates that about 11 nucleotides were occluded by a RecQ helicase monomer.
The precise value of the binding site size is important for determining the affinity (K) and cooperativity () parameters . Circles, 1 nM; squares, 10 nM; rhombuses, 30 nM. The lines represent simulated theoretical isotherms based on the method of Epstein (33) using the following values: n ϭ 10, K ϭ 6.7 Ϯ 0.95 ϫ10 6 M Ϫ1 , ϭ 25.5 Ϯ 1.2, and A max ϭ 0.23. B, steady-state fluorescence anisotropy versus the fractional saturation of RecQ. The degree of saturation was estimated directly from the plateau region of the titration curve (circles). At saturation, the moles of RecQ binding to DNA are equal to the total moles of fluorescein-labeled DNA, assuming that no protein-DNA complexes precipitations were formed. The moles of RecQ helicase bound with DNA at a given total RecQ concentration were calculated from the corresponding anisotropy value. Alternatively, the fractional saturation at a given anisotropy was determined using the method of macromolecular binding density function analysis (squares, see text).
for interactions between helicase and DNA. We, therefore, further determined the stoichiometry value of RecQ helicase binding to an oligonucleotide using the fluorescence anisotropy method under equilibrium binding conditions. We have previously determined the apparent dissociation constants for oligonucleotides of lengths of 13-, 21-, 36-, and 45-mer. Thus, the stoichiometry for DNA binding can be obtained under conditions where the oligonucleotide concentration is kept much higher than the dissociation constant (29,30). Fig. 3 shows the results of fluorescence anisotropy titrations of fluorescein-labeled 21-and 45-mer oligonucleotides with increasing RecQ helicase concentrations. In both cases fluorescence anisotropy increases sharply and then saturates. Binding stoichiometry was estimated empirically from the intersection of asymptotes as shown in Figs. 3, A and B. In the case of fluorescein-labeled 21-mer DNA the anisotropy increases from 0.055 to a plateau value of 0.18, demonstrating that RecQ helicase binds to DNA with 2:1 stoichiometry (Table II) and, thus, suggesting that the binding site size of RecQ helicase is 10 nucleotides. This conclusion was further supported by results obtained from global fits of the data in Fig. 3 to Equation 1.
We next titrated fluorescein-labeled oligonucleotides of lengths of 13-, 36-and 45-mer with increasing concentrations of RecQ helicase. The stoichiometry values determined by both direct inspection and global regression fits are summarized in Table II. We noted that there is a clear linear relation between the stoichiometry and the length of ssDNA used in the determination (Fig. 3C). The slope value of 11.4 Ϯ 1.5 nucleotides indicates that the DNA binding site size is less than 11 nucleotides per enzyme. Additionally, when fluorescein-labeled 13and 36-mer ssDNA was titrated with RecQ helicase in the presence of 1 or 2 mM AMPPNP or ADP, the stoichiometry values are essentially the same as in the absence of these nucleotide cofactors (Table II).
To further determine the number of nucleotides directly engaged in interactions with the DNA binding site of RecQ helicase, a series of ssDNA of different lengths from 7-to 45-mer were examined. Consistent with the above observation, oligonucleotides of more than 10 nucleotides in length bind tightly to the enzyme at saturating concentrations of RecQ helicase (500 nM) (Fig. 4A). However, the binding amplitude of the 7-mer oligonucleotide is lower compared with that of the 21mer oligonucleotide. In addition, the 7-mer oligonucleotidehelicase complex is unstable and dissociates gradually as a function of time. However, the complex can be stabilized in the presence of 1.5 mM AMPPNP (Fig. 4B), indicating a possible allosteric regulation of the enzyme by the nucleotide cofactor. Therefore, we conclude that the number of nucleotides directly engaged in interactions with RecQ helicase should be 8 Յ n Յ 11 bases.
Determination of Association Constants and Cooperativity-To determine the association constant (K) and cooperativity parameter () we constructed a model-free binding isotherm of versus L F (Fig. 5A) from the anisotropy-based binding isotherms shown in Fig. 1A (31). The binding isotherms were analyzed by the Epstein Equation, which describes more precisely the binding properties of a large ligand to a finite lattice of DNA with an arbitrary length (33). We fit the isotherms under the conditions that n ϭ 10, DNA length is 21 nucleotides, and RecQ helicase binds to DNA as a monomer, as we have shown previously (27). The best fit results give an association constant K ϭ 6.7 Ϯ 0.95 ϫ10 6 M Ϫ1 and a cooperativity parameter ϭ 25.5 Ϯ 1.2. These parameters were used to fit the binding isotherms of Fig. 1A. Similarly, the association constants and cooperativity parameters for RecQ helicases binding to fluorescein-labeled 36-and 45-mer oligonucleotides were  (Table II) of the data. C shows the linear relationship between the stoichiometry and the oligonucleotide length.
where R s is the migration distance of the complex in the presence of RecQ protein, R 0 in the absence of the protein, and R u the upper limit of the migration distance. determined in the same way. The results are summarized in Table III. These data clearly indicate that the association constant and the cooperativity of RecQ helicase vary from 6.7 Ϯ 0.95 ϫ10 6 to 7.5 Ϯ 2.1 ϫ 10 6 M Ϫ1 and from 25 to 38, respectively.
