HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 26, 19498-19504, June 30, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19498    most recent M910091199v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Déclais, A.-C. Articles by Duguet, M. Search for Related Content PubMed PubMed Citation Articles by Déclais, A.-C. Articles by Duguet, M. Social Bookmarking                      What's this?

Reverse Gyrase, the Two Domains Intimately Cooperate to Promote Positive Supercoiling^*

Anne-Cécile Déclais, Janine Marsault, Fabrice Confalonieri, Claire Bouthier de La Tour, and Michel Duguet^§

From the Laboratoire d'Enzymologie des Acides Nucléiques, Institut de Génétique et Microbiologie, UMR 8621 CNRS, Bât. 400, Université de Paris Sud, Centre d'Orsay, 91 405 Orsay Cedex, France

Received for publication, December 20, 1999, and in revised form, March 8, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Reverse gyrases are atypical topoisomerases present in hyperthermophiles and are able to positively supercoil a circular DNA. Despite a number of studies, the mechanism by which they perform this peculiar activity is still unclear. Sequence data suggested that reverse gyrases are composed of two putative domains, a helicase-like and a topoisomerase I, usually in a single polypeptide. Based on these predictions, we have separately expressed the putative domains and the full-length polypeptide of Sulfolobus acidocaldarius reverse gyrase as recombinant proteins in Escherichia coli. We show the following. (i) The full-length recombinant enzyme sustains ATP-dependent positive supercoiling as efficiently as the wild type reverse gyrase. (ii) The topoisomerase domain exhibits a DNA relaxation activity by itself, although relatively low. (iii) We failed to detect helicase activity for both the N-terminal domain and the full-length reverse gyrase. (iv) Simple mixing of the two domains reconstitutes positive supercoiling activity at 75 °C. The cooperation between the domains seems specific, as the topoisomerase domain cannot be replaced by another thermophilic topoisomerase I, and the helicase-like cannot be replaced by a true helicase. (v) The helicase-like domain is not capable of promoting stoichiometric DNA unwinding by itself; like the supercoiling activity, unwinding requires the cooperation of both domains, either separately expressed or in a single polypeptide. However, unwinding occurs in the absence of ATP and DNA cleavage, indicating a structural effect upon binding to DNA. These results suggest that the N-terminal domain does not directly unwind DNA but acts more likely by driving ATP-dependent conformational changes within the whole enzyme, reminiscent of a protein motor.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Reverse gyrase is one of the most typical examples of the originality of biological macromolecules produced by hyperthermophilic organisms. Initially discovered in an Archaeon (see Refs. 1 and 2 for a review), it has been further isolated from a variety of archaea and bacteria and has been recognized as a marker of hyperthermophiles, independently of the domain of life to which it belongs (3-5). The basic activity performed in vitro by this enzyme is the production of positive supercoils in a closed circular DNA, a thermodynamically unfavored reaction that is driven by ATP. This activity provided possible explanations for the specific presence of reverse gyrase in hyperthermophiles, since positive supercoiling may stabilize the double helix at high temperatures, preventing local opening of the helix or allowing duplex "renaturation" after the passage of a transcription complex (6). An additional "raison d'être" of reverse gyrase could be the use of positive supercoiling to assemble chromatin-like structures in some hyperthermophiles (7). More recently, a similar activity was proposed to be present in eukaryotes (8), where it could be involved in disrupting nucleosomal structures or in destabilizing illegitimate recombination intermediates (9), thus giving reverse gyrase a more universal status.

A crucial question that could shed some light on the function of reverse gyrase remains unanswered. What is the mechanism by which this sophisticated enzyme is able to catalyze a positive supercoiling reaction? Several years ago, two unexpected results somewhat clarified the problem. One was the finding that reverse gyrase, despite its unique ATP dependence, was a type I topoisomerase, transiently bound to the 5' end of the cleaved strand (10), as the other members of the type I-5' family (11). The second was that stoichiometric binding of reverse gyrase to a form II DNA in the absence of ATP decreased the linking number of DNA after closure by a ligase, a property that was interpreted as DNA unwinding and was reminiscent of helicases (10).

An additional clue to the reverse gyrase mechanism was later provided by the cloning and sequencing of the enzyme from Sulfolobus acidocaldarius in our laboratory (12). The sequence that came out was clearly divided into two halves. The N-terminal half exhibited the usual signatures of a family of DNA/RNA helicases, and the C-terminal half was similar to the type I-5' topoisomerases, confirming the previous results. All of the other recently sequenced reverse gyrase genes also contain the signatures of these two protein families, although the gene structure can be somewhat different (13). Based on this structure in two domains, we proposed a "dynamic" model for the mechanism in which we assumed that the N-terminal domain had an ATP-dependent helicase activity; helix tracking by this domain was supposed to produce a wave of positive supercoiling ahead of reverse gyrase and of negative supercoiling behind (14). Specific relaxation of these negative supercoils by the topoisomerase I domain was supposed to finally produce net positive supercoiling (12). This model supported the idea that in eukaryotes the function of reverse gyrase could be fulfilled by the association of a helicase and a topoisomerase (9).

However, recent results (15, 16) on the sequence specificity of reverse gyrase are difficult to reconcile with the above model, which implies that the protein moves rapidly along the DNA. Moreover, the model was exclusively based on protein sequence comparisons, so that the precise mechanism of reverse gyrase is still a matter of hypotheses. In order to confirm these, we expressed separately in Escherichia coli the entire 1247 amino acid putative coding sequence of S. acidocaldarius reverse gyrase and its putative helicase-like (amino acids 1-612) and topoisomerase (residues 610-1247) domains.

