SuZfoZobus shibatae - The Journal of Biological Chemistry

3 downloads 58 Views 5MB Size Report
All the purification steps were carried out at 6 “C. S. shibatae B12 frozen cells (20 g) were ...... olysis always occurred. 4) Ditirachium album proteinase K is ... particular, topoisomerase 11' of E. coli has lost the N-terminal part of the B subunit of ...
THEJOURNAL OF B I O ~ I CCHEMISTRY AL 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 7, Issue of February 16, pp. 5255-5263, 1994 Printed in U S A .

Purification and Characterizationof Reverse Gyrase from SuZfoZobus shibatae ITS PROTEOLYTIC PRODUCT APPEARSAS AN ATP-INDEPENDENT TOPOISOMERASE* (Received for publication, June 22, 1993, and in revised form, October 8, 1993)

Marc Nadal, Elisabeth Couderc, Michel Duguet, and Christine JaxelS From the Laboratoire d‘Enzvmolonie des Acides Nucldiques, Znstitut de GVndtique Microbiologie, batiment 400, Universite Paris Sud, 91405 Orsay Ceiex, Frince

In contrast, type I topoisomerases are subdivided into two Sulfolobus shibatae B12 is a thermophilic archaebacterium that contains an inducible virus named SWl. distinct families: the family of the E. coli topoisomerase I (o The viral DNA has been shown to be positively super- protein) (Wang, 1971) and thefamily of the eukaryotic nickingcoiled before encapsidation. We have previously puri- closing enzyme(Champoux and Dulbecco, 1972). The major fied an archaebacterial DNA topoisomerase from SuZ- differences between these two groups are the polarity of the foZobus crcidocazdarius DSM 639, reverse gyrase, likely enzyme linking to the DNA and the ability of the enzyme to relax positively supercoiled DNA. Thus, the o-liketopoisomerresponsible for this positive supercoiling reaction. In the end of the DNA, requires magnesium and order to studyan homogeneous system containing bothase is linked to 5’ is unable to relaxpositively supercoiled DNA. In contrast, the reversegyraseandoneof its preferentialsubstrate, SSVl DNA, we have purified this enzyme from S. shi- eukaryotic topoisomerase is linked to the 3’ end and relaxes batae. During the course of the purification, we havede- both negative and positive supercoiled DNA. tected another topoisomerase activity. In order to sepa-The enzymatic properties of gyrase and protein o suggested that in bacterial cells, the level ofDNA supercoiling is conrateandpurifythesetwotopoisomerases,wehave devised a new purification procedure. Purified S. shi- trolled by a balance between these two types of topoisomerases batae revers? gyrase is a 124-kDa monomer, with a Stokes (Menzel and Gellert, 1983; Drlica, 1984). of 6.2 S. It radius of 43 A and a sedimentation coefficient In thermophilic archaebacteria, a new DNA topoisomerase is able to perform a DNAreverse gyrationper se at 10m~ named “reverse gyrase” was first described in two different NaCl in aMg- and ATP-dependent manner. The other to- strains of Sulfolobus (Kikuchi and Asai, 1984; Forterre et al., poisomerasqis a monomer of about 40 kDa, with a Stokes 1985). This enzyme catalyzesthe formation of positively superof 4 S. This coiled DNA in the presence of ATP and magnesium. Surprisradius of25 A and a sedimentation coefficient additional topoisomerase activityis Mg-dependent and ingly, reverse gyrase, which is able to perform a reaction of ATP-independent and catalyzes only a relaxation reac- gyration per se,is a type I enzyme(Forterre et al., 1985; Nakasu tion of negatively supercoiledDNA at 150m~ NaCl. This and Kikuchi, 1985). The fact that Jaxel et al. (1989) showed new ATP-independent topoisomerase activity seems to that reverse gyrase is linked through the 5’ end of the DNA be a proteolysis product of reverse gyrase. suggested that it is related to the w-protein family. This was confirmed by the comparison of the deduced protein sequences (Confalonieri et al., 1993). More recently, reverse gyrase activity was detected in all DNA topoisomerases are ubiquitous enzymes that occupy a critical place in DNA metabolism (for reviews see Wang (1985, thermophilic archaebacteria tested (Collin et al., 1988; Sle1987) and Maxwell and Gellert (1987)). One of their actions is sarev, 1988; Bouthier dela Tour et al., 1990) and, surprisingly, to change and regulate the superhelical density of circular also in thermophilic eubacteria (Bouthier de la Tour et al., DNA. They act by introducing a transient single strand break, 19911, suggesting that thepresence of this enzyme is a charan for type I topoisomerase, and a transient double strand break, acteristic of life at high temperature rather than archaebacfor type I1 enzyme. In addition, type I1 enzyme requires the terial feature. By looking carefully at the dataobtained in theseorganisms, hydrolysis of ATP to realize a n enzymatic cycle. In Escherichia coli,the topoisomerase I1 (DNAgyrase)is able we noted major discrepancies between results. For example, to introduce negative supercoilsin closed circular DNA (Gellert reverse gyrase was described as more active either at low salt et al., 1976). In contrast, the eukaryotic enzyme does not ex- (our results) or at high salt concentration (Nakasu and Kikuhibit supercoiling activity but relaxes both negativeor positive chi, 1985; Slesarev, 1988). Moreover, the level of superhelical supercoils (Miller et al., 1981). Sincethe amino acid sequences density inthermophilic archaebacterial cells is not clear: Nadal et al. (1986) reported that the genome of the virus SSVl isoof some type I1 topoisomerases representative of the three urkingdoms (Eubacteria, Eukaryotes, andArchaebacteria) are lated from Sulfolobus shibatae exists in the positive superheI1 topoisomerases are closely lical state whereas Charbonnier et al. (1992)reported that available, it is clear that all type SSVl DNA and the plasmid pGT5 are rather relaxed. Finally, related. in addition to reverse gyrase, an ATP-independent topoisomerase activity (o-like) appeared in some thermophilic archaebac* This work was supported by Grants 554 and 1354 from Unit6 de terial(Slesarev et al., 1991) and thermophilic eubacterial Recherche associ6e. The costs of publication of this article were destrains (Bouthierde la Tour et al., 1991). This ATP-independent frayed in part by the payment of page charges. This article must there- activity seems present only in particular strains andwas not fore be hereby marked “advertisement”in accordance with 18 U.S.C. detected in Sulfolobus acidocaldarius cells (Nadal et al., 1988). Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed. “el.: 33-1-69-41-62- Thus, both the enzymatic characterization and the putative 05; Fax:33-1-69-41-72-96. role of topoisomerases in thermophilic organisms are not clear.

