Purification and Characterization of DNA Topoisomerase V AN ENZYME FROM THE HYPERTHERMOPHILIC PROKARYOTE METIWVOPYRUS M D L E R I THAT RESEMBLES EUKARYOTIC TOPOISOMERASE I*

DNA topoisomerase V is a novel prokaryotic enzyme related to eukaryotic topoisomerase I. The enzyme is a type I DNA topoisomerase and is recognized by poly- clonal antibody against human topoisomerase I. We describe ita purification *om the hyperthermophilic methanogen Methanopyncs kandleri. The enzyme has high activity in crude extracts and is present in at least 1,600 copiedcell. Topoisomerase V migrates as a 110-kDa polypeptide in SDS-polyacrylamide gel electrophoresis and as a 142-kDa globular protein in gel filtration. It is active up to at least 100 "C on both positively and nega- tively supercoiled DNA and is not inhibited by single-stranded DNA The enzyme works from 1 to 680 PLM NaCl and up to 3.1 M potassium glutamate. It acta processively at low ionic strength and distributively at high NaCl or KC1 concentration. Magnesium is not required and does not stimulate the enzymatic activity. Under DNA dena- turing conditions, topoisomerase V catalyzes an unlinking

DNA topoisomerase V is a novel prokaryotic enzyme related to eukaryotic topoisomerase I. The enzyme is a type I DNA topoisomerase and is recognized by polyclonal antibody against human topoisomerase I. We describe ita purification *om the hyperthermophilic methanogen Methanopyncs kandleri. The enzyme has high activity in crude extracts and is present in at least 1,600 copiedcell. Topoisomerase V migrates as a 110-kDa polypeptide in SDS-polyacrylamide gel electrophoresis and as a 142-kDa globular protein in gel filtration. It is active up to at least 100 "C on both positively and negatively supercoiled DNA and is not inhibited by singlestranded DNA The enzyme works from 1 to 680 PLM NaCl and up to 3.1 M potassium glutamate. It acta processively at low ionic strength and distributively at high NaCl or KC1 concentration. Magnesium is not required and does not stimulate the enzymatic activity. Under DNA denaturing conditions, topoisomerase V catalyzes an unlinking reaction which results in substantial reduction in the linking number of closed circular DNA The driving force for this process is DNA melting. Camptothecin is not nearly as good an inhibitor for topoisomerase V as it is for eukaryotic topoisomerase I. The unique OCCUP rence of two major type I topoisomerases (reverse gyrase and topoisomerase V) in M. kandkri may shed new light on the evolution of this family of enzymes and supports the concept of a distant but significant relationship between some hyperthermophilic organisms and eukaryotes.
Type I DNA topoisomerases can be divided into two nonrelated groups (Table I ( ). Group A comprises enzymes making a transient covalent complex with the 5' end of the broken DNA strand. These enzymes act on negatively, but usually not on positively, supercoiled DNA and are inhibited by single-stranded competitor DNA. They can act on positively supercoiled DNA if a single stranded loop is inserted into the molecule or if a region of a positively supercoiled circle melts at high temperature (1,16,17,19,(21)(22)(23). A divalent cation is * This work was supported in part by grants of the Leibniz F'reis (to Alexander von Humboldt and Alfred P. Sloan Fellowships (to A. I. S.). K. 0. S.) and the National Science Foundation (to J. A.

L.) and by
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 8 'Ib whom correspondence should be addressed: Molecular Biology Inst., University of California, Los Angeles, 405 Hilgard Ave., Los Angeles, CