The Escherichia coli RecQ Helicase Functions as a Monomer*

The RecQ helicases belong to an important family of highly conserved DNA helicases that play a key role in chromosomal maintenance, and their defects have been shown to lead to several disorders and cancer in humans. In this work, the conformational and functional properties of the Escherichia coli RecQ helicase have been determined using a wide array of biochemical and biophysical techniques. The results obtained clearly indicate that E. coli RecQ helicase is monomeric in solution up to a concentration of 20 (cid:1) M and in a temperature range between 4 and 37 °C. Furthermore, these properties are not affected by the presence of ATP, which is strictly required for the unwinding and translocating activity of the protein, or by its nonhydrolyzable analogue 5 (cid:1) -adenylyl- (cid:2) , (cid:3) -imidodiphosphate. Consistent with the structural properties, functional analysis shows that both DNA unwinding activity and single-stranded DNA-stimulated ATPase specific activity were independent of RecQ concentration. The monomeric state was further confirmed by the ATPase-deficient mutants of RecQ protein. The rate of unwinding was unchanged when the wild type RecQ helicase

Genetic information is stored into the double-stranded DNA molecule that is stabilized through specific hydrogen-bonded base pairs. However, replication and repair as well as recombination of the DNA molecule require a single-stranded DNA to be available at least transiently. In cells, unwinding and separation of the complementary strands of duplex DNAs (dsDNA) 1 into single-stranded DNAs (ssDNA) are catalyzed by a class of enzymes known as helicases. DNA helicases destabilize and unwind the duplex DNA through a series of energetic states, driven by the binding and hydrolysis of NTP, usually ATP, and subsequent release of NDP and inorganic phosphate. Thus, DNA helicases convert the chemical energy into mechanical energy for DNA unwinding and translocation along the nucleic acid lattice (for reviews, see Refs. [1][2][3]. During unwinding of dsDNA, and to translocate processively without dissociation from DNA, the helicase must use at least two DNA-binding sites to keep contact with the DNA lattice; one binds to ssDNA for translocating, whereas the other binds to dsDNA for DNA unwinding. These two DNA binding sites may be located at different domains within a single polypeptide of a monomer or be held by two different polypeptides within a dimer or an oligomer for providing multiple DNA-binding sites. Corresponding to the first possibility, the "inchworm" model was proposed, in which the helicase is assumed to possess two nonidentical DNA binding sites (4); the "leading" site binds both ssDNA and dsDNA and interacts with the duplex to be unwound during successive unwinding cycles, whereas the "lagging" site interacts only with ssDNA. The disruption of the dsDNA at the leading site and the translocation of the enzyme are the result of conformational change of the enzyme modulated by binding and hydrolysis of ATP. Recent crystal structures of complexes of PcrA helicase with a partial dsDNA duplex substrate show that the ssDNA and dsDNA bind, respectively, on two domains of this monomeric helicase (5). These data provided direct proof to support an inchworm mechanism (6).
Corresponding to the second possibility, the active "rolling" model requires that the enzyme be oligomeric and at least dimeric. Each protomer possesses an identical DNA binding site. Both sites could bind either ssDNA or dsDNA, and binding of ssDNA and dsDNA cannot occur simultaneously in the same subunit. Binding of ATP leads to the enzyme interacting alternatively with the ssDNA and dsDNA at the junction region. Furthermore, hydrolysis of ATP destabilizes hydrogen bonds between the base pairs of the duplex. This model was originally based on the observed allosteric effects of ATP and ADP on the ssDNA and dsDNA binding properties of the Escherichia coli Rep dimer (7). However, the crystal structures of Rep helicase bound to ssDNA alone or bound to both ssDNA and ADP have revealed that the protein remained monomeric; no proteinprotein interactions were observed (8).
An essential difference between the "inchworm" and "rolling" model is that an oligomer, at least a dimer, is absolutely required for translocation and unwinding in the case of the rolling model, whereas a monomeric form or any oligomeric form could function in the inchworm model. Thus, the knowledge of the oligomeric structure of a helicase is of fundamental concern in understanding the mechanism by which the protein unwinds DNA.
The objective of this work is to investigate the structural state of the E. coli RecQ helicase in solution, a protein of 610 amino acids, which is highly conserved across a wide variety of organisms including Saccharomyces cerevisiae Sgs1 (9), Schizosaccharomyces pombe Rqh1 (10), and Homo sapiens RecQL (11), Bloom syndrome protein (12), and Werner syndrome protein (13). Members of this family play a key role in the maintenance of chromosomal stability, and their defect in humans leads to several disorders including Bloom syndrome (cancer predisposition) and Werner syndrome (premature aging condition) (14). The protein, originally identified in E. coli via a mutant resistance to thymine starvation, is implicated in the RecF and RecE recombination pathway (15) and is characterized by a central domain that contains seven so-called helicase signature motifs, including a putative ATP binding sequence in motif I and a DEXH box in motif II. The protein displays a 3Ј-5Ј polarity in DNA unwinding, since it has been observed using nonsymmetrical DNA substrates (16), and has a broad range of DNA substrates such as duplex DNA with blunt ends or with 5Ј or 3Ј overhangs, nicked or forked DNA, and three-or four-way junctions, as well as G4 DNA (17)(18)(19).