5255

Characterization of Reverse Gyrase from S . shibatae

5256

For these reasons, we have studied reverse gyrasefrom the Sulfolobus strain carrying the SSVl virus: S. shibatae (Yeats et al., 1982;Martin et al., 1984;Grogan et a l . , 1990).Now a homogeneous system is available in order to study both the topoisomerases activitiesin vitroand therole of these enzymes in vivo. In this article, we describe an efficient purification of reverse gyrase isolated from thermophilic organisms. This procedure may be useful in thecase of other hydrophobic proteins purification. During the course of the purification from S. shibatae strain, we detected another topoisomerase activity: a n activity of relaxation of negatively supercoiled DNA in a Mg-dependent and ATP-independent manner. Here we detail thephysical and enzymatic properties of these two proteins. We show that this new activitylikely results from a proteolytic cleavage of reverse gyrase. EXPERIMENTALPROCEDURES

Materials Plasmid pTZ18R was prepared as described by Duguet et al. (1983). ATP,NAD, and ethidium bromide were purchased from Boehringer (Mannheim, Germany). The protein markers used in this work were from Sigma, as well as polyminP,’DTT, protease inhibitors, phenylmethanesulfonyl fluoride, and single-stranded DNA Cellulose. PhenylSepharose, heparin-Sepharose, blue-Sepharose, and Sephadex G-150 were purchased from Pharmacia. Phosphocellulose P11 was from Whatman. SDS was fromServa, pure sucrose from British Drug House, and agarose (A37 NA) from Industrie Biologique Fraqaise. Proteinase K and the other chemicals were purchased from Merck.

Methods Cell Culture S. shibatae B12 was obtained from Dr. Zillig (Miinchen, Germany). The cells were grown at 75 “C in a 300-liter fermenter (Laboratoire $Extraction et de Fermentation, Centre National de la Recherche Scientifique, Gimvette, France) as described previouslyby Mirambeau et al. (1984).

Topoisomerase Assays The standard reaction mixture (20 pl) contained 50 IIIM Tris-HC1, pH 8.0,0.5 IIIM DTT, 0.5 IIIM EDTA, 10 IIIM MgC12, 0 (for ATP-independent topoisomerase)or 1m~ ATP (for reverse gyrase), 120 m~ NaC1,30 pg/ml bovine serum albumin, and 0.3 pg of pTZ18R. After addition of 2 pl of the fraction t o be assayed, the mixture was incubated for 30 min at 75 “C.The reaction was stopped by cooling the samples and by adding 1%SDS, 0.25 mg/ml bromphenol blue, and 15%sucrose. Dilutions of the various enzymatic fractions were performed in each case in the same buffer as the fraction tested, with the addition of 1mg/ml bovineserum albumin. Reaction products were analyzed by 1.2% agarose gel electrophoresis in TEP buffer (36 m~ Tris, 30 m~ NaH,P04, 1mM EDTA, pH 7.8) at 4.8 V/cm for 2 h 30 with buffer recirculation. They were stained with 2 p g / d ethidium bromide and destained with 1 m~ magnesium sulfate. Gelswere photographed under UV illumination at 312nm (transilluminator with cooling fan Spectroline) by using a model N432 CCD Camera coupled to a numerization card CAM8803 (Orkis, Les Milles, France) for Macintosh computer (Apple). The images were converted to digital pictures (256 gray levels) and conserved in the storage software “Opaline”(Orkis). For the quantitation, 1unit of enzyme was definedas the amount of protein required to relax 50% of the negatively supercoiled pTZ18R input under the standard assay conditions. Thus, we determined the amount of remained negatively supercoiled DNA. For that, the digitalized images of electrophoreses were analyzed with the software “Scan Analysis” (Biosoft,UK), on a Macintosh computer (Appleh2

Protein Determination Protein concentration was determined by using a Bio-Rad protein The abbreviations used are: polymin P, poly(ethy1enimine);DTT, dithiothreitol. In order to improve the copyof the photographic prints, we enhanced the contrast of digitalized pictures before printing.

assay according to the instructions of the manufacturer, with bovine serum albumin as a standard.

Polyacrylamide Gel Electrophoresis SDS-polyacrylamide gradient (520%) gel electrophoresis was performed accordingto Hames (19811, and proteins were revealed by silver staining (Oakley et al., 1980). Molecular size markers were myosine (205 m a ) , P-galactosidase (116 m a ) , phosphorylase b (97.4 m a ) , bovine serum albumin (66 m a ) , ovalbumin (45 m a ) , and carbonic anhydrase (29 m a ) .

Purification of lbpoisomerases Purification Bufers-Buffer A consisted of 50 IIIM NaH,PO,/ Na2HP0,, pH 7.0, 1m~ DTT, 1 IIIM EDTA. Buffer B consisted of buffer

A + 10%ethylene glycol. Buffer C consisted of buffer A + 60% ethylene glycol. Buffer D consisted of 25 IIIM NaH2POJNa2HP04,pH 7.0, 0.5IIIM DTT, 0.5 IIIM EDTA, 100 m~ NaCI. All the purification steps were carried out at 6 “C. S. shibatae B12 frozen cells (20 g) were disrupted by thawing in 300 mlof buffer A containing 1.2 M NH4CI, 1IIIM EGTA, 10 m~ sodium bisulfite, 0.5 m~ phenylmethanesulfonylfluoride, 1 mg/ml leupeptin, 2 mg/ml pepstatin A, 0.125pg/ml (2S,3R)-amino-2-hydroxy-4-(4-nitrophenyl)butanoyl-~leucine, and 2 pg/ml chymostatin. Cell lysis was achieved by homogenization with an Ultraturax (TP 18/2) apparatus. The resulting solution was centrifuged at 24,000 x g for 5 min. The pellet was resuspended in the same buffer (200 ml), homogenized,and centrifuged as indicated for the lysis. The two supernatants were pooled (fraction I), and polymin P was added to a final concentration of 0.36%. Aftergentle mixing during 30 min, the solution was centrifuged at 24,000 x g for 20 min (fraction 11). The supernatant was further clarified by ultracentrifugation at 90,000 x g for 45 min (fraction 111). Ammonium sulfate was added to fraction I11 to a final concentration of 70% saturation and centrifuged at 24,000 x g for 20min. The pellet was dissolvedwith ammonium sulfate a t 35% saturation in buffer A and centrifuged in the same conditions. The supernatant was diluted with 2.4 M NaCl in buffer A to give a final concentration of 0.8 M ammonium sulfate and 1.2 M NaC1. This solution was saved as fraction Tv.Fraction IV was loaded on a phenyl-Sepharose column (4.4 x 20 cm) equilibrated with 0.8 M ammonium sulfate and 1.2 M NaCl in buffer A. The column was first washed with 3 liters of the same buffer at a flow rate of 120 ml/h and then with 3 liters of 0.25 M NaCl in buffer A. It was further developed with a linear gradient of 6 6 0 % ethylene glycol (2 x 3 liters) inbuffer A containing 0.25 M NaCl and thenwashed with 1.6 liters of 60% ethylene glycol and 0.25 M NaCl in buffer A. Finally, the phenyl-Sepharose column was washedwith 1% Triton X-100 in buffer A (1liter). Purification of Reverse Gyrase-Active fractions were pooled, concentrated with polyethylene glycol 35000 and dialyzed against buffer A containing 10% ethylene glycol (buffer B), and referred to as fraction V. Fraction V was loaded onto a phosphocellulose column (4.4 x 16 cm) equilibrated with buffer B and washed with 1.8 liters of the same buffer at a flow rate of 120 mVh.The bound proteins were eluted with 1M NaCl in buffer B. Activefractions were pooled,concentrated with polyethylene glycol 35000, and dialyzed against buffer B containing 0.2 M NaCl (fraction VI). Fraction VI was applied onto a heparin-Sepharose column (2.2 x 10 c m ) equilibrated with 0.2 M NaCl in buffer B at a flow rate of25 mVh. The column was washedfirst with 400 ml of the same buffer, then with 400 ml of buffer A containing 60% ethylene glycol (buffer C) and 0.2 M NaCI, and developed with a lineargradient of 0.2-1 M NaCl(2 x 300 ml) in buffer C. Activefractions were pooled, dialyzedagainst buffer C, and saved as fraction VII. Fraction VI1 was applied onto a blue-Sepharose column (1.6 x 7.5 cm)equilibrated with buffer C at a flow rate of 12 ml/h. The column was first washed with 150 ml of the same buffer, then developed with a lineargradient of ( M . 8 M NaCl(2 x 115 ml) in buffer C. Active fractions were pooled and saved as fraction VIII. Fraction VI11 was concentrated and equilibrated with buffer A containing 0.6 M NaCl on anhniconmicroconcentrator (Centricon30).This fraction was loaded on a 5-20% sucrose gradient in buffer A containing 0.6 M NaCl and centrifuged at 210,000 x g for 40 h in aSW 41 Beckman rotor.Fractions of 400 pl were collected. Active fractions were pooled, concentrated, and finally equilibrated with buffer D (fraction M).