We have previously determined that the DNA binding site of RecQ helicase encompasses only 10 nucleotide residues and that a 13-mer DNA binds with high affinity to RecQ helicase. Thus, the study of RecQ helicase binding to a 13-mer oligonucleotide will provide the information about helicase binding to DNA without any interference from protein-protein interactions. The model-free binding isotherm obtained with 13-mer oligonucleotide is fit best with Epstein's non-cooperative binding equation with an association constant K ϭ 6.4 Ϯ 1.2 ϫ 10 6 M Ϫ1 (Fig. 5B). This result is expected because only 1 protein monomer can bind to the 13-mer oligonucleotide, and therefore, there are no protein-protein interactions.
Effect of Salt and pH on RecQ Binding to DNA- Fig. 6A demonstrates that the formation of the helicase-DNA complex is very sensitive to the concentration of sodium chloride; its formation was inhibited by sodium chloride concentrations exceeding 80 mM. About 50% of DNA-protein complexes were formed in the presence of 65 mM NaCl. However, sodium chloride was necessary for helicase binding to DNA, since in sodium chloride-free binding buffer no RecQ-DNA complex was ob-served. In fact, RecQ helicase aggregates and forms a tubular structure under this condition. Because the extent of complex disruption by high concentrations of NaCl reflects the affinity of a protein for DNA, we next examined the stability of the complex at high concentrations of NaCl (Fig. 6A, inset). We found that the helicase-DNA complex could not be disrupted to the same extent by the same NaCl concentration that was needed to inhibit the binding. The midpoint of salt titration for dissociation of the RecQ helicase by NaCl (130 mM) is 2-fold higher than that for inhibition by NaCl.
We next investigated the pH dependence of RecQ binding to DNA. Fig. 6B shows that RecQ helicase displays an optimum  DNA binding activity between pH 6.2 and 8.0. Similar results were obtained with dsDNA or DNA of different lengths (data not shown).
Effect of Salt Concentration on K and -The salt dependencies of the binding parameters K and were further determined quantitatively. Fig. 7A shows the binding isotherms obtained with the 21-mer oligonucleotide in the presence of different concentrations of NaCl. The binding parameters were obtained by theoretical fits of the binding isotherms. A log-log plot of the binding constant versus the salt concentration is   linear and characterized by a slope of Ѩlog K/Ѩlog[NaCl] ϭ Ϫ3.7 Ϯ 0.2 (Fig. 7B). The value of the slope indicates that there is a net release of 3.7 ions upon the complex formation. Thus, electrostatic interactions between the oligonucleotide and the protein monomer may contribute to the stability of the protein-DNA complex.
We next studied the salt dependence of 13-, 36-, and 45-mer DNA binding to the RecQ helicase. The association constant and cooperativity parameters were calculated similarly as reported above. For each case a plot of log K versus log[NaCl] was linear, and the number of ions released upon DNA binding was determined. An interesting phenomenon was observed; the number of ions released increases linearly as the length of ssDNA increases (Fig. 7C). Extrapolation of the plot toward zero oligomer length gives a value of 1.5 that is independent of the oligomer length.
Single-stranded DNA and Duplex DNA Bind Competitively to the Same DNA Binding Site on RecQ Helicase-To determine whether ssDNA and dsDNA bind to the same DNA-binding site we performed competitive binding studies. The protein-DNA complex was titrated with ssDNA or dsDNA under standard titration conditions. It is reasonable to suppose that if the helicase possesses the same DNA binding site for ssDNA and dsDNA, then fluorescein-labeled ssDNA (or dsDNA) and nonfluorescein-labeled dsDNA (or ssDNA) should competitively bind to the helicase. Therefore, the competitive experiments were performed with preformed RecQ-fluorescein-labeled ssDNA (or dsDNA) complex, and the complex was then titrated with non-fluorescein-labeled dsDNA (or ssDNA). The competitive binding of the non-fluorescein-labeled dsDNA (or ssDNA) over fluorescein-labeled ssDNA (or dsDNA) was monitored by measuring the change in the fluorescent anisotropy. As shown in Fig. 8, we observed that dsDNA binding could displace ssDNA and ssDNA binding completely displaced dsDNA, indicating that ssDNA and dsDNA bound to the same site on RecQ helicase. It is worth noting that the concentration of dsDNA needed to displace ssDNA to a given extent is higher than that of ssDNA needed to displace dsDNA to the same extent. This suggests that RecQ helicase possesses a higher affinity for ssDNA than dsDNA. The results shown in Table IV further support this conclusion and are consistent with previous observations (34).