In this report, both structural and functional analyses show that the RecQ helicase functions as a monomer. These results should be helpful in understanding the mechanism of action of E. coli RecQ helicase and possibly of other members of the RecQ helicase family.

EXPERIMENTAL PROCEDURES
Reagents-Chemicals were reagent grade, and all solutions were prepared using ELGApure water. ATP and AMPPNP were purchased from Sigma. Buffer A is 25 mM Tris-HCl, pH 7.5, 50 mM NaCl, 3 mM MgCl 2 , 0.1 mM dithiothreitol.
Nucleic Acid Substrates-Oligonucleotides were purchased from Genset and further purified by polyacrylamide gel electrophoresis. The sequences of the oligonucleotides used in this study are listed in Table  I. A 100 M working stock of double-stranded DNA was prepared by mixing equal concentrations of each oligonucleotide in 10 mM Tris-HCl (pH 7.5 at 25°C), 100 mM NaCl, followed by heating the mix for 5 min at 85°C and then cooling it slowly to room temperature.
Protein Expression-A full coding region of RecQ DNA was prepared by PCR using E. coli chromosome DNA. A 2.2-kb PCR product was cloned into pGEM®-T easy vector, and the sequence of this insert was shown to be identical to the published RecQ gene sequence (20). The DNA corresponding to the coding sequence of the RecQ helicase gene was amplified again by PCR-mediated amplification, using a set of oligonucleotides corresponding to the 5Ј-and 3Ј-ends of the gene. These primers had additional bases at each end for creation of restriction sites: NdeI at the 5Ј-end (5Ј-CAT ATG AAT GTG GCG CAG GCG GAA GTG TTG-3Ј) and XhoI at the 3Ј-end of the gene (5Ј-GCG CAT GTT GAT GGC GAT GAC GAA GAG TAG CTC GAG-3Ј) for expression of RecQ helicase bearing a hexahistidine tag at the NH 2 -terminal end or c-myc tag sequence at the 3Ј-end of the gene (5Ј-GCG CAT GTT GAT GGC GAT GAC GAA CAA AAA CTC ATC TCA GAA GAG GAT CTG TAG CTC GAG CGG-3Ј) for expression of RecQ helicase with c-myc epitope at the amino terminus. The amplified product was digested with NdeI and XhoI enzymes, and the appropriate size fragment (2100 bp) was gelpurified and subcloned, respectively, into NdeI/XhoI sites of the vector PET-15b and pNB5 (21). The PET-15b vector encodes a hexahistidine tag at the amino terminus that allows purification of the expressed protein on a nickel-chelating column. A thrombin cleavage site was adjacent to the histidine tag. The constructed plasmid was expressed in E. coli strain BL 21(DE3). The bacteria were grown at 37°C in terrific broth supplemented with 50 g/ml ampicillin. The expression of the protein was induced by adding isopropylthio-␤-D-galactoside to 0.5 mM at low log phase (A 600 ϭ 0.6). The culture was then incubated for 3 h at 37°C. Cells were harvested by centrifugation at 3,000 ϫ g for 15 min at 4°C. For expression of the RecQ helicase bearing a c-myc tag at the carboxyl terminus, but without the histidine tag at the amino terminus, the pNB5 vector containing the RecQ gene was transformed into EK 1104. The protein expression and purification were essentially as described by Benaroudj et al. (21).
Protein Purification-His-tagged RecQ helicase was overexpressed in E. coli BL21 (DE3) and purified under native conditions. Briefly, harvested cells were suspended in 30 ml of suspension buffer (20 mM Tris-HCl, pH 7.9, 0.5 mM imidazole, 500 mM NaCl) and were lysed using a French pressure cell. The lysate was then sonicated in order to shear DNA into small fragments. The lysate was cleared by centrifugation at 70,000 ϫ g for 30 min at 4°C. The supernatant was applied to the column charged with histidine binding resin (Novagen). The column was washed with 20 mM Tris-HCl (pH 7.9) buffer containing 500 mM NaCl, 60 mM imidazole. The proteins bound to the column were eluted stepwise using 20 mM Tris-HCl (pH 7.9) buffer containing 100, 200, 300, 400, or 500 mM imidazole. RecQ helicase-containing fractions, identified by both DNA-dependent ATP hydrolysis and helicase activity assays, were pooled. The histidine tag was cleaved using biotinylated thrombin during a dialysis step. The removal of biotinylated thrombin was accomplished using streptavidin-agarose magnetic beads (Novagen, Madison, WI). RecQ helicase was further purified by FPLC size exclusion chromatography (Superdex 200; Amersham Biosciences). Finally, ion exchange chromatography (DEAE Sephadex A-50) was used to remove the contaminating DNA. The active fractions were pooled, dialyzed against storage buffer (25 mM Tris-HCl, pH 7.5, 3 mM MgCl 2 , 500 mM NaCl, 2 mM dithiothreitol), and stored at Ϫ80°C. The protein was pure as judged by Coomassie staining and electrospray mass spectrometry. Protein concentration was determined spectrophotometrically using an extinction coefficient at 280 nm of 1.54ϫ10 4 M Ϫ1 cm Ϫ1 .