Partial Purification of the ATP-independent Topoisomerase Active fractions were pooled,diluted with buffer A containing 0.12 M NaCl and 85% ethylene glycol to give the final concentration of 0.2 M NaCl and 60% ethylene glycol, and referred to as fraction V’ (650ml, 97 mg of protein). Fraction V was loaded onto a heparin-Sepharose column (3.3 x 21 cm) equilibrated with 0.2 M NaCl in buffer C at a flow rate

Characterization of Reverse Gyrase from S. shibatae of 60 ml/h. The column was first washed with 1.5 liters of the same buffer, and the bound proteins were eluted with 1 M NaCl in buffer C. Active fractions were pooled, dialyzed against buffer C, and saved as fraction VI' (130 ml, 5.78 mg of protein). Fraction VI' was applied onto a blue-Sepharose column (2.6 x 22 cm) equilibrated with buffer C at a flow rate of 40 mlh. The column was first washed with 950 ml of the same buffer, then developed with a lineargradient of 0 . 8 M NaCl(2 x 700 ml) in buffer C. Active fractions were pooled and saved as fraction VII'. Fraction VII' was applied onto a single-stranded DNA Cellulose column (1.6 x 5 cm) equilibrated with buffer C at a flow rate of 12 ml/h. The column was first washed with 100 ml of the same buffer, then developed with a lineargradient of 0-1 M NaCl(2 x 120 ml) in buffer C. Active fractions were pooled and saved as fraction VIII'. Fraction VIII' was concentrated and equilibrated with buffer A containing 0.6 M NaCl on an Amicon microconcentrator (Centricon 30). This fraction was loaded on a 5-20% sucrose gradient in buffer A containing 0.6 M NaCl and centrifuged at 240,000 x g for 42h in SW a 41 Beckman rotor. Active fractions were pooled,concentrated, and finally equilibrated with buffer D (fraction E').

Determination of the Sedimentation Coeficient and the Stokes Radius The sedimentation value was determined by the procedure used for the laststep of purification accordingto Martin and Ames (1961).300 pl of protein samples were prepared in buffer B containing 0.2% Triton X-100. 15 pl of reverse gyrase fraction E, 42.5 p1 of the ATP-independent topoisomerase fraction VI', and protein standards were loaded independently onto a 5-20% sucrose gradient in buffer A containing 0.6 M NaCl and centrifuged at 230,000 x g for 40 h in aSW 41 Beckman rotor. The Stokes radius was determined by gel filtration through Sephadex G-150 (0.5 x 35.5-cm column) equilibrated with buffer B containing 0.6 M NaC1. Amixture of 40 plcontaining 0.2% Triton X-100 was loaded onto the column and eluted at a flow rate of 0.5 ml/h. For reverse gyrase, 10 pl of fraction M was used and 10 pl of the fraction VI' for the ATPindependent activity. The excluded volume vas determined by using blue dextran (Mr,2 x 109. For calculation of the Stokes radius, we used the equation of Laurent and Killander (1964). The protein standards used in these two experiments were as follows: catalase, 11.3 S, 52 A; alcohol dehydrogenase, 7.4 S, 46 A; bovine sealbumin, 4.3 S, 35 A ovalbumin, 3.6 S, 27 A; cytochrome c, 1.9 S, 17 A.

I

5257

+ Polvmin P

supernatant pellet

UI supematant pellet

Em supernatant

supematant pellet phenyl-Sephamse

fractions 30-4096 Ethylene Glycol /:asphe

cellulose

fractions 45-5596 Ethylene Glycol heparinSephmse

Exl Pparjn-Sepharose

Eyn

1 1 ux

blue-Sephamse

IWn

sucrose gradient

EYI: blue-Sephamse

1

EYE singlestranded DNA cellulose

1 1

Eyrm

sucrose gradient

EMI

Towlsomerase

FIG.1. Scheme of the different purification stages for S. shibatae reverse gyrase and ATP-independent topoisomerase. The chromatographies using heparin-Sepharose, blue-Sepharose, and single-stranded DNA cellulose werecarried out in the presence of 60% ethylene glycol.