In the present paper, we show that full-length recombinant reverse gyrase, when expressed at 37 °C in E. coli, is able to sustain efficient positive supercoiling at 75 °C in the presence of ATP, without requiring any preheating treatment. As expected, mutation of the putative active site tyrosine 964 into a phenylalanine completely abolished this activity. In addition, the putative topoisomerase I domain alone, expressed in E. coli, exhibited an ATP-independent DNA relaxation activity on negative supercoils at 75 °C. By contrast, we failed to detect any helicase activity for the N-terminal domain as for the whole reverse gyrase. However, simple mixing of the two domains reconstitutes the ATP-dependent positive supercoiling typical of reverse gyrase. The interaction between the domains seems rather specific, since the topoisomerase domain cannot be replaced by another thermophilic topoisomerase I, and the helicase-like domain cannot be replaced by a thermophilic helicase.

Finally, we show that the stoichiometric DNA unwinding activity is not an intrinsic property of the helicase-like domain, contrary to our previous estimation (17). Like the supercoiling activity, DNA unwinding requires the presence of both domains, be they separately expressed or linked in a single polypeptide. These results suggest a mechanism in which the N-terminal domain does not directly unwind DNA but is more likely a protein motor, driving conformational changes within the whole enzyme.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials

Bacterial strains used in this work were DH5 (cloning) and XL mutS Kan^r (mutagenesis). Culture media were from Life Technologies, Inc. Restriction enzymes, T4, and Thermus DNA ligases were purchased from New England Biolabs, and oligonucleotides were from Genosys.

Methods

Construction of Expression Vectors-- Reverse gyrase and its putative domains were produced in E. coli as glutathione S-transferase (GST)¹ fusion proteins, with the pGEX expression system described by Smith and Johnson (18). The reverse gyrase complete coding sequence from S. acidocaldarius was inserted in two steps into the pGEX-2TH vector. (i) We used polymerase chain reaction to insert a BamHI site immediately 5' to the first codon and to amplify the 5' terminus of the gene from a recombinant  phage carrying the entire gene (12). The amplified fragment was digested by BamHI and BsaBI. (ii) The same  phage was digested by BsaBI and HindIII to prepare the 3' part of the gene, and both fragments were inserted into the BamHI-HindIII site of the vector.

The design of the reverse gyrase domains and their boundaries was exclusively based on the sequence: we used secondary structure predictions (Mac Molly package, Soft Gene, GmbH) to define a flexible loop, located about 15 amino acids upstream the first topoisomerase motif, as the junction between the two putative domains. We thus defined the helicase-like domain as amino acids 1-612 and the topoisomerase domain as amino acids 610-1247. The corresponding DNA segments were inserted into the cloning vector by using the same strategy as for the complete enzyme.

Site-directed Mutagenesis-- Mutagenesis of the putative active site tyrosine into a phenylalanine was performed with the Chameleon^TM kit of Stratagene, as described by the manufacturer, to yield the RG-Y964F and RG Top Y352F mutants.

Expression and Purification of Recombinant Proteins-- Two liters of LB broth were inoculated with E. coli DH5 strain carrying the various plasmid constructions. Cells were grown at 37 °C with aeration to an A₆₀₀ of 0.8, and isopropyl-1-thio--D-galactopyranoside was added to a final concentration of 0.1 mM. After 2 h, cells were quickly cooled on ice and harvested. The cell paste (4 g) was resuspended in 10 volumes of lysis buffer (50 mM Tris, pH 7.4, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 10% glycerol, 1% Nonidet P-40), and the cells were lysed in a French press at 9,000 pounds/square inch. Cellular debris were removed by centrifugation at 10,000 × g for 15 min, and the fusion proteins were purified on glutathione-agarose (from Sigma) using a batch procedure modified from Smith and Johnson (18); the crude extract was stirred with 1.6 ml of glutathione-agarose for 1 h at 4 °C, and the resin was harvested by centrifugation at 500 × g for 10 min. After extensive washing with 1 M NaCl in 50 mM Tris, pH 8.0, the fusion proteins were eluted by 20 mM reduced glutathione. The GST tag was removed by incubation for 3 h at 26 °C with thrombin. In the case of the full-length reverse gyrase, an additional purification step on phenyl-Sepharose was necessary to remove fragments produced by proteolysis or incomplete expression. The sample, adjusted to 300 mM NaCl, was loaded onto a 0.5-ml phenyl-Sepharose CL-4B column (Amersham Pharmacia Biotech) equilibrated in buffer A (50 mM Na₂HPO₄/NaH₂PO₄, pH 7.0, 1 mM DTT, 1 mM EDTA, 300 mM NaCl). After the column was washed by 25% ethylene glycol in the same buffer, reverse gyrase was eluted by 60% ethylene glycol. The various fractions were analyzed by SDS-polyacrylamide gel electrophoresis and checked for topoisomerase activities.