Mutagenesis-Site-directed mutagenesis was performed using the QuikChange TM kit (Stratagene) according to the manufacturer's instructions. Amino acid substitutions were made in conserved motif I (K55A) and in conserved motif II (D148A) using oligonucleotides C and D ( Table I). The mutations were confirmed by DNA sequencing (MWG Biotech). The mutant helicases were expressed in E. coli BL21 (DE3) cells and purified as described above for wild type enzyme.
Immunoprecipitation-Anti-His tag monoclonal antibodies were immobilized onto CNBr-activated Sepharose 4B at 2.5 mg of protein/ml of packed beads (22,23). For homodimerization studies, 0.2 ml of the mix of 5 or 10 M His-tagged and c-myc-tagged helicase was incubated at 4°C in buffer A for 1 h. The samples were mixed with 150 l of anti-His tag antibody beads. The beads were then washed twice with 5 ml of 10 mM Hepes, pH 7.4. The bound proteins were eluted from the beads by the addition of 180 l of 0.1 M acetic acid containing 0.2% Triton X-100.
SDS-PAGE and Western Blotting-SDS-polyacrylamide gel electrophoresis was carried out on 10% polyacrylamide gels. For Western blotting analysis, 23 l of the samples from immunoprecipitation experiments were analyzed by SDS-PAGE. The gels were electroblotted onto Immobilon membranes. For immunoblotting analysis, the blots were blocked in phosphate-buffered saline, containing 0.05% Tween 20 and 5% nonfat milk, and incubated with anti-c-myc monoclonal antibodies (dilution 1:1000) for 45 min. The blots were then washed in phosphate buffer (0.05% Tween 20) and further incubated in blocking buffer containing 1:1000 anti-mouse immunoglobulin peroxidase for 30 min. The blots were washed again in phosphate buffer (0.05% Tween 20) and visualized using the ECL Western blotting detection kit.
ATPase and Helicase Assay-The ATPase activity was measured according to the literature (24). One unit of ATPase activity is defined as the amount of RecQ protein required to hydrolyze 1 nmol of ATP/min at 37°C (16). Helicase activity was determined by Xu et al. (25) using DNA substrate B (Table I). Helicase-DNA Binding Activity Assay-Binding of RecQ helicase to DNA was analyzed by fluorescence polarization using a Beacon fluorescence polarization spectrophotometer (PanVera). RecQ helicase (200 nM) was added to a 150-l aliquot of buffer A containing 1 nM of 21-mer fluorescein-labeled DNA (substrate A). Each sample was incubated for 5 min at 25°C, after which fluorescence polarization was measured. In order to ensure that the mixture had reached equilibrium, the sample was further incubated for 30 min, and a second reading then was taken. No significant change was observed between the two measurements, indicating that equilibrium was reached.
Size Exclusion Chromatography-Size exclusion chromatography was done at 25°C, using an FPLC system (Amersham Biosciences), on a Superdex 200 (analytical grade) column equilibrated with buffer A. Elution was performed using the same buffer. RecQ helicase was also analyzed using the buffer containing 300 mM NaCl. Fractions of 0.5 ml were collected at a flow rate of 0.4 ml/min, and absorbance was measured at 280 and 260 nm. For ATP and AMPPNP experiments, 1 mM ATP or AMPPNP was present in all buffers. It is well known that proteins migrate through a gel filtration column as a function of both the molecular weight and the molecular shape. The R S values (R S designates the Stoke radius of the protein) of the RecQ molecule in different conditions were determined from the plot of log R S versus K av using the different Stokes radii of the standards. The partition coefficient K av was calculated using the formula where V e is the elution volume of the sample, V 0 is the excluded volume of the column, and V t is the total volume of the column. The excluded volume, V 0 (7.52 ml), and the total volume, V t (23.5 ml) were measured by calibration with dextran blue and thymidine. The calibration graph of log R S versus K av was constructed using a high and low molecular weight calibration kit from Sigma (26): cytochrome c (12.4 kDa; R S ϭ 12 Å), carbonic anhydrase (29 kDa; R S ϭ 22.5 Å), albumin (67 kDa; R S ϭ 35.5 Å), phosphorylase b (97.4 kDa; R S ϭ 38.75 Å), and thyroglobulin (669 kDa; R S ϭ 85 Å). Assuming similar shape factors, the plot calibration of log M r versus K av allowed the determination in a first approximation of the molecular weight of helicase.
Sedimentation Velocity-Sedimentation velocity experiments were performed at 4°C with a Beckman Optima XL-A analytical ultracentrifuge equipped with an An Ti 60 titanium four-hole rotor with twochannel 12-mm path length centerpieces. Sample volumes of 400 l in the buffer A were centrifuged at 55,000 rpm, and radial scans of absorbance ( ϭ 280 nm) were taken at 10-min intervals. The experimental base lines were measured for each sample at the end of the run. RecQ protein remained stable over the course of the run as shown by both the ATPase and the helicase activity assays. Data analysis and determination of the hydrodynamic parameters (the diffusion constant D and the sedimentation coefficient s) were performed using the computer programs supplied by Beckman, SEDNTERP and SVEDBERG supplied by John Philo. The partial specific volume v, 0.7286 ml/g, was calculated from the amino acid composition. Molecular mass is related to D and S by the equation, where is the solvent density, and R and T are the molar gas constant and the absolute temperature, respectively. Sedimentation Equilibrium-Sedimentation equilibrium experiments were carried out as described by Bailey et al (27). at 4°C using different rotor speeds (from 4,500 to 15,000 rpm). Enzyme samples were dialyzed extensively against buffer A. During the dialysis, the buffer was changed three times. The reference cells were filled by the buffer used for the last dialysis. Radial scans of the absorbance at 280 nm were taken at 2-h intervals, and samples were judged to be at equilibrium by the absence of systematic deviation in overlaid successive scans and when a constant average M r was obtained in the plot of M r versus centrifugation time. RecQ protein remained stable over the course of the run as judged by both the ATPase and the helicase activity assays.