NaCl as ionic strength along the purification stages. These conditions permitted the visualization of both DNA topoisomerase activities (Fig. 2C). In the case of the ATP-independent Supercoiled pTZ18R (40 pg) was submitted to electrophoresis 1.2% activity, it is essential to use high voltage during electrophoreagarose gel containing 22 pdml chloroquine in TBE (89 m~ Tris, 89 m~ sis in order to clearly separate pTZ18R supercoiled DNA subboric acid, 2 m~ EDTA, pH 8.0) buffer. The electrophoresis was performed at 1.9 V/cm for 40 h. After staining with ethidium bromide, strate and therelaxation DNA product obtained at 75 "C. Takinto account the difference between the incubation topoisomers, visualized under W illumination with low intensity, were ing cut out of the gel and electroeluted in a IBI unidirectional electroelutor temperature (75 "C) and our migration temperature (about (model UEA) for 30 min at 100 V in 0.5 x TBE buffer. Eluted DNA was 40 "C), we could predict that the topoisomers relaxed at 75 "C extracted with butanol and chloroform,and ethanol-precipitated. About appeared negatively supercoiled duringthe electrophoresis 2 pg each of the seven main topoisomers were recovered and controlled (see Fig. 5A and Jaxel et al. (1989)). on electrophoresis gel in the presence of chloroquine. The procedure to purify the reverse gyrase and the ATPDigestion of Reverse Gyrase by Proteinase K independent activity of S. shibatae was summarized in Fig. 1 For the Western blot analysis, 1pl(70 ng) of reverse gyrase (fraction and Table I . The first four steps of purification (until ammonium sulfate M)was incubated in buffer D (10 pl)with different amounts of proteinase K(O.1-5 pg) at 65 "C in the presence of 0.05% Triton X-100. Proteins were stage) were performed essentially as described for the S . acisubmitted to electrophoresis (Laemmli, 1970) on a 7% SDS-polyacryl- docaldarius enzyme (Nadal et al., 1988). Some modifications amide gel and electrotransferred on a nitrocellulose membrane accord- were introduced to improve the efficiency of lysis; only 20 g of ing to the method of Svoboda et al. (1985).The membrane was incubated S. shibatae were used, and thebuffer volumes were increased. during 1 h with immune serum against S.acidocaldarius DSM 639 reSince proteolysis is more important in S. shibatae strain, we verse gyrase (Nadal et al., 1988)at a dilution of 1/800and finally stained to decrease it. Durwith an immunodetection kit (Life Technologies, Inc.) using alkaline used additionalprotease inhibitors in order phosphatase. Biotinylated molecular mass markers are:phosphorylase ing phenyl-Sepharose chromatography, we characterized two b (97.4 m a ) , catalase (58.1 kDa), alcohol dehydrogenase (39.8 m a ) , topoisomerase activities: one ATP-dependent and one ATP-incarbonic anhydrase (29 m a ) , trypsin inhibitor (20.1 m a ) , lysozyme dependent. After loading fraction IV on a phenyl-Sepharose, (14.3 m a ) , P-galactosidase (116 kDa), and lactoperoxidase (85 kDa). the column was washed at a low ionic strength (0.25 M NaCl) to For the enzymatic activities, the same amount of reverse gyrase was allow the eliminationof numerous proteins. No significant tomixed with 7 pg of proteinase K in the same buffer. Then, the assays Along the folwere performed as described under "ropoisomeraseAssays"; 2 p1 of the poisomerase activity eluted during these stages. mixture were added at the indicated dilution to the standard reaction lowing 040% ethylene glycol linear gradient(shown in Fig. 2), mixture and incubated for different incubation time at 75 "C. the first activity eluted at 3040% and the second one at 4555% ethylene glycol. Fig. 2C describes the activities of the RESULTS various fractions recovered during elution by ethylene glycol. Comments on the Purifications-In order to avoid aggrega- Dilutions of each fraction allowed us to determine theposition tion or nuclease activity (Nadal et al., 1988), we used 120 mM of the two peaks (fractions 280 and 360). The first activity is Preparation of pTZ18R with a Unique Linking Number

Characterization of Reverse @rase from S. shibatae

5258

TABLE I Purification of S. shibatae reverse gyrase Fraction

step

Volume

Total

ml

Total protein mg

activity x

units

Specific activity x

unitslmg

I

Crude extract 520 NS" ND6 ND I1 Polymin P 2444 510 ND ND I11 Ultracentrifugation 2088 475 ND ND rv Ammonium sulfate 955 600 24 22,900' V Phenyl-Sepharose 52.60 675 213 11,200 VI Phosphocellulose 135 4.85 1078 5230 VI1 Heparin-Sepharose 1.20 150 1958 2350 VI11 Blue-Sepharose 0.46 60 3522 1620 0.70 0.05d 5480 274 M Sucrose gradient NS, not significant. ND, not determined. e This activity does not correspond to the reverse gyrase activity alone but tothe addition of ATP-independenttopoisomerase and reverse gyrase activities. Amount of proteins determined by amino acid composition.

characteristic of reverse gyrase, producing positive supercoiled DNA in anATP-dependent manner at the higher protein concentration (Fig. 2C, fractions 260, 280, and 300). Even at the higher concentration, no ATP-independent activity was detectable. The second activity relaxes the negatively supercoiled substrate at 75 "C in an ATP-independent manner (Fig. 2C, fractions 340,360, and 380). The patternobtained with diluted fractions is characteristic of this enzyme and is identical with or without ATP. It is notable that this enzyme is not inhibited positively supercoiled DNA (data by ATPand is not able to relax not shown). Therefore, the reaction product obtained in the presence of ATP was always the addition of an ATP-independent activity plus theATP-dependent one due to reverse gyrase. Fraction V (pool reverse gyrase) represented5.5% and fraction V' (pool ATP-independent activity) 10.3%of the proteins initially bound to phenyl-Sepharose. Moreover, the patternof prolargely differentfrom those of the teins presentat this step was first steps (Fig. 3, A and B; compare lanes I , 11, 111,and N with lanes V and V). For the well known reverse gyrase, it is possible to distinguish it in thefractions 260-320 and in thepool phenyl-Sepharose (Figs. 2B and 3 A ) . Phenyl-Sepharose constitutes the more efficient chromatography to separate the two topoisomerase activities. During preliminary attempts to purify S. shibatae enzymes, we performed the nextchromatographic stages in thepresence of 10% ethylene glycol as previously described (Nadal et al., 1988). We then observed that for all the Sepharose supports, topoisomerase activities were not eluted even at high ionic strength. The only possibility was to increase ethylene glycol nonspecific up to60%.These conditionsallowed us to eliminate interactions between the Sepharose matrix and theproteins. Purification of Reverse Gyrase-Fraction V was loaded at low ionic strength on a phosphocellulose column. This resulted in the elimination of 90% of the proteins loaded (Table I), while reverse gyrase remainedbound and was further eluted with1 M NaCl (fractionVI). One part ofATP-dependent topoisomerase activity was always detected in thenon-adsorbed fraction. The properties of this fraction were different from those of the reverse gyrase (not shown). In fraction VI, reverse gyrase represented a significant part of the protein content(Fig. 3A, compare lanes Vand VI).Fraction VI was then loaded at 0.2 M NaCl at 0.5 M on heparin-Sepharose, and reverse gyrase was eluted NaCl (fraction VII). About 75% of the loaded proteins were eliminated during thischromatography (Table I). At this point, the pattern of remained proteins was clearly simplified (Fig. 3A, compare lanes VI and VZZ).The last chromatography was achieved on blue-Sepharose. 60%of the protein content of fraction VI1 were eliminated in the flow-through, while reverse gyrase eluted at 0.44 M NaCl (fraction VIII). During preliminary attempts to purify these enzymes, we were able to sepa-