Helicase Assay-- This assay was essentially based on the displacement of a labeled oligonucleotide from a partial duplex, with modifications for an adaptation to high temperatures. Two types of substrates were used. In the first substrate (19), a 82-mer oligonucleotide, 5'-end-labeled using T4 polynucleotide kinase (Biolabs) was annealed to M13mp18 single-strand circles through its central part (nucleotides 24-59) leaving both a 5' tail (residues 1-23) and a 3' tail (residues 60-82) unhybridized. Other oligonucleotides, untailed or simply tailed, were also used. In the second type of substrate, described by Tanguy Le Gac et al. (20), a 5'-labeled 59-mer oligonucleotide was partially hybridized (over 34 nucleotides) to a 90-mer, forming an asymmetric Y-shaped substrate, with a 5'-labeled 25 nucleotide tail and a 3'-unlabeled 56 nucleotide tail.

The reaction mixtures (10 µl) contained 20 mM Tris, pH 7.5, 50 mM NaCl, 3 mM MgCl₂, 1 mM DTT, 10% glycerol, 1 mM ATP, 0.02 pmol of substrate, and various amounts of the recombinant proteins to test (up to 500 ng). For the second type of substrates, a 27-mer oligonucleotide, complementary to the 90-mer, was added in 4-fold excess to trap the displaced 90-mer. After 10-30 min at various temperatures, the products were analyzed by 12% polyacrylamide non-denaturing gel electrophoresis. PcrA helicase from Bacillus stearothermophilus used as a control was a generous gift of Dr. D. B. Wigley (Oxford).

Assay for Topoisomerase Activities-- The same assay was used to monitor the relaxation of negatively supercoiled DNA and positive supercoiling, both activities corresponding to an increase of the DNA linking number. The recombinant proteins (0.5-10 ng) were incubated in a 20-µl reaction mixture containing 50 mM Tris, pH 8.0, 10 mM MgCl₂, 120 mM NaCl, 0.5 mM DTT, 30 µg/ml bovine serum albumin with 400 ng of negatively supercoiled pTZ18 (Amersham Pharmacia Biotech) at 75 °C for 30 min in the presence or absence of 1 mM ATP. Reactions were stopped by quick cooling on ice and addition of 0.5% SDS, 10 mM EDTA, 5% glycerol, and 0.02% bromphenol blue. The incubation products were analyzed by bidimensional 1.2% agarose gel electrophoresis as described previously (21), with addition of 3 µg/ml chloroquine in the gel and the buffer for the second electrophoresis. For some experiments, monodimensional, 1.2% agarose gel electrophoresis was performed.

DNA Unwinding Assay-- This assay measures the change in the DNA linking number after stoichiometric binding of a form II DNA with a protein and closure by ligase. In our case, the assay was performed at 75 °C, essentially as described previously (10). Briefly, reaction mixtures (20 µl each in siliconized tubes) contained 20 mM Tris-HCl, pH 7.5, 30 mM NaCl, 25 mM potassium acetate, 10 mM magnesium acetate, 10 mM DTT, 1 mM NAD, 30 µg/ml bovine serum albumin, 33 ng of pTZ 18 form II DNA with an average of one single-strand break per circle, and various amounts of proteins to be tested. Two microliters of a dilution of Taq DNA ligase (Biolabs, 2 units) in 10 mM Tris-HCl, pH 7.5, 100 mM KCl, 20% glycerol were introduced, prior to incubation at 75 °C, as a drop on the inner wall of each siliconized tube but not mixed with the reaction medium. After 10 min of incubation at 75 °C, each tube was quickly agitated to mix the ligase drop with the other components and further incubated for 5 min at 75 °C. This procedure allowed us to avoid variations in the linking number due to temperature changes. After incubation, the tubes were cooled, centrifuged for 30 s, and treated with 50 mM EDTA, 1% SDS, and 0.5 mg/ml proteinase K for 30 min at 50 °C. The reaction products were separated by bidimensional gel electrophoresis, transferred to Hybond N⁺ membrane (Amersham Pharmacia Biotech), and hybridized with a ³³P-labeled pTZ18 probe (random priming, Roche Molecular Biochemicals). The products were revealed either by autoradiography or by using a PhosphorImager (Molecular Dynamics).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Full-length Recombinant Reverse Gyrase Sustains ATP-dependent Positive Supercoiling at High Temperature-- The entire 1247 amino acid coding sequence of reverse gyrase from S. acidocaldarius was expressed in E. coli as a fusion protein with GST and partly purified by affinity chromatography on glutathione-agarose (see "Experimental Procedures"). In all conditions tested, the level of expression was low, and after the fusion polypeptide was cleaved by thrombin, an additional purification step on phenyl-Sepharose was needed to remove a large part of the remaining contaminants (Fig. 1A). These contaminants, eluted at 25% ethylene glycol (lane 1), are fragments of reverse gyrase since they react with antibodies directed against reverse gyrase in Western blots but have no detectable activity on supercoiled DNA. The main polypeptide, eluted at 60% ethylene glycol, had the expected size for full-length reverse gyrase on polyacrylamide gel electrophoresis (Fig. 1A, lanes 3 and 4). This recombinant protein was able to fully convert negatively supercoiled plasmid DNA to positive supercoils at 75 °C in the presence of ATP, as shown on the bidimensional analysis of Fig. 1B. This result suggests that, although expressed in E. coli at 37 °C and inactive at this temperature, recombinant reverse gyrase is correctly folded to be active at high temperature as is the enzyme from Sulfolobus. As expected, the mutant enzyme obtained by converting tyrosine 964 into a phenylalanine totally lacks topoisomerase activity, confirming the identity of the active site tyrosine (see Fig. 1C).