The experimental data were fitted to a model for a single homogeneous species following the equation, where ϭ W 2 /2RT and W ϭ M(1 Ϫ ), A(r m ) is the absorbance of the solute at the meniscus, A(r) is the absorbance at a radial distance r from the center of rotation, and W is the buoyant molecular weight. M r and are the relative molecular weight and partial specific volume of solute, respectively, is the density of the solvent, is the angular velocity, R and T are the molar gas constant and the absolute temperature, respectively, and ␦ is the base line offset. The partial specific volume for the protein was calculated from the amino acid composition of the protein sample, and the densities of the different buffers were determined from published tables (28,29).
Time-resolved Fluorescence Measurements-The time-resolved fluorescence anisotropy was obtained from the two polarized emission decays Ivv(t) and Ivh(t), using the time-correlated single-photon counting technique. Ivv corresponds to the emission intensity (I) when both excitation and emission polarizers are vertical (v). Ivh corresponds to the emission intensity (I) when excitation polarizer is vertical (v) and emission polarizer is horizontal (h). The excitation light pulse source was a Ti-sapphire subpicosecond laser (Tsunami, Spectra Physics) associated with a third harmonic generator tuned at 299 nm for tryptophan fluorescence. The repetition of the laser was set down to 4 MHz, and to ensure the single-photon counting condition, the counting rate never exceeded 40 kHz. The excitation light pulse was triggered by a Hamamatsu photodiode (S4753). The fluorescence emission was detected through a monochromator (SpectraPro 150, ARC) set at the appropriate wavelength (⌬ ϭ 15 nm) (emission wavelength was 350 nm). A time-correlated single photon-counting SPC-430 card (Becker-Hickl GmbH) was used for the acquisition. The function of the instrumental response of the laser pulse (100 ps) was recorded by detecting the light scattered by a water solution. The time scaling was 11 ps/ channel, and 4096 channels were used. The two polarized components of the emission decay were collected until the total count of the Ivv component reached 15-20 million for tryptophan fluorescence or 25-35 million for fluorescein fluorescence. The microcell (volume 50 l) was thermostated with a Haake type-F3 circulating bath. Unless otherwise specified, the Trp fluorescence measurement mixture included 20 mM Tris-HCl, pH 7.4, 3 mM MgCl 2 , 0.1 mM dithiothreitol, and helicase RecQ at the appropriate concentration in the presence of either 75 or 500 mM NaCl. The data were analyzed by the quantified maximum entropy method (30,31), which displays the lifetime () and the correlation time () distributions according to Equations 3 and 4, where ␣ i represents the relative population of fluorophores with lifetime i , and j represents the initial anisotropy of molecules characterized by the rotational correlation time j . The decay of the total fluorescence intensity, I T (t), is calculated from both polarized components by the equation, where G represents the correction factor for the difference in the monochromator transmission between parallel and perpendicular polarized components (G-factor ϭ Ivv/Ivh). The anisotropy (r) is then defined by the following relation.
For a spherical macromolecule, the hydrated molecular volume (V) is related to by the Perrin equation, where represents the viscosity, T the absolute temperature, and k is the Boltzman constant.

Purification and Characterization of the RecQ Helicase-The
RecQ helicase was overexpressed in and purified from E. coli. Gel electrophoresis of the protein in denaturing conditions gave a single band corresponding to a molecular mass of about 70 kDa, consistent with the value determined from the amino acid sequence (68,290 Da). The purity and the molecular mass of the protein were further confirmed by electrospray mass spectrometry, which gives a mass of 69,000 Ϯ 250, with purity exceeding 95%. The protein purified as above is active as an ATPase with a specific activity of 420 units/g of protein, which is comparable with the previous determinations (16). The protein also displays DNA unwinding activity with an apparent unwinding rate of 2 bases/s, consistent with previous results (19).