rate the reverse gyrasefrom the ATP-independent activity by using blue-Sepharose chromatography.In fraction VIII,reverse gyrase represented the major protein (more than 50% of the proteins stained)(Fig. 3A, lane VIII).Along the phosphocellulose, heparin- and blue-Sepharose chromatographies, almost no ATP-independent activity was detected. In orderto improve the reverse gyrasepurity, sedimentation in a sucrose gradient was carried out. Reverse gyrase, sedimenting in the middle of By scanning the gradient, was totally purified (fraction E). lane M of the polyacrylamide gel (Fig. 3 A ) , we determined that reverse gyrase representedmore than 95% of the stained proteins. Isolated from S. shibatae B12, reverse gyrase appeared, by SDS-PAGE, as a sin5le polypeptide of 124 kDa (Fig. 3 A ) . The Stokes radius is43 A and the sedimentationcoefficient about 6.2 S. According to Siege1 and Monty (1966), these dataallowed us to calculate a frictional coefficient flfo of 1.3. This result suggests a globular protein. We noted that, as previously described (Bouthier de laTour et al., 1990), the antibodies raised against S. acidocaldarius reverse gyrase cross-react, although with a lower extent, with the protein isolated fromthe S. shibatae strain (Fig. 7, lane 6). Partial Purification of the ATP-independent TopoisomeraseFraction V' was applied on a heparin-Sepharose column at 0.2 M NaCl. This resulted in the elimination of 94%of the proteins loaded, while the ATP-independent activity remained bound and was further elutedwith 1 M NaCl (fraction VI'). In a previous purification, the ATP-independent activity was eluted at 0.64 M NaCl. In fraction VI', the pattern of remained proteins was simplified (Fig. 3B, compare lanes V and VI' ). Fraction VI' was loaded on blue-Sepharose at 0 M NaCl and ATP-independent enzyme was eluted at 0.3 M NaCl (fraction VII'). Some ATP-independent activity did not bind to blue-Sepharose. At this point, a few bands remained on the SDS-polyacrylamide gel (Fig. 3B, lane VII').The last chromatography was achieved onto single-stranded DNA cellulose as described by Depew et al. (1978) for protein w. Fraction VII' was loaded at 0 M NaC1, and the activity eluted at 0.4 M NaCl (fraction VIII'). Once again, some activity was found in non-adsorbed fractionsand no more activity was obtainedby increasing theionic strength of elution (up to 4.5 M NaC1). The fraction IX', obtained after sedimentation in a sucrose gradient, was characterized by a particularly weak activity, and we were not able to determine theprotein composition of this preparation by SDS-PAGE analysis. Nevertheless, during this stage, we removed proteins of about 120 and 100 kDa, which sedimented at the bottom of the gradient, anda smaller protein, which sedimented at the top (Fig. 3B, lane VIZI' 1. The only band that could be correlated with ATP-independent activity was the40-kDa protein. In addition, this correlation has

Characterization of Reverse Gyrase from S. shibatae

5259

A

E

ze4 d 0

I

M 260270280290300310320330340

I I

I

I

I

I

\

M 330360370380400

I

I

I

I

I

I

I

I

I

"

FIG. 2. Phenyl-Sepharose column chromatography. A, elution profile of proteins. Fractions of 15 ml were collected. 8,pooled reverse gyrase fractions; 0,pooled ATP-independent topoisomerase fractions. B , SDS-polyacrylamidegel electrophoresis of the eluted fractions: M,molecular mass markers; lunes 260400,analysis of 40 pl of fractions 260-400. C, activities of the fractions visualized by agarose gel electrophoresis of the DNA. The followine dilutions of each fraction were used: 1. 1/4. 1/16. 1/256. were performed in the presence (upper part) or absence . . 1/64. Incubations (lower part) i f ATP. ReZ, relaxed DNA; sc, supercoiled DNA.

been observed during various chromatographies issued from independent purifications. The hypothesis of a size of about 40 kDa is consistent with the Stokes radius (25A) and thesedimentation coefficient (4 S). These values agree with a globular monomeric enzyme (flfo = 1.1). Moreover this hypothesis is confirmed by the Western blot analysis of fraction VI', since a polypeptide of about 40 kDa was recognized by antibodies raised against S. acidocaldarius reverse gyrase (Fig. 7, lane 1 ). In conclusion, we noted a significant loss of activity during the last purification stages (from the blue-Sepharose chromatography). This effect was observed throughout the previous performed purifications. Several chromatographies on phosphocellulose gave rise to non-reproducibleelutions (not shown). Thus, the enzyme behavior was not reproducible and suggested a phenomenon of proteolysis.

Enzymatic Characteristics of Both Topoisomerases-As shown in Fig. 4, reverse gyrase and itsATP-independent activity counterpart changed the linking number of the DNAby steps of one. Consequently, they are type I topoisomerases. By bidimensional gel electrophoresis (Fig. 5),we studied the extent of the reaction catalyzed by these two enzymes.Panel A, relaxation with Mg(I1) and without ATP, clearly shows that the ATP-independent activity catalyzes only the relaxation of the DNA. In contrast, as shown in panel C , the purified reverse gyrase introduces a number of positive superhelical turns in DNA in an ATP- and magnesium-dependent manner (panel B shows relaxed DNA a t 37 "C that allowed us to visualize the negative and positive branches of the two-dimensional gelarch). As we have reported for reverse gyrase purified from S. acidocaldarius (Nadal et al., 1988),the enzyme isolated from S. Shibatae catalyzes positive supercoiling of the DNA at high temperature andis able to perform a reaction of gyrationper se.

Characterization of Reverse e r a s e from S. shibatae

5260

A

B I

I1

111

I

I

I

v

IV

I

I

VI

I

VI1 VI11

I

I

IX

I

M

I

I

V’ VI’

I I

VII’ VIII’

I

I

45 -

FIG.3. SDSpolyacrylamide gel electrophoresisof the fractions obtainedat each stageof topoisomerases purification.A and B, silver staining of the gels (in order to detect low amount of contaminants, we over-stained the gel). A, reverse gyrase purification. Lune I , 5 pl of crude extract;lane II, 5 pl of polymin P fraction; lane III, 5 pl of ultracentrifugation stage;lane N , 6.5 pl of ammonium sulfate stage;lane V,40 pl of phenyl-Sepharose pool; lane V I , 40 pl of phosphocellulose step; lane V I I , 150 pl of heparin-Sepharose pool; lane VIII, 60 p1 of blue-Sepharose pool; lane E,5 p1 of sucrose gradient pool. E , ATP-independent topoisomerase purification. M , molecular mass markers; lane V’, 40 pl of phenyl-Sepharose pool; lane VI’,60 pl of heparin-Sepharose step; lane VU’,200 pl ofblue-Sepharosepool; lane VIII’,200 p1 of single-stranded DNA cellulose pool. 1

2

3

4

5

6

7

Znd Dim

7 FII

4

scg

Re1

=

=

SC@

L

sc

”t

J

RG.4. Variations in thelinking number of the DNA. DNA with a unique linking number was incubated without enzyme (lane 1 ) or with reverse gyrase a t 120 m NaCl (lane 4 ) or ATP-independent activity at 150 m NaCl (lane 7) in the standardreaction mixture. Controls pTZ18R were incubated without (lane 2) or with 16 (lane 3) and 2 (lane 5 ) units of reverse gyrase. Lune 6 is control pTZ18R incubated with ATP-independent activity. Rel, relaxed DNA, sc, supercoiled DNA, FII, form I1 DNA.