View larger version (46K): [in this window] [in a new window]   Fig. 1.   Purification and activities of reverse gyrase expressed in E. coli. A, polyacrylamide gel electrophoresis of the proteins eluted from phenyl-Sepharose: silver staining. Lane 1, elution with 25% ethylene glycol. Lanes 2-4, fractions eluted with 60% ethylene glycol. The migration of the size markers (kDa) is indicated by arrows. B, bidimensional gel electrophoresis of pTZ 18 DNA (400 ng) after incubation at 75 °C and 120 mM NaCl with various fractions from the phenyl-Sepharose column. In this bidimensional analysis, negatively supercoiled topoisomers constitute the left branch of the arch, and positively supercoiled topoisomers constitute the right branch. The left upper band is form II DNA. Panels a and b, incubation with fractions from lanes 3 and 4 of the phenyl-Sepharose column, containing 10 and 6 ng of reverse gyrase, respectively. Panels c and d, activity of side fractions containing about 3 and 0.5 ng of reverse gyrase, respectively. Upper panels, + symbols indicate incubations in the presence of ATP; lower panels  symbols indicate the absence of ATP. C, monodimensional agarose gel electrophoresis of pTZ18 (400 ng) after incubation with reverse gyrase mutant Y964F (10 ng) in the absence () or presence (+) of ATP. Conditions of incubation are indicated under "Experimental Procedures." A control with the plasmid alone (200 ng) is shown in the left lane (pTZ).

Activities of the Separately Expressed Topoisomerase and Helicase-like Domains of Reverse Gyrase-- Since sequence analysis (12) suggested that the N-terminal part of reverse gyrase had helicase signatures, and the C-terminal part exhibited topoisomerase I motifs, we designed two polypeptides, covering each half of the whole reverse gyrase sequence. The rationale for the design of these proteins is based on secondary structure predictions and the presence of a highly conserved region in the N terminus of the topoisomerase part (see "Experimental Procedures"). We then expressed both putative domains separately, using the same GST fusion expression system. The helicase-like domain appears as a main band of the expected size, whereas the topoisomerase domain invariably appears as a triplet (Fig. 2A), reminiscent of reverse gyrase degradation fragments (see Fig. 1A, lane 1).

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Reconstitution of reverse gyrase activity by mixing the two domains. A, polyacrylamide gel electrophoresis (silver staining) of the helicase-like (Hel) and topoisomerase (Top) domains produced in E. coli and purified on glutathione-agarose. The migration of the size markers (kDa) is indicated by arrows. B-D, bidimensional gel electrophoresis. Incubations were at 75 °C in the presence of 80 mM NaCl. B, analysis of pTZ 18 (400 ng) after incubation with the helicase domain (Hel, 6 ng, upper panel), with the topoisomerase domain (top, 6 ng, middle) or with the mixture of 6 ng of each domain (bottom) giving a molar ratio of about 1:1. C, analysis of pTZ 18 (400 ng) after incubation with increasing helicase-like domain at fixed amount (3 ng) of the topoisomerase domain. Panels a-f, molar helicase to topoisomerase ratio of 0 (a), 1:16 (b), 1:4 (c), 1:2 (d), 1:1 (e), 2:1 (f). For each experiment,  and + indicate incubations in the absence or in the presence of ATP, respectively. D, incubation of pTZ 18 (400 ng) with low amounts of the topoisomerase domain (1 ng) in the absence (panel g) or in the presence (panel h) of the helicase domain (2 ng).

As expected, the putative topoisomerase I domain was able to partly relax a negatively supercoiled plasmid in an ATP-independent reaction at 75 °C (Fig. 2B, middle). This result, and the complete lack of activity of the putative active site mutant Y352F (data not shown), confirmed that the C-terminal part of reverse gyrase was indeed a topoisomerase domain, as suspected from its sequence. However, this domain exhibited a poor relaxation efficiency (see Fig. 2B, middle, C, panel a, and D, panel g).

The putative helicase domain was tested for helicase activity against a variety of substrates formed by untailed, 5'-tailed, 3'-tailed, or doubly tailed oligonucleotides annealed to M13 single-strand circles (see "Experimental Procedures"). As shown in Fig. 3, neither the helicase-like domain nor the full-length reverse gyrase exhibited helicase activity on the doubly tailed substrate, whereas the thermophilic helicase PcrA from B. stearothermophilus (22) was able to partially release the 82-mer oligonucleotide (panel 3). The same result was obtained with the other substrates in a variety of experimental conditions, at 37, 50, 60, or 70 °C (not shown). In another set of experiments, a Y-shaped substrate formed by partial annealing of a 90-mer oligonucleotide to a 5'-labeled 59-mer was used in place of M13 substrates (see "Experimental Procedures") to avoid possible trapping of the proteins by the large amount of M13 single-strand present. Again, in our hands, only the PcrA helicase was able to displace the 59-mer oligonucleotide.

View larger version (49K): [in this window] [in a new window]   Fig. 3.   Oligonucleotide displacement assay for the helicase-like domain and the full-length reverse gyrase. Autoradiography of the products after polyacrylamide gel electrophoresis. The substrate is an 82-mer oligonucleotide partially annealed to M13 single-stranded circle. All incubations are for 10 min at 70 °C in the presence of 1 mM ATP. Lane 1, substrate (40 ng) incubated alone; lane 2, boiled substrate; lane 3, substrate incubated with PcrA (200 ng); lane 4, substrate incubated with reverse gyrase (400 ng); lane 5, substrate incubated with the helicase-like domain (200 ng). The bands corresponding to the annealed substrate and to the free oligo are indicated by arrows.