Effect of Mg 2ϩ on the DNA Binding of RecQ Helicase-It is well known that Mg 2ϩ is required for ATP binding and hydrolysis. However, it is not clear whether Mg 2ϩ also contributes to DNA binding to RecQ helicase. It was reported that Mg 2ϩ ions greatly enhanced DNA binding of the Rad 51 protein, which promotes strand exchange between circular ssDNA and linear dsDNA (35). We, therefore, examined the effect of Mg 2ϩ on DNA binding of RecQ helicase. A series of titrations of 21-mer ssDNA with RecQ helicase in the absence or in the presence of MgCl 2 at different concentrations is shown in Fig. 9A, in which the solid lines are the computer fits according to the Epstein equation. Fig. 9B shows the dependence of the 21-mer DNA intrinsic binding constant on [MgCl 2 ] (log-log plot). The plot is linear, and the slope is Ѩlog K/Ѩlog[MgCl 2 ] ϭ Ϫ0.34, the magnitude of which is significantly lower than 3.7 obtained in the presence of NaCl alone. Because no great difference was observed between the association constants determined in the absence and presence of Mg 2ϩ ions, we conclude that Mg 2ϩ has only a slight effect on DNA binding of RecQ helicase.
Effect of the Nucleotide Cofactor on the Interactions of RecQ Helicase with DNA-We had previously observed that the binding of the 7-mer oligonucleotide to RecQ helicase was enhanced by AMPPNP (Fig. 4B). The allosteric effects were further studied here. Fig. 10A shows the fluorescence anisotropy titrations of fluorescein-labeled 21-mer ssDNA with RecQ helicase in the presence of 1.5 mM AMPPNP, 1.5 mM ADP and in the absence of nucleotide cofactor. The solid lines are computer fits using the Epstein Equation (33). The obtained intrinsic binding constants are 11.2 Ϯ 0.89 ϫ10 6 , 6.5 Ϯ 1.2 ϫ 10 6 , and 6.8 Ϯ 1.1 ϫ 10 6 M Ϫ1 for binding in the presence of AMPPNP, ADP, and in the absence of nucleotide cofactors, respectively. However, when the same experiments were performed with fluoresceinlabeled dsDNA, no significant differences were observed between the intrinsic binding constants in the presence and in the absence of nucleotide cofactors (Table IV). In addition, we systematically investigated the effect of nucleotide cofactor concentration on the intrinsic binding constant obtained with ssDNA and dsDNA. Fig. 10B shows that the allosteric effect of AMPPNP was only observed when RecQ helicase binds to ssDNA in the presence of 1.5 mM AMPPNP. No detectable allosteric effect was observed with dsDNA in the presence of low concentrations of nucleotide cofactors (below 1.5 mM). Some modest inhibition effects were observed in the presence of high concentrations of nucleotide cofactors (above 1.5 mM).