In order to make it efficiently, the enzyme requires lessthan 10 m NaCI. In theseconditions, the reaction is processive. At 120 m NaCI, the enzyme is less efficient and works in a distributive manner (datanot shown). In contrast, as shown in Fig. 6, the ATP-independent activity absolutely requires magnesium is active only a t high salt concentrations (part B ) . (part A) and Stability oftheProteinsS. shibatae topoisomerases are less stable than S. acidocaldarius DSM 639 reverse gyrase. We noted a degradation of partially purified fractions of S. shibatae reverse gyrase (fractions V-VIII) after several months a t 4 “C. Purified fraction M was stable more than 1year at 4 “C. Nevertheless, transfer of enzyme from one tube to another decreased its activity. We supposed that reverse gyrase binds “nonspecifically” to the surface of the tube. This binding was prevented by 0.05% Triton X-100.The same phenomenon was more prominent for the ATP-independent topoisomerase, probably due to its higher hydrophobicity. This might in part explain the very low amount of enzyme recovered at theend of its purification. In order to avoid enzymatic interferences, we used Triton only when we have previously verified that the addition of Triton does not change the properties of enzymes. Amino AcidComposition-The amino acid compositions of S. shibatae reverse gyrase fromtwo independent purifications were compiled in Table 11. As for the enzyme isolated from S. acidocaldarius, we noted the absence of cysteine residues

A B C WG. 5.Tho-dimensional gel analysisof the topological conversions catalyzed by reverse gyrase and ATP-independent activity. We used a large minigel (13.5x 11.5 cm) in order to perform all the analyses on the same gel. The first dimension was carried out in TEP buffer for 4 h a t 3.5 V/cm. The gel was soaked for 2 h in TEP buffer added with 5 pg/ml of chloroquine (freshly prepared) and thenelectrophoresed for 14 h at 0.7 Vkm. The gel was stained as described under “Methods.” A , ATP-independent activity in the standardreaction mixture except that we used 150 m NaCl and 0.6pg of DNA substrate. E , Relaxed DNA at 37 “C prepared with Calf thymus topoisomerase I incubated without magnesium. C, reverse gyrase activity in the standard reaction mixture (1pl of fraction E,70 ng), except that we used 0.6pg of DNA.Rel, relaxed DNA, sc-, negatively supercoiled DNA, sc+, positively supercoiled DNA, *, form I1 DNA.

(Nadal et al., 1988).Nevertheless, it is possible that theenzyme contains an extremely low content of cysteine. In thesame way, we noted the presence of numerous hydrophobic residues such as leucine or isoleucine. This high level of hydrophobic residues is in a good agreement with the hydrophobic properties of the enzyme (see “Comments on the Purifications” and “Stability of the Proteins” under “Results”). Activity of Proteolyzed Reverse GyraseThe fact that the polypeptide of 40 kDa present in theATP-independent fraction (fraction VI’) was recognized by antibodies raised against reverse gyrase prompted us to study the relationship between these two enzymes. Thus, we tried to degrade purified reverse gyrase by using various proteases. We used a nondenaturing reaction mixture to be able to check the potential activity of proteolyzed reverse gyrase. V8 protease was unable to degrade the enzyme a t 37 “C (data not shown). The possibility that reverse gyrase (a thermophilic enzyme) could be more sensitive to proteolysis a t higher temperature prompted us to check its degradation by proteinase K. Indeed, this protease is able to work a t 65 “C. Onthe Western blot shown in Fig. 7, wecan see the pattern of reverse gyrase proteolysis obtained with differ-

Characterization of Reverse e r a s e from S. shibatae

A

1

2

3

4

5

6

7

8

5261

1

2

3

4

5

6

I

I

I

I

I

I

!-

- 116

- 91.4 - 85

B

I

2

3

4

5

6

7

x

9

FIG.6. Enzymatic properties of the ATP-independent topeisomerase.We used ATP-independent activityin kinetic conditions. A, requirement for magnesium. Standard reaction mixture without (lane I ) or with enzyme with,respectively, 0 , 2 , 4 , 6 , 8 , 10, and 12mM MgClz (lanes 2-8). B, requirement for sodium chloride. Standard reactionmixture with enzyme(lanes 1-11) with, respectively, 10,30,50,70,90, 110, 130, 150, 170, 190, and 210 m~ NaCl. Rel, relaxed DNA, sc, supercoiled DNA. TABLE I1 Amino acid composition (%) of S. shibatae reverse gyrase An aliquot of fraction M was hydrolyzed with 6M HCl, 0.25% phenol, and 0.02% mercaptoethanol at 110 "C during 24 h. The amino acid mixture was analyzed by a Beckman 7003 system according to the recommendations of the manufacturer. Residue

90

ASX

10.2 5.15 9.15 12.15 3.25 14.35 4.6

Thr Ser Glx Pro GlY Ala CYS Val Met Is0 Leu

Tyr Phe His LYS Arg

I

1011

0

5.25 10.05 3.05 3.3 1.35 5.4 5.1

ent proteaseconcentrations after 30 min of incubation at 65 "C. At this temperature, the degradation was very efficient. Depending on the extentof proteolysis, first we observed a polypeptide of about 60 kDa, andsecond we noticed the accumulation of a polypeptide of about 40 kDa (Fig. 7, lanes 5 and 4 ) . Products of lower molecular mass were observed with highest amount of proteinase K (Fig. 7, lanes 2 and 3 1. This observation was confirmed by gel filtration cpromatoqaphies because the Stokes radius changes from 32 A to 18 A depending on the extent of proteolysis (data not shown). In order to check the activity of proteolyzed reverse gyrase, we performed a kinetic of proteolysis at 65 "C, followed by a n incubation at 75 "C with the DNA (data not shown). As expected, reverse gyrase activity decreased as a function of time and we detected a DNA relaxation activity in anATP-independent process. This activity was detectable after only 30 min of incubation time at 75 "C in thepresence of protease. We checked the eventualcontamination of reverse gyrase by

- 39.8

FIG.7. Western blot analysisof S. ehibotae reverse gyraseand ATP-independent topoisomerase.Lane 1, 20 p1 of the fraction VI' (ATP-independent enzyme eluted from heparin-Sepharose chromatography); lanes 2-6, 1 p1 of the fraction M (reverse gyrase of sucrose gradient pool) digested by 5 pg (lane Z ) , 2.5 pg (lane 3), 1pg (lane 4 ) ,0.1 pg (lane 51,and 0 pg (lane 6 ) of proteinase K at 65 "C during 30 min. The arrow shows nonspecific staining dueto Laemmli's buffer.

the ATP-independent enzyme (Fig. 8A). This contamination might be due to a degradation or a n oxidation of the enzyme in the storage tube.More certainly, it was the result of the stoichiometric activityof reverse gyraseas previously described by Jaxel et al. (1989).Under the sameexperimental conditions, a 30-min incubation a t 75 "C allowed an high reverse gyrase activity (activeup to1/1000 dilution, Fig. 8 A , lane 5)and a low ATP-independent one (active up to 1/80 dilution, Fig. 8 A , lane 9).In terms of supercoiled DNA disappearance, the ratio reverse gyrase/ATP-independent activities was about 15. The effect of proteinase K on the ATP-independent activity is shown in Fig. 8 ( B and C).After only 5 min at 75 "C in the absence ofATP, the incubation without protease resulted in the appearance of form I1 DNA (Fig. 8B, upper part, lane 11, corresponding to the stabilization of cleavable complex at a stoichiometric ratio reverse gyrase/DNA. Comparison of the upper and lower parts in Fig. 8B indicated that proteinase K produced a very efficient ATP-independent activity after 5 min of incubation (active up to11320 dilution; Fig. 8B,lowerpart, lane 6). Moreover, this activity considerably increased after 20 min of incubation (Fig. 8C, lower part ). If we looked at the reverse gyrase activityfor the corresponding time of proteolysis, the activity seemed to be slightly decreased after 5 min of incubation (data not shown). The inhibitory effect was clear after 20 min of proteolysis, particularly for the positive supercoiling activity (Fig. 80,compare upper and lowerparts, lane 1 ). The activity of positive DNA supercoiling is the only one to be specific of reverse gyrase. In terms of DNA relaxation, theATP-dependent process is replaced by an ATPindependent process as a function of proteolysis. All these experimentsstrongly suggested that, in theS. shibatae strain, the ATP-independent activity is the result of reverse gyrase proteolysis. DISCUSSION