Mixing Separately Expressed "Helicase-like" and "Topoisomerase" Domains Reconstitutes Efficient Reverse Gyrase Activity-- We next tried to reconstitute reverse gyrase by simply mixing the two domains at a molar ratio of 1:1. Indeed, incubation of this mixture with plasmid DNA at 75 °C produced ATP-dependent positive supercoiling as efficiently as the full-length enzyme (compare Fig. 2B, bottom with Fig. 1B, left). In this test, the helicase-like domain alone had no apparent activity on the DNA substrate, whereas the topoisomerase domain exhibited a low relaxation activity (Fig. 2B, top and middle panels). Remarkably, this latter activity is stimulated by the presence of the helicase-like domain in the absence of ATP (Fig. 2B, compare middle to bottom).

The appearance of a reverse gyrase activity was monitored by adding increasing amounts of the helicase-like domain to a fixed amount of the topoisomerase domain (Fig. 2C, panels a-f); a plateau of activity was obtained for a 1/1 Hel/Top molar ratio (panel e). Addition of excess "helicase" had no detectable effect (panel f). The same result was obtained by testing increasing amounts of the topoisomerase domain at fixed helicase concentration (not shown). Interestingly, excess of the topoisomerase domain did not change the topoisomer distribution, confirming that this domain was not able to relax positive supercoils.

Unexpectedly, when amounts of the topoisomerase domain, low enough to present no visible activity on the substrate (Fig. 2D, panel g), were supplemented by the helicase-like domain, an efficient reverse gyrase activity was nevertheless reconstituted (panel h). This indicates a strong activation of the topoisomerase activity through interactions between the domains. Finally, reconstitution did not need a preincubation of the mixed domains at low or high temperature, suggesting that when expressed in E. coli, the domains already had a conformation allowing mutual recognition or can acquire it very rapidly when the temperature is increased. Together, these results suggest that the separately expressed domains form a heterodimer that mimics reverse gyrase.

The Reconstituted "Reverse Gyrase" Is Less Stable Than the Full-length Enzyme-- The stability of the reconstituted activity toward ionic strength and temperature was compared with that of reverse gyrase. As shown on Fig. 4, the activity of the full-length recombinant enzyme is poor at 30 mM salt, maximal around 120 mM, and still important at 200 mM, which is similar to the behavior of the endogenous reverse gyrase from Sulfolobus (21). By contrast, the activity exhibited by the mixture of the two domains is highest at low salt and practically nonexistent at 200 mM. The same type of result was obtained when increasing the incubation temperature up to 90 °C; the mixed domains lost their ability to produce positive supercoiling, whereas reverse gyrase retained some activity (not shown). These results suggest a relatively tight association between the domains, which is, however, not strong enough to retain activity in high salt or temperature conditions in the absence of a covalent link between them.

View larger version (46K): [in this window] [in a new window]   Fig. 4.   Activity of the reconstituted reverse gyrase in the presence of increasing salt concentrations; comparison with full-length reverse gyrase:bidimensional agarose gel electrophoresis of the incubation products. NaCl concentrations are indicated. Hel, helicase-like domain (3 ng); top, topoisomerase-like domain (3 ng); RG, full-length recombinant reverse gyrase (4 ng). ATP is present in all incubations.

The Association of the Two Domains in the Reconstituted Reverse Gyrase Is Specific-- The next question that we addressed was how specific the interaction is between the two domains. In other words, supposing that reverse gyrase is a modular enzyme, is it possible to replace either module by a functional equivalent? To test this possibility, we replaced the topoisomerase domain by the type I topoisomerase from Thermotoga maritima, a hyperthermophilic bacterium (23, 24). This choice had two advantages as follows: (i) this topoisomerase works at 75-80 °C, and (ii) it has approximately the same size (633 amino acids) as the topoisomerase domain of reverse gyrase (635 amino acids). Incubation of the Thermotoga topoisomerase I with the helicase-like domain of reverse gyrase in the usual reaction mixture did not reveal any positive supercoiling activity (Fig. 5, top), and all conditions tested failed to detect this activity. The same result was obtained when the full-length reverse gyrase mutated in its active site tyrosine (Y964F) was incubated with the Thermotoga topoisomerase I (Fig. 5, bottom). However, in both cases, the presence of the helicase domain, free or included in the reverse gyrase mutant, appeared to stimulate the relaxation activity of the Thermotoga topoisomerase, independently of the presence of ATP (compare panels 1 and 2 to 3 and 4 and panelss 5 and 6 to 7 and 8). The possibility to replace the helicase-like domain by a "true" thermophilic helicase, the PcrA helicase from B. stearothermophilus (22) in the reconstitution assay was also tested and failed (not shown).