DISCUSSION
In this report we have characterized the DNA binding properties of E. coli RecQ helicase, a typical member of the RecQ helicase family, using the steady-state fluorescence anisotropy assay under equilibrium binding conditions. The DNA binding assay that we used in this study is rapid and convenient since physical separation of free and DNA-bound protein is not required. This assay is also a reliable method for surveying the binding behavior of RecQ helicase under these conditions, because both direct analysis of the relationship between the fluorescence anisotropy signal and fractional saturation and the macromolecular binding density function analysis show that the fluorescence anisotropy increases linearly with the fractional saturation. Based on this assay we have performed rigorous, quantitative studies of the interactions between RecQ helicase and oligonucleotides of different lengths. To our knowledge such a protein-DNA binding study has not been previously reported for any RecQ family helicase. The major results of our studies are as follows. 1) Under optimal conditions RecQ helicase has a binding site size of 10 nucleotides and binds DNA with an intrinsic binding constant K ϭ 6.7 Ϯ 0.95 ϫ 10 6 M Ϫ1 and low cooperativity ( ϭ 25.5 Ϯ 1.2). 2) RecQ helicase can bind to ssDNA or dsDNA at the same site. 3) The was suggested that an oligomer or at least a dimer is absolutely required for helicase translocation and unwinding activity (36 -38). However, several lines of evidences indicate that some  helicases function as a monomer (39 -41). The extensive structural and kinetic analysis of PcrA helicase show that the active form of this protein is a monomer (42,43). Using a wide array of biochemical and biophysical methods we have previously shown that RecQ helicase is a monomer in solution in the absence of DNA (27). Recently, the high resolution three-dimensional structure of the RecQ protein further confirmed its monomeric nature (44). In our previous study we did not address whether RecQ helicase could undergo dimerization or higher order oligomerization upon binding to DNA, as suggested for Rep helicases that form a homodimer upon binding to either ssDNA or dsDNA (45). Our direct stoichiometry studies shown in this report indicate that the RecQ helicase binds the nucleic acid as a monomer. This conclusion is supported by the following results. (i) The stoichiometry of 13-mer DNA-RecQ helicase complex is 1, indicating that RecQ helicase binds to DNA as a monomer not as a dimer or oligomer; (ii) the stoichiometry of 1:10 of RecQ helicase-nucleotides complex remains the same independent of the DNA length (Fig. 3C), suggesting that RecQ helicase binds to DNA as a monomer whatever the length of DNA. From these observations it is clear that RecQ helicase binds sequentially to the DNA substrate as a monomer in forming the protein-nucleic acid complex. It is worth noting that RecQ helicase remains a monomer with DNA even in the presence of nucleotide cofactors (Table II).
Low Cooperativity Displayed by RecQ Helicase Is Consistent with Its Structure and Functional Properties-The most striking result of this study is that RecQ helicase displays a low cooperativity with a value of ϳ25. First, such a behavior is expected from its quaternary structural nature. As it was mentioned above, we have previously shown that RecQ helicase is a monomer both in the absence and in the presence of DNA. Thus, no essential protein-protein interactions were observed under different conditions. Therefore, the low cooperativity exhibited by RecQ helicase is consistent with its monomeric nature. It is interesting to note that herpes simplex virus type I-coded ssDNA-binding protein, ICP8, identified as a monomer, also displayed a low cooperativity (24). Second, from the definition of the term of cooperativity, high cooperativity means that a protein will have a high affinity for long stretch ssDNA, and the binding of molecules adjacent to each other could be greatly enhanced. Consistent with this notion, most ssDNAbinding proteins bind DNA with a relatively high cooperativity over a wide range of values from 10 2 to 10 5 so that the protein can efficiently cover a relatively long stretch of ssDNA at low protein concentrations (46,47). In the case of the RecQ helicase family proteins, the proteins exercise their diverse functions through heterologous protein-protein interactions. It is well established that the RecQ family proteins interact physically and/or functionally with several proteins including topoisomerases, polymerase ␦, proliferating cell nuclear antigen, replication protein A, and FLAG endonuclease 1 (see the Introduction). High cooperativity would enhance the homologue protein-protein interactions and could probably inhibit the heterologous interactions with other proteins necessary for normal RecQ helicase functions. Thus, it is conceivable that low cooperativity is necessary for RecQ helicase to perform its functions in cells.
Effect of Nucleotide Cofactors and Mechanism of the Helicase Functioning-Nucleotide cofactors such as ATP not only provide energy for unwinding activity but also function as an allosteric effector to regulate helicase affinity toward DNA substrates (1, 2). It is well documented that a common feature of the well characterized helicases such as the E. coli Rep helicase, phage T7 and T4 helicases, DnaB and PriA proteins is that their affinity for ssDNA or dsDNA is modulated either by ATP or by its hydrolysis product, ADP (48 -52). The binding and release of the DNA substrate as a function of nucleotide binding and hydrolysis is a central concept for models proposed so far. For example, the rolling model was proposed based on the observations that ADP favored helicase binding to ssDNA and ATP favored simultaneous binding to ssDNA and dsDNA (36). We have compared RecQ with several helicases whose affinity for ssDNA and dsDNA were modulated by ATP and ADP (Table IV). Based on our studies, we find that RecQ helicase displays a special feature; its affinity for ssDNA is enhanced only by AMPPNP in a narrow concentration range of nucleotide cofactor, and ADP has no detectable effect on its affinity for either ssDNA or dsDNA. These results imply that the mechanism of RecQ helicase modulation by nucleotide cofactors may be different from other helicases. Is there a common feature for helicase binding activity modulation by nucleotide cofactors? As can be seen from Table IV, the same nucleotide cofactor may exercise different effects on different helicases. For example, AMPPNP may greatly increase the affinity of DnaB helicase for ssDNA, whereas that of Rep helicase is decreased to 35.3% by AMPPNP. Thus, it is not conceivable to give a general feature for effects of ATP or ADP on all helicases.