This article describes a new purification procedure that allowed us to obtain highly purified reverse gyrase. For this, we prevented nonspecific hydrophobic interactions of the protein with the Sepharose matrix by adding 60%ethylene glycol during chromatography procedures. These conditions are likely extendable to thepurification of any hydrophobic protein. Reverse gyrase isolated from S. acidocaldarius (Nadal et al., 1988)or from S. shibatae presented essentially the same physical and enzymatic properties. Their amino acid compositions

Characterization of Reverse Gyrase from S. shibatae

5262

A

B

C

1

2

3

I-

I

I

4

5

I

61

I

2

D 3

I

4

I

5

I

6

I

I

I

1

2

3

4

5

6

' UI ,

FIG.8 . Activity of proteolyzed reverse gyrase. Activity of reverse gyrase (fraction E) visualized by agarose gel electrophoresis of the DNA. A, the enzyme was incubated at 75 "C for 30 min in the presence (lanes 1 - 5 ) and in the absence (lanes 6-11)ofATP. The following dilutions of the fraction were used: 1/200 (lane1 ), 1/400 (lane 2 ) , 1/600 (lane3 ) , 1/800 (lane41, l/lOOO (lane 5),l/10 (lane 6), 1/20 (lane 7), 1/40(lane 8).1/80 (lane 9),1/160 (lane l o ) ,and 11320 (lane 11 1. B , the enzyme was incubated at 75 "C for 5 min without ATP in the absence (upperpart)and in the presence (lower part) of proteinase K. The following dilutions of the fraction were used: 1/10 (lane 1 ), 1/20 (lane 2 ) , 1/40 (lane 3 ), 1/80 (lane4 ) , 1/160 (lane 5),and 1/320 (lane 6 ) .C , same assays as in B except that theincubation time was 20 min. D, the enzyme was incubated at 75 "C for 20 min with ATP in the absence (upper part) and in the presence (lower part) of proteinase K. The following dilutions of the fraction were used: 1/10 (lane 1 ), 1/40 (lane 2), l/160 (lane 3 ) , l/640 (lane4 ) , 1/2560 (lane 5),and 1/10240 (lane 6 ) .Rel, relaxed DNA sc, supercoiled DNA; FZZ, form I1 DNA.

are highly similar andS. shibatae reverse gyrase isrecognized by the antibodies raised against S. acidocaldarius enzyme. In addition, the molecular mass of reverse gyrase isolated from S. shibatae is also approximately the same as that of S. acidocaldarius enzyme. Moreover, reverse gyrases from other bacterial strainsalso have the same size (Nakasu andKikuchi, 1985; Slesarev, 1988; Bouthier de la Tour et al., 1990). For the S. acidocaldarius enzyme, i t is interesting to note a difference between the molecular mass deduced from the sequence (143 kDa) and thatof the purified enzyme (128 kDa). This value is found reproducibly, and the difference is not the resultof partial proteolysis but maybe due to the amino acid composition of thermophilic enzymes. In our hands, the optimal DNA positive supercoiling catalyzed by reverse gyrase is always obtained at low salt concentration. The difference between our results and the results obtained by Nakasu and Kikuchi (1985) and Slesarev (1988) may be due to thepresence of spermidine both in their purification buffers and in their enzymatic assays. Spermidine is known to interactwith DNA at low temperature (37 "C) and to inhibit d i k e topoisomerases activities (Srivenugopaland Morris, 1985). Thus, it is possible that spermidine interferes with reverse gyraseactivity. In addition, at high salt concentration, it is difficult to distinguish between the two enzymes. Even in the presence ofATP, the ATP-independent enzyme can relax DNA and it then modifies the starting substrategiven to reverse gyrase. Consequently, reverse gyraseactivity is overestimated. During reverse gyrase purification, we detected a n ATP-independent topoisomerase activity. For the following reasons, we suspected that this was a degradation product. 1)ATPindependent activity and reverse gyrase are immunologically related. 2) The absence of reproducibility of chromatographies

suggests that the ATP-independent activity corresponds to a mixture of polypeptides, of which the major form is about 40 kDa. 3) During the attempts of purification of reverse gyrase from S. acidocaldarius (Forterre et al., 1985) we also found ATP-independent activity. Since we used phenyl-Sepharose as the first chromatographic stage, we maintained a high ionic strength during the first stages of purification, which abolished proteolysis (Nadal et al., 1988). Consequently, we failed to detect any ATP-independent activity in S. acidocaldarius. We used the same purification conditions for reverse gyrase of S. shibatae and so considerably limited proteolysis; however, in spite of the useof additional protease inhibitors, partialproteolysis always occurred. 4) Ditirachium album proteinase K is able to degrade S. shibatae reverse gyrase in vitro, giving rise to a n ATP-independent activity. All these results are consistent with a degradation of reverse gyrase resulting in a n ATP-independent activity. That is why we did not presenta precise quantitation of this activity during purification. This kindof protein degradation is a reminiscence of the initial purifications of most of the topoisomerases. In particular, topoisomerase 11' of E. coli has lost the N-terminal part of the B subunit of gyrase. Consequently, the ATP-motor was removed and thegyration ability was abolished (Brown et al., 19791, as in our ATP-independent activity (reverse gyrase prime). The small size of the ATP-independent activity (40 kDa) is consistent with a DNA nicking-closing activity, since the molecular mass of topoisomerases isolated from viruses is in the same range (Shuman and Moss, 1987; Upton et al., 1990). Due to proteolysis, we did not obtainenough homogeneous polypeptide carrying the ATP-independent activity to perform microsequencing. Nevertheless, by looking at the deduced sequence of reversegyrase from S. acidocaldarius (Confalonieri et al.,