View larger version (65K): [in this window] [in a new window]   Fig. 5.   Replacement of the topoisomerase domain by a type I-5' topoisomerase. Panels 1-4, the topoisomerase domain is replaced by T. maritima topoisomerase I (0.1 ng). Panels 1 and 2, T. maritima topoisomerase I alone; panels 3 and 4, mixture of the helicase domain (2 ng) with T. maritima topoisomerase I. Panels 5-8, the helicase-like domain was replaced by the catalytically inactive Y964F reverse gyrase mutant. Panel 5, T. maritima topoisomerase I alone; panels 6-8, mixture of T. maritima topoisomerase I with 5, 10, and 20 ng of Y964F, respectively. ATP is absent from incubations 1 and 3.

Stoichiometric DNA Unwinding Also Requires Cooperation of the Two Domains-- In a previous work, we showed that in the absence of ATP and at high temperature, the stoichiometric binding of Sulfolobus reverse gyrase to an open circular DNA resulted in a decrease of the DNA linking number after closure by a thermophilic ligase. This was interpreted as an unwinding of the DNA upon reverse gyrase binding (10), a property that is shared by some helicases. We decided first to test this property with the recombinant reverse gyrase mutated on the active site tyrosine (Y964F), so that there was no possible DNA cleavage or interference with the positive supercoiling activity. In addition, it was possible to check the effect of ATP and non-hydrolyzable analogs on this unwinding activity. Incubation at 75 °C of an open circular plasmid with increasing amounts of this mutant in the absence of ATP, followed by covalent closure with the ligase from Thermus, resulted in the progressive appearance of a negatively supercoiled plasmid (Fig. 6A). At a ratio of about 40 protein molecules per DNA circle, a population migrating as a highly supercoiled plasmid was obtained. This result indicates that unwinding by reverse gyrase does not require ATP nor DNA cleavage.

View larger version (90K): [in this window] [in a new window]   Fig. 6.   Stoichiometric DNA unwinding by catalytically inactive reverse gyrase and its domains. Autoradiographies of bidimensional gels blotted and hybridized with 33P-labeled pTZ18 probe. A, covalent closure of form II DNA (33 ng) at 75 °C and 30 mM NaCl by the Thermus thermophilus DNA ligase in the presence of increasing amounts of reverse gyrase Y964F mutant. Panel T is a form II control incubated without ligase. Panels 0, 4, 10, 20, and 40 indicate the total number of reverse gyrase molecules per DNA circle. ATP is omitted. B, same experiment as in A with the mixed domains, helicase-like domain, and catalytically inactive topoisomerase domain (Y352F) in a 1:1 ratio. Indicated is the total number of protein molecules per DNA circle. A trace of residual Form I is present in this Form II preparation (see T control). C, influence of ionic strength on the unwinding activity. Reverse gyrase Y964F (lower panel), mixed domains, and helicase plus Y352F (upper panel) were incubated as in A at a ratio of 40 molecules per DNA circle in the presence of indicated salt concentrations. An additional 25 mM potassium acetate is present in all incubations. T indicates controls without ligase.

We next asked whether the helicase-like domain itself was able to unwind DNA, so that its binding would provide a substrate for the -like topoisomerase domain, which is one of the proposed models for the reverse gyrase mechanism (17). By using the same protocol as above, we were not able to detect any DNA unwinding ability for the helicase domain itself, even at molar ratios up to 80 (not shown). This result came as a surprise, and we then tested whether this unwinding could be performed by the topoisomerase domain alone, using the catalytically inactive mutant (Y352F, see "Experimental Procedures"). Once again, we could not detect any such activity.

Finally, we asked whether we could mimic the unwinding property of reverse gyrase by mixing the helicase-like domain with the mutant Y352F topoisomerase domain. Fig. 6B shows that this was indeed the case, although less efficiently than with reverse gyrase. Again, maximal effect was obtained for equimolar proportions of the domains, the degree of negative supercoiling increased to a plateau when increasing amounts of the helicase-like domain were added to a fixed concentration of the Y352F topoisomerase domain and vice versa. This unwinding activity was not influenced by the presence of ATP, AMPPNP, or ADP (data not shown). Finally, the stoichiometric unwinding activity of the reconstituted complex is less stable than that of reverse gyrase, as we observed for the catalytic positive supercoiling activity (Fig. 6C).

Together, these results support the view that the two domains form a specific complex where they intimately cooperate to mimic the full-length reverse gyrase and that the positive supercoiling activity is not the simple addition of two different activities.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The experiments described in this paper bring new insights into the mechanism of reverse gyrase. Table I summarizes the main activities of the recombinant fragments of reverse gyrase.

(i) The full-length reverse gyrase, expressed as a recombinant protein in E. coli, is able to sustain ATP-dependent positive supercoiling at 75 °C, just as the wild type Sulfolobus reverse gyrase does. This suggests that no major post-translational modification of the enzyme occurs in Sulfolobus. Replacement of the putative active site tyrosine by a phenylalanine abolished all topoisomerase activities. However, this mutant Y964F is still able to promote stoichiometric DNA unwinding in a ligation-mediated assay in the absence of ATP (Table I). This indicates that reverse gyrase binds and unwinds DNA, independently of ATP binding and DNA cleavage.