Analysis of RecQ Helicase by Immunoprecipitation-RecQ helicase was first studied qualitatively by immunoprecipitation analysis. We reasoned that if RecQ helicase functions as an oligomer, the protein-protein interactions in the oligomer could be revealed by immunoprecipitation analysis. For this purpose, two different recombinant proteins were prepared; one was His-tagged at the NH 2 -terminal end, and the other had a c-myc epitope at the COOH-terminal end. A complex between the purified His-tagged and c-myc-tagged helicases was allowed to form at different concentrations of salt or different temperatures (between 20 and 42°C) and in the absence or in the presence of ATP. Anti-His tag monoclonal antibodies were used to immunoprecipitate RecQ helicase in the complexes. The immunoprecipitated samples were then analyzed by Western blotting with antibodies against c-myc epitope for detection of the c-myc-tagged helicase. The immunoprecipitation of Histagged helicase by the anti-His tag monoclonal antibodies failed to co-precipitate the helicase having the c-myc epitope, even in the presence of ATP (results not shown). However, it is still possible that the purified helicases preformed a stable oligomer so tightly that no monomer exchange between the preformed oligomers can occur. We therefore co-expressed the two tagged helicases in the same E. coli cell, and then analyzed the cell extracts by immunoprecipitation analysis. No proteinprotein interaction could be detected (results not shown). These results, taken together, suggest that RecQ helicase is mainly monomeric both in vitro and in vivo conditions. Analysis of RecQ Helicase by Size Exclusion Chromatography-In order to further characterize the structure of the RecQ helicase, the protein free in solution was analyzed by size exclusion chromatography. These experiments were first performed in the presence of low concentrations of NaCl (50 mM), conditions in which both the optimum ATPase and DNA unwinding activities are observed. In these conditions, RecQ helicase elutes as a protein of about 57 kDa having a Stokes radius of 29 Å (Fig. 1 and line 1 in Table II). Since the molecular weight obtained was smaller than expected for a protein of 610 amino acid protein and molecular weight of 68,290 Da according to the amino acid sequence, proteolysis may have occurred during the experiment. This was ruled out by SDS-PAGE electrophoresis and electrospray mass spectrometry of the helicase recovered from size exclusion chromatography, which gave the expected molecular weight. One likely explanation for this discrepancy is that RecQ helicase may interact partially with the column material at low salt concentration, and in fact similar results have been reported by other groups for T4 Dda helicase on gel filtration columns (32,33). Thus, when the experiment was performed at high concentrations of NaCl (300 mM), RecQ helicase eluted as a single species corresponding to a molecular mass of about 67,000 Da and a Stokes radius of 34 Å (Fig. 1, inset; Table II, line 3). This value is in accordance with the one obtained (33.6 Å) using the following empirical relation: log(Rs) ϭ 0.369 log(M r ) Ϫ 0.254 (34). Similar molecular weight and Stoke radius were obtained at high protein concentrations (up to 20 M) (Table II, (Table II, lane 5). Therefore, the RecQ helicase is most likely a monomer in solution, whatever the salt and protein concentrations (up to 20 M), and in the absence or in the presence of ATP.
Analysis of RecQ Helicase by Analytical Ultracentrifugation-To obtain further information about the quaternary structure of RecQ helicase in solution, analytical ultracentrifugation experiments were performed. First, sedimentation velocity confirmed the monomeric nature of RecQ helicase as indicated by the presence of a single sharp boundary. Moreover, from 3.6 to 10 M, the sedimentation coefficient was rather constant (not shown), as expected for a nonassociative single particle. Extrapolating this data to infinite dilution gave a sedimentation coefficient, s°4 ,w of 3.42 S and, when corrected for temperature, an s°2 0,w of 5.28 S (see Table III). The diffusion coefficient was 6.89 cm 2 /s and allowed the calculation for both the Stokes radius (R s ϭ 31.21 Å) and a mass of 69,120 g/mol according to Equation 1 (see "Experimental Procedures"). Thus, the molecular weight obtained by sedimentation velocity was consistent with the one determined by size exclusion chromatography and corresponded to that of a monomer. The calculated frictional ratio f/f 0 of 1.1 suggests that the RecQ helicase monomer is rather globular (Table III).
Second, sedimentation equilibrium of RecQ helicase, at different rotor speeds and protein concentration, was performed. The data for the RecQ helicase fit well to a single species model that yielded a molecular mass of 69,420 Ϯ 380 Da (Table IV and Fig. 2), in good agreement with the molecular weight obtained by size exclusion chromatography and sedimentation velocity. This is further confirmed using the DISCREEQ program (35), which indicated that 100% of RecQ helicase was monomeric. These data are consistent with the above observations that RecQ helicase is monomeric. based on the depolarization of light that occurs during the rotational diffusion of macromolecules or biological complexes. The extent of light depolarization between excitation and emission times is then related to the molecular size of the macromolecule. The analysis of the TFA data by the maximum entropy method displays the distribution of rotational correlation times (), which are related to the hydrodynamics volumes (see Equation 7). Such a distribution is shown in Fig. 3. RecQ helicase was studied using intrinsic tryptophan fluorescence. RecQ helicase contains 4 tryptophan residues distributed along the sequence of the entire protein. The lifetime distribution as revealed by the maximum entropy method analysis ( 1 ϭ 0.37 Ϯ 0.04 ns (39%); 2 ϭ 1.6 Ϯ 0.1 ns (33%); 3 ϭ 4.05 Ϯ 0.2 ns (21%); 4 ϭ 7.35 Ϯ 0.25 ns (7%)) leads to an average lifetime of 2.05 ns. As shown in Fig. 3, a 1 M RecQ helicase preparation displayed a long correlation time of about 50 ns at 20°C, consistent with a low order of oligomerization, mainly a monomeric form (see "Discussion"). Another peak was detected at about 4.9 ns. This peak corresponds most likely to a domain motion that is independent of the global tumbling of the entire protein. The flexibility motions of Trp were characterized by a correlation time distribution below 1 ns (not shown). Between 0.5 and 20 M, the RecQ helicase displayed a similar long rotational correlation time, indicating that its oligomeric state is not strongly dependent upon the protein concentration in this range (Table V). Since most of the ultracentrifugation data were obtained at 4°C, the TFA experiments were repeated at different temperatures (Table VI). From 4 to 37°C, the long correlation time decreased from 90 Ϯ 13 to 38 Ϯ 5 ns. To take into account the modulation of correlation times by external factors such as the temperature and viscosity, all of the data were normalized at 20°C according to the Perrin equation (Equation 7). As indicated in Table VI, the normalized rotational correlation times did not change significantly as a function of temperature. a The distribution coefficient K av was measured from the retention time on the column and was used to determine both the apparent molecular weight and the Stokes radius from the calibration curves (see "Experimental Procedures").