Characterization of Reverse Gyrase from S. shibatae 1993),we can conclude that our nicking-closing activity is carried by the C-terminal part of the protein. Indeed, the putative tyrosine involved in theactive site islocalized in this part of the topoisomerase. Our results allowed us to propose that the reverse gyrase contains two domains; one of them is a classical topoisomerase I (the ATP-independent activity), and the other one is an ATP-motor responsible for the extent of the reaction. This second domain could correspond to the putative “helicase” domain as proposed by Confalonieri et al. (1993). What is thebiological significanceof this proteolytic product of reverse gyrase? This activity is not dependent on energy, and its optimal enzymatic conditions are totally different from those of reverse gyrase. For instance, the ATP-independent topoisomerase needs high salt concentration to be active. It should be mentioned that non-degraded reverse gyrase is unable to relax DNAwithout ATP. These two enzymes are working in the same direction; the ATP-independent activity increases the linking number of DNA up to that of relaxed form, and reverse gyrase generates positively supercoiledDNA. It is possible that post-translational regulation of reverse gyrase, like proteolysis, modulates the level of superhelical density of DNA in the cell. On another hand, Slesarev et al. (1991) described the purification of an ATP-independent topoisomerase activity from D. amylolyticus and proposed to name this enzyme topoisomerase 111. In addition, Bouthier de la Tour et al. (1993) described the purification of an ATP-independent topoisomerase I from a thermophilic eubacterium. Since these results were obtained in different bacterial cells that contain reverse gyrase, we propose two hypotheses. 1)Two distinct proteins, reverse gyrase and ATP-independenttopoisomerase, are associated with two genes (or two transcripts of the same gene), and the amount of each enzyme depends on a lot of parameters such as growth conditions. 2) The ATP-independent activity is the resultof the degradation of reverse gyrase. These two ways of life may coexist in the cell. This work would allow us to better characterize the enzymatic properties of reverse gyrase, i.e. nicking-closing, ATPase, and gyration-dissociated activities. Since we used S. shibatae, the strainharboring the SSVl virus, it is possible to investigate the role of topoisomerases in vivo. The positively supercoiled state of SSVl DNA strongly suggests that thisDNA is a natural substrate for reverse gyrase. The determination of catalytic sites of reverse gyrase on the genome of the SSVl virus of S. shibatae is under investigation. Acknowledgments-We thank M. Thomas (Laboratoire #Extraction et de Fermentation du CentreNational de la Recherche Scientifique,A. Escaut, Giflvette, France) for growth of S.shibatae. Amino acid composition was performed by M. Hattab and L. Camoin (Institut Cochin de

5263

G6nBtique MolBculaire, Paris, France). A. Capurro and collaborators (Orkis) continuously support our efforts for better quantifications and analyses of our electrophoreses images. We thank D. Thevenet, C. Temple, and F. Willaume for technical assistance and P. Mollat and C. Cullin for critical reading of the manuscript. REFERENCES Bouthier de la Tour, C., Portemer, C., Nadal, M., Stetter, K., Forterre, P., and Duguet, M. (1990) J. Bacteriol. 172,68034808 Bouthierdela Tour, C.,Portemer, C., Huber, R., Forterre, P., andDuguet, M. (1991) J. Bacteriol. 173, 3921-3923 Bouthier dela Tour, C., Portemer, C., Forterre, P, Huber, R., and Duguet,M. (1993) Biochim. Biophys. Acta, 1216,213-220 Brown, P. O.,Peebles, C. L., and Cozzarelli, N. R. (1979) P m . Natl. Acad. Sci. U.S. A. 76,6110-6114 Champoux, J. J., and Dulbecco, R. (1972) P m . Natl. Acad. Sci. U.S. A. 69, 143 Charbonnier, F., Erauso, G . , Barbeyron, T., F’rieur, D., and Forterre, P. (1992) J. Bacterid. 174, 61034108 Collin, R. G . , Morgan, H. W., Musgrave, D.R., and Daniel, R. M. (1988) FEMS Microbiol. Lett. 66, 235-240 Confalonieri, F., Elie, C., Nadal, M., Bouthier de la Tour, C., Forterre, P., and Duguet, M. (1993)Proc.Natl. Acad. Sci. U.S. A. 90,47534757 Depew, R. E., Liu, L. F., and Wang, J. C. (1978) J . Biol. Chem. 263, 511-518 Drlica, K. (1984) Microbiol. Reu. 48, 273-289 Duguet, M., Lavenot, C., Harper, F., Mirambeau, G., and deRecondo, A,”. (1983) Nucleic Acids Res. 11, 1059-1075 Forterre, P., Mirambeau, G.,Jaxel, C., Nadal, M., and Duguet,M. (1985) EMBO J . 4,2123-2128 Gellert, M., Mizuuchi, K., ODea, M. H., and Nash, H.A. (1976) Proc. Natl. Acad. Sci. U.S. A. 73, 3872-3876 Grogan, D., Palm, P., and Zillig, W. (1990)Arch. Microbiol. 154,594-599 Hames, B. D. (1981)Gel Electrophoresis 0fPmteins:APracticalApproach. (Hames, B. D., and Rickwood, D, eds) pp. 71-76, IRL, Oxford Jaxel, C., Nadal, M., Mirambeau, G., Forterre, P., Takahashi, M., and Duguet, M. (1989) EMBO J. 8,31354139 Kikuchi, A., and Asai, K. (1984) Nature 309,677481 Laemmli, U.K. (1970)Nature 227,680485 Laurent, T. C., and Killander, J. (1964) J. Chromatogr 14, 317 Martin, A., Yeats, S . , Janekovic, D., Reiter, W.-D., Aicher, W., and Zillig, W. (1984) EMBO J. 3,2165-2168 Martin, R. G., and Ames, B. N. (1961)J. Biol. Chem. 236. 1372-1379 Maxwell, A., and Gellert,M. (1987) A d u . Protein Chem. 38,69-105 Menzel, R., and Gellert, M. (1983) Cell 34, 105113 Miller, K. G., Liu, L. F., and Englund, P. T. (1981)J. Biol. Chem. 266,9334-9339 Mirambeau, G., Duguet, M., and Forterre, P. (1984)J . Mol. Biol. 179, 559-563 Nadal, M., Mirambeau, G., Forterre, P., Reiter, W.-D., and Duguet, M. (1986) Nature 321, 256-258 Nadal, M., Jaxel, C., Portemer, C., Forterre, P., Mirambeau, G., and Duguet, M. (1988) Biochemistry 27,9102-9108 Nakasu, S., and Kikuchi, A. (1985) EMBO J. 4, 27062710 Oakley, B. R., Kirsch, D. R., and Moms,R. N. (1980) Anal. Biochem. 106,361-363 Siegel, L. M., and Monty, K J. (1966)Biochim. Biophys. Acta 112, 346-362 Slesarev, A. (1988) Eur J. Biochem. 173,395399 Slesarev, A. I., Zaitzev, D. A,, Kopylov, V. M., Stetter, K. O., and Kozyavkin, S . A. (1991) J. B i d . Chem. 266,12321-12328 Shuman, S., and Moss, B. (1987) Proc. Natl. Acad. Sci. U.S. A. 64, 7478-7482 Srivenugopal, K. S., and Moms, D. R. (1985) Biochemistry 24,476W771 Svoboda, M., Meuris, S . , Robyn, L., and Christophe,J. (1985)Anal. Biochem. 161, 1&23 Upton, C., Opgenorth, A., Traktman, P., and McFadden, G. (1990) Virology 176, 439-447 Wang, J. C. (1971) J. Mol. Biol. 66, 523-533 Wang, J. C. (1985) Annu. Reu. Biochem. &4,66&697 Wang, J. C. (1987) Biochim. Biophys. Acta 909,l-9 Yeats, S . , McWilliam, P.,and Zillig, W. (1982) EMBO J . 1, 1035-1038