(ii) The C-terminal half of Sulfolobus reverse gyrase, expressed in E. coli, exhibits a topoisomerase I activity, independent of the presence of ATP and specific of negative supercoils, confirming the relevance of type I-5' topoisomerase signatures found in the sequence. However, this activity is low when compared with the other topoisomerases I and is more reminiscent of the activity found in topoisomerase III. It is also comparable to the residual relaxation activity of reverse gyrase in the absence of ATP (21). Thus, our previous hypothesis to explain the low activity of reverse gyrase in the absence of ATP, a repression of the topoisomerase domain by the other domain, now appears unlikely. On the contrary, DNA relaxation by the topoisomerase domain is slightly stimulated by the helicase-like domain in the absence of ATP (Fig. 2B), suggesting a positive control of the helicase-like domain over the topoisomerase domain within reverse gyrase. As for the full-length enzyme, replacement of the same tyrosine 964 (position 352 of the topoisomerase domain) totally suppressed topoisomerase activity. Finally, contrasting with the complete reverse gyrase, the C-terminal domain is not able to sustain stoichiometric unwinding (Table I).

(iii) Our attempts to demonstrate DNA helicase activity of the N-terminal half of reverse gyrase or of the whole enzyme failed, both with an oligonucleotide annealed to M13 circles and with a Y-shaped substrate. However, since some helicases have very limited strand displacement effect, other experiments are needed to confirm the lack of helicase activity. The N-terminal domain also failed to exhibit DNA unwinding in a ligase-mediated assay (Table I). This suggests that the domain itself is not a DNA-unwinding module, contrary to our previous proposition from studies on Sulfolobus shibatae reverse gyrase (17).

(iv) An important result of the present study is the successful reconstitution of reverse gyrase activity from its putative domains. Reconstitution of topoisomerase activities from separate subunits has been obtained in several cases (13, 25), but it is one of the first examples of reconstitution from separately expressed putative domains designed solely on the basis of amino acid sequence comparisons. This result indicates that the separate domains are able to naturally adopt a fold similar to their fold within the full-length reverse gyrase. This is consistent with the finding that reconstitution does not need preincubation of the domains. Maximum activity is obtained for equimolar amounts of the domains, suggesting that they form a heterodimer that mimics reverse gyrase. In the unique case of the two-subunit reverse gyrase from Methanopyrus kandleri, Krah et al. (13) reconstituted positive supercoiling activity by heating the mixture of the two subunits. However, in this case, the natural subunits do not coincide to the previously defined sequence domains; for instance, the large ATP-binding subunit (RgyB) encompasses a part of the topoisomerase domain. This may explain the need of a preheating treatment to reach the proper oligomeric structure for activity. Finally, although in the present work with S. acidocaldarius reverse gyrase we can reconstitute efficient positive supercoiling without preheating, this activity is more sensitive to ionic strength and temperature. This finding is consistent with a gain in thermodynamic stability corresponding to an additional peptide bond. From an evolutionary point of view, this would provide a selective advantage for a reverse gyrase formed by the fusion of two ancestral domains in a unique chimeric polypeptide.

(v) In reconstitution experiments, the topoisomerase domain cannot be replaced by the thermophilic topoisomerase from T. maritima, whereas the N-terminal domain cannot be replaced by the thermophilic helicase PcrA from B. stearothermophilus (Table I). These findings put a doubt on the validity of the dynamic model initially proposed, which supposed the cooperation of a true helicase with any type I-5' topoisomerase. On the contrary, a mutual and specific recognition of the two domains to promote positive supercoiling seems more likely.

(vi) Finally, the mix of the two domains also reconstitutes stoichiometric DNA unwinding (Table I). This finding is of considerable importance to understanding the mechanism of reverse gyrase; it suggests a strong cooperation of the two domains at the structural level to unwind the double helix.

In 1995 (2), we proposed the following model for the mechanism of positive supercoiling by reverse gyrase. (i) The binding of a high energy form of the enzyme induces a partition between an overwound DNA domain and an underwound region where the protein is bound. (ii) The underwound region is relaxed. (iii) The enzyme is released from the DNA, which is now overwound. At that time, reverse gyrase was seen as a modular enzyme, in which the helicase-like domain was responsible for the local DNA unwinding upon binding, and the topoisomerase domain relaxed the underwound substrate thus created.

The present work rules out this simplistic hypothesis. The two domains of reverse gyrase do not function independently of each other. Indeed, the high efficiency of the positive supercoiling reaction of the reconstituted enzyme is in strong contrast with the weak or nonexistent activities of the domains by themselves. Moreover, they form a specific complex since neither of them can be replaced by a functional equivalent. Finally, we show that the underwound DNA region is tightly maintained by the protein, since it is not accessible to an exogenous type I-5' topoisomerase. Indeed, there is no in vitro complementation of the inactive reverse gyrase mutant Y964F by the topoisomerase domain or by the topoisomerase I from T. maritima. Thus, we propose a different view on the role of the N-terminal domain and its ATP-binding site. We believe that there is an intimate interaction between the two domains of reverse gyrase, the ATP binding and hydrolysis being used to promote conformational changes within the whole enzyme at each cycle of topoisomerization. In this sense, the N-terminal domain would be more reminiscent of ATP-driven chaperones than of ATP-dependent DNA helicases. This possibility is still compatible with sequence data, where the conserved helicase motifs are mainly devoted to ATP binding and hydrolysis and to ATP-driven protein conformational changes (26). This hypothesis has to be supported by the direct demonstration of ATP-dependent conformational changes in reverse gyrase.