b The experiment was performed with elution buffer containing 300 mM NaCl.

TABLE III Hydrodynamic parameters of RecQ helicase
The molecular weight is obtained by the Svedberg equation. The hydrodynamic parameters were calculated using the partial specific volume value determined from the amino acid composition. s°n ,w is the corrected sedimentation coefficient at a given temperature in water. f and f o are the experimental and theoretical frictional coefficients, respectively. D o 20,w is the corrected diffusion coefficient at 20°C in water.   Therefore, this result shows that RecQ helicase is mainly monomeric in the 4 -37°C temperature range.
ATPase Activity and DNA Unwinding Activities Assay Versus RecQ Helicase Concentration-If RecQ helicase functions as a monomer, then the ATPase and helicase specific activities of RecQ helicase should not be influenced by the increasing concentration of protein. The ssDNA-stimulated ATPase activity of RecQ helicase has been studied as a function of RecQ protein concentration. As shown in Fig. 4, the k cat value is not affected by protein concentration up to 80 nM. Similar results were observed in the presence of ssDNA and dsDNA of different sequences or different lengths or at a higher temperature (42°C), suggesting that RecQ helicase monomers are an active form of the enzyme, at least as an ATPase. We next characterized the DNA unwinding activity of RecQ helicase at different protein concentrations. The time courses of the helicase reactions are shown in Fig. 5A. The unwinding rate was obtained from the fit of the unwinding kinetics curve by using a single exponential equation. Consistent with the ATPase assay, the DNA unwinding rate was unchanged as RecQ protein concentration increased (Fig. 5B). All of these experiments indicate that the ATPase and helicase specific activities do not increase by increasing the protein concentration.
Effect of the ATPase-deficient RecQ Helicase on the Wild Type RecQ ATPase and Helicase Activities-In order to further confirm that RecQ helicase functions as a monomer and no cooperative interactions between the subunits existed, two ATPasedeficient RecQ helicases were prepared for this purpose. It is well established that specific amino acids in the helicase conserved motifs I and II were involved in the ATPase activity of the enzyme (36,37). With the intention of preparing ATPasedeficient mutants, the amino acid lysine 55 in the helicase conserved motif I and asparagine 147 in motif II were mutated separately. If only one animo acid that is highly conserved in the nucleotide-binding loop is mutated in each modified RecQ protein, the modification should not change the three-dimensional structure of the protein. Such ATPase-deficient mutant should be still able to bind to DNA and involved in the proteinprotein interactions but unable to unwind duplex DNA substrates. We reasoned that if RecQ helicases function as a mon-  omer, the catalytic efficiencies for ATPase and DNA unwinding activities should not be influenced in the presence of increasing concentrations of ATPase-deficient helicase.
As expected, both mutants show no detectable ATPase and helicase unwinding activity (Table VII). The circular dichroism spectrum of both wild type and mutant proteins displayed no difference (results not shown), ruling out the possibility of structural change as the result of the mutation. In order to confirm that the ability of the mutants to bind to DNA was not impaired by a single amino acid substitution, the binding of the mutated RecQ helicases to DNA was studied under the equilibrium binding conditions (Fig. 6). The data obtained from wild type helicase fit well to a hyperbolic equation, whereas the binding isotherms determined from both modified RecQ helicases are sigmoidal. To determine the apparent K d values, the data were fitted to the Hill equation. The apparent dissociation constants are 25, 29, and 56 nM for wild type, K55A, and D148A, respectively (Table VII). These results indicated that the DNA binding activity of the ATPase-deficient mutants were essentially unchanged compared with that of the wild type helicase.
The ATPase activity of wild type RecQ helicase was determined in the presence of increasing amounts of either the K55A or D148A mutant protein. We observed that the rate of ATP hydrolysis by wild type RecQ helicase did not change as the mutant protein concentration was increased (results not shown), suggesting that there were nonprotein-protein interactions involved in ATPase activity.