Finally, the present hypotheses on the enzyme mechanism might provide new clues on the open questions of the biological function of reverse gyrase and its possible presence in eukaryotes. Indeed, many Archaeal genes involved in DNA transactions have equivalents in eukaryotes. We have previously proposed that, in eukaryotes, equivalents of reverse gyrase might be built by the association of a topoisomerase with a DNA helicase and could be involved in the processing of replication and recombination intermediates (9). Alternatively, as suggested here, equivalents of reverse gyrase in eukaryotes could result from the association of a topoisomerase with a protein motor and could act to displace proteins from the DNA (27). Such an association has been recently found in the bacterium Chlamydia (28) where an SWI-like module is fused to the type I topoisomerase.

	ACKNOWLEDGEMENTS

We are indebted to Mark Dillingham and Dale B. Wigley for their help in the helicase assays and their generous gift of PcrA helicase, to Guiseppe Villani for Y-shaped helicase substrate, and C. Jaxel and M. Nadal for fruitful discussions.

	FOOTNOTES

^* This work was supported by CNRS Grant UMR 8621, Association pour la Recherche sur le Cancer Grant ARC 6198, and Action Microbiologie du Ministère de la Recherche.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: CRC Nucleic Acid Structure Research Group, Dept. of Biochemistry, University of Dundee, MSI/WTB Complex, Dow St., Dundee DD1 5EH, UK.

^§ To whom correspondence should be addressed. Tel.: 33 1 6915 6216; Fax: 33 1 6915 7296; E-mail: duguet@igmors.u-psud.fr.

Published, JBC Papers in Press, March 27, 2000, DOI 10.1074/jbc.M910091199

	ABBREVIATIONS

The abbreviations used are: GST, glutathione S-transferase; DTT, dithiothreitol; ADPPNP, 5'-adenylyl-,-imidodiphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Valenti, G. Perugino, A. D'Amaro, A. Cacace, A. Napoli, M. Rossi, and M. Ciaramella Dissection of reverse gyrase activities: insight into the evolution of a thermostable molecular machine Nucleic Acids Res., July 9, 2008; (2008) gkn418v1. [Abstract] [Full Text] [PDF]

				M. Wagner, G. Price, and R. Rothstein The Absence of Top3 Reveals an Interaction Between the Sgs1 and Pif1 DNA Helicases in Saccharomyces cerevisiae Genetics, October 1, 2006; 174(2): 555 - 573. [Abstract] [Full Text] [PDF]

				T.-s. Hsieh and J. L. Plank Reverse Gyrase Functions as a DNA Renaturase: ANNEALING OF COMPLEMENTARY SINGLE-STRANDED CIRCLES AND POSITIVE SUPERCOILING OF A BUBBLE SUBSTRATE J. Biol. Chem., March 3, 2006; 281(9): 5640 - 5647. [Abstract] [Full Text] [PDF]

				A. Valenti, A. Napoli, M. C. Ferrara, M. Nadal, M. Rossi, and M. Ciaramella Selective degradation of reverse gyrase and DNA fragmentation induced by alkylating agent in the archaeon Sulfolobus solfataricus. Nucleic Acids Res., January 1, 2006; 34(7): 2098 - 2108. [Abstract] [Full Text] [PDF]

				T.-s. Hsieh and C. Capp Nucleotide- and Stoichiometry-dependent DNA Supercoiling by Reverse Gyrase J. Biol. Chem., May 27, 2005; 280(21): 20467 - 20475. [Abstract] [Full Text] [PDF]

				A. Napoli, A. Valenti, V. Salerno, M. Nadal, F. Garnier, M. Rossi, and M. Ciaramella Functional interaction of reverse gyrase with single-strand binding protein of the archaeon Sulfolobus Nucleic Acids Res., January 26, 2005; 33(2): 564 - 576. [Abstract] [Full Text] [PDF]

				A. Napoli, A. Valenti, V. Salerno, M. Nadal, F. Garnier, M. Rossi, and M. Ciaramella Reverse Gyrase Recruitment to DNA after UV Light Irradiation in Sulfolobus solfataricus J. Biol. Chem., August 6, 2004; 279(32): 33192 - 33198. [Abstract] [Full Text] [PDF]

				E. Shor, S. Gangloff, M. Wagner, J. Weinstein, G. Price, and R. Rothstein Mutations in Homologous Recombination Genes Rescue top3 Slow Growth in Saccharomyces cerevisiae Genetics, October 1, 2002; 162(2): 647 - 662. [Abstract] [Full Text] [PDF]

				A. C. Rodriguez Studies of a Positive Supercoiling Machine. NUCLEOTIDE HYDROLYSIS AND A MULTIFUNCTIONAL "LATCH" IN THE MECHANISM OF REVERSE GYRASE J. Biol. Chem., August 9, 2002; 277(33): 29865 - 29873. [Abstract] [Full Text] [PDF]

				K. Y. Kwan and J. C. Wang Mice lacking DNA topoisomerase IIIbeta develop to maturity but show a reduced mean lifespan PNAS, April 25, 2001; (2001) 101132498. [Abstract] [Full Text]

				R. J. Bennett, M.-F. Noirot-Gros, and J. C. Wang Interaction between Yeast Sgs1 Helicase and DNA Topoisomerase III J. Biol. Chem., August 25, 2000; 275(35): 26898 - 26905. [Abstract] [Full Text] [PDF]

				K. Y. Kwan and J. C. Wang Mice lacking DNA topoisomerase IIIbeta develop to maturity but show a reduced mean lifespan PNAS, May 8, 2001; 98(10): 5717 - 5721. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/26/19498    most recent M910091199v1 Alert me when this article is cited