We next performed a series of unwinding experiments to determine the effect of the mutant protein on the activity of wild type RecQ helicase. In order to give to the DNA substrate an equal opportunity to bind to the protein, all of the experiments described below were performed by premixing the wild type RecQ helicase and mutant helicase, and then the premixed solution was added to the reaction solution containing DNA. First, the unwinding activity of wild type enzyme was measured under the condition in which the different concentrations of wild type RecQ and mutant RecQ were mixed at a 1:1 ratio. Each kinetic curve was fit to a single exponential equation; the apparent unwinding rates were then plotted against the mutant helicase concentrations. As shown in Fig.  7A, the apparent unwinding rate of wild type RecQ helicase remains unchanged, whatever the mutant concentration. In a second experiment, the wild type RecQ helicase was held constant at 30 nM, and the unwinding activity was measured in the presence of increasing amounts of mutant helicase. As shown in Fig. 7B, a mild decrease of the apparent unwinding rate was detected as mutant K55A RecQ concentration increases. A similar phenomenon was observed with D148A RecQ mutant. The observed decrease of the apparent rate can be attributed either to interacting subunits or to simple competition between wild type and mutant protein for binding the DNA substrates. As will be discussed under "Discussion," these decreases arise from the competition between the two proteins for DNA binding rather than protein-protein interactions. Finally, we performed single-turnover kinetics using the premixed proteins to determine the unwinding rate constants. In this experiment, a large excess of nonspecific "trapping" DNA was added with ATP solution to prevent helicase from reassociating with the DNA substrates, thus ensuring that only the first cycle of unwinding activity is observed. Under these conditions, the unwinding rate constants of wild type helicase are not significantly influenced by the increasing concentration of mutant helicase (Fig. 7B, square). Taken together, these results provide strong evidence that no protein-protein interactions were involved in the DNA unwinding process. DISCUSSION We have investigated in this work the structural and functional properties of the E. coli RecQ helicase in solution by immunoprecipitation, size exclusion chromatography, analytical ultracentrifugation, and time-resolved fluorescence anisotropy as well as enzyme kinetics. The results presented shed some light on the hydrodynamic and conformational properties of this protein.
Conformational Properties of RecQ Helicase Free in Solution-First, all of the data of this study show that RecQ helicase is monomeric in solution. Immunoprecipitation analysis showed that no helicase-helicase interactions occur even in the presence of ATP or AMPPNP. Size exclusion chromatography and sedimentation velocity confirmed this result, since the R S values are consistent with a monomeric 68,290 Da protein.
Sedimentation equilibrium experiments further confirmed these observations by a direct determination of the M r of the enzyme. Furthermore, the long correlation time as determined by TFA (about 55 ns at 20°C) is compatible with the monomeric state, since, for instance, serum albumin, which has similar mass (66 kDa), has a correlation time in the same range (46 ns), and the ␥-interferon (34 kDa), half of the mass of RecQ helicase, has a correlation time divided by 2 (27 ns). TFA experiments established that RecQ helicase exists as a monomer up to a concentration of at least 20 M and that this monomer is stable from 4 to 37°C.
Size exclusion chromatography and sedimentation velocity experiments allowed the determination of the RecQ helicase Stokes radius. The values found for RecQ helicase are comparable with the above methods and gave R S equal to 34 and 31.1 Å, respectively. This R S knowledge allows deeper insight into the molecular shape of the protein. In fact, the expected theoretical R S for an unhydrated spherical protein characterized by  an M r of 68,290 g/mol is 27.1 Å. Assuming an hydration of 0.365 g of H 2 O/g of protein, which is a common value in the literature, the theoretical R S for RecQ helicase is equal to 31 Å, and the ratio R S (experimental)/R S (theoretical) is comprised between 1 and 1.09, suggesting that RecQ helicase is globular.
The three-dimensional structure of RecQ helicase is not yet available. However, the crystallized helicases PcrA and Rep shared homology with RecQ helicase and fold into a globular form that is reminiscent of a crab claw (5,8,38). These two proteins comprise two domains. Each domain is composed of a large central parallel ␤-sheet flanked by ␣-helices. Moreover, the comparison of the secondary structure of RecQ helicase with those of PcrA and Rep determined from the crystal structures shows that most of the secondary structural elements are conserved in the RecQ, PcrA, and Rep helicases (5,8). Thus, it is possible that RecQ helicase folds into a very similar three-dimensional structure compared with PcrA and Rep, suggesting that the globular shape may be a common feature of this protein family.
The Monomer Is the Functional Form of RecQ Helicase-To further confirm that RecQ helicase functions as a monomer in solution, we have studied both ATPase and DNA unwinding specific activities as the enzyme concentration increases. In an equilibrium condition system, increasing protein concentra-tions will shift the equilibrium toward formation of oligomer. If the enzyme functions as an oligomer, then its specific activity will be enhanced at higher concentrations of the protein, whereas, if the specific activity of the enzyme is independent of the protein concentration, this may suggest that the enzyme functions as a monomer. We found that the ATPase activity of RecQ helicase is not affected by increasing concentrations of the enzyme. In fact, above 70 nM, ATPase specific activity decreases as protein concentration increases. At a concentration of 200 nM, the ATPase specific activity of RecQ helicase was decreased by 30%. Similar results were also observed in previously published work (Fig. 8C of Ref. 19). However, the exact reason for the decrease in ATPase activity at high protein concentrations is not clear now. Consistent with the structural analysis, the DNA unwinding activity was independent of protein concentrations, indicating that no protein-protein interactions were involved in the unwinding process. We therefore conclude that the active form of RecQ helicase is a nomomer.
This conclusion was further tested by the mutant inhibition experiments. To examine whether a dimer or an oligomer is required for both the helicase and ATPase activities, 2 amino acid residues highly conserved in the motif I and motif II of the nucleotide-binding loop were mutated, respectively. The mutations did not affect the ability of K55A or D148A to bind to DNA. However, their ATPase and helicase activities are undetectable. The ATPase specific activity does not decrease in the presence of the increasing mutant protein concentration, indicating that RecQ ATPase activity is independent of subunits interactions. The helicase activity also remains unchanged in the presence of mutant helicase protein under conditions in which wild type and mutant helicases were kept at stoichiometry (Fig. 7A). However, we detected a mild decrease in helicase activity as mutant concentration increased while wild type helicase concentration was constant. The observed decrease of helicase activity may result from either the interacting subunits or the competition between wild type and mutant helicases for DNA substrate. In order to distinguish between these two possibilities, these data were analyzed in a quantitative manner according to Equation 8 (39), where H WT represents the helicase rate of the wild type RecQ protein, and E WT and E MUT are the concentrations of wild type and mutant helicase, respectively. n is the number of subunits in the complex. If n ϭ 1, this suggests that the enzyme function as a monomer, whereas if n is Ն2, the enzyme may function as a dimer or oligomer. The best fit of the data shown in Fig. 7B provides a value for n of 0.91 Ϯ 0.1. For comparison, the expected reduction in helicase activity, if RecQ helicase functions as a dimer or a hexamer, is also shown in Fig. 7B (dashed and dotted  lines). Clearly, the value for n Ն 2 fails completely to fit the data. This analysis shows that RecQ helicase unwinds DNA as a monomer. The observed mild decrease in helicase activity certainly  RecQ  610  100  Monomer  This work  RecQ1  649  42.3  Dimer  51  WRN  1432  43  Trimer  52  BLM  1417  45  Hexamer  40 a Helicase sequences are from GenBank™. The identity was determined using the Alignp program (InfoBiogen), and the sequence of E. coli RecQ helicase was taken as the reference. WRN, Werner syndrome protein; BLM, Bloom syndrome protein. results from the fact that the wild type helicase dissociates from the DNA during the course of the unwinding reaction, and the mutant protein could bind to DNA and affect the unwinding rate. If this is true, then the specific activity of RecQ helicase should not be affected by the addition of the mutant under condition in which only single cycle unwinding is performed. As expected, a single-turnover experiment shows that the catalytic constant of wild type helicase is independent of the concentration of mutant RecQ helicase. Taken together, these functional analyses clearly indicate that no subunits interactions are involved in the unwinding processes.
Based on the observation that RecQ helicase displays positive cooperativity in ATP hydrolysis with a Hill coefficient of 3.3, Harmon and Kowalczykowski (19) proposed a multimeric structure of at least three subunits as the active state. Moreover, on the analogy of the hexameric BLM helicase structure (40), the same authors suggested an hexameric structure for RecQ helicase. Although oligomeric enzymes may show a Hill coefficient higher than 1, the reciprocal is not always true. Enzymes that display a positive cooperativity with a Hill coefficient higher than 1 do not always correspond to an oligomeric structure. It is well established that there is a class of monomeric enzymes that display positive cooperativity for substrate binding (41)(42)(43)(44)(45). This positive cooperativity has been explained by a theoretical "mnemonical" model (43). This model assumes that monomeric enzyme exists mainly in one conformation state in the absence of ligands and that the binding of ligand induces a conformational transition to a new conformation that possesses a high affinity for the substrate. Considering the fact that the conformational state of helicases is usually modulated upon binding of DNA or nucleotide cofactor (1), it is not impossible that a monomeric helicase exhibits a positive cooperativity under certain conditions. By contrast, our direct structural and functional analysis of RecQ helicase, using a wide array of biochemical and biophysical method, clearly indicated that RecQ helicase functions as a monomer.
Functional Implication of the Monomeric Nature of RecQ Helicase-Whether the active form of helicases must be oligomeric is still uncertain and under debate (1,6). Previous biochemical studies on RecB protein, UvrD, HCV RNA helicase, and PcrA helicases revealed that these enzymes function as monomers (46 -49). In addition, the available crystal structures of several helicases such as Rep, PcrA, and the hepatitis C virus RNA helicase (5,8,50) crystallized with DNA and/or nucleotide show that these proteins are monomeric. The work presented here further enforces the idea that helicases function as monomers (6,49). The information available on the quaternary structure of RecQ helicase family is summarized in Table VIII. We can note that RecQ helicases adopt different quaternary structures from monomer to hexamer (40,51,52). This observation may suggest that (i) oligomeric structure may not be essential for the basic helicase activity or (ii) the quaternary structure of RecQ helicase family appears to be determined by the NH 2 -or COOH-terminal amino acid sequences, since the central domains of the RecQ helicases display high homologies (Table VIII). The fact that helicases adopting different quaternary structures display different enzymatic properties suggests that quaternary structure plays an important role in determining the mechanism by which helicases recognize the specific DNA substrate. The quaternary structure may also influence its unwinding rate and its processivity. Thus, monomeric helicases could retain full essential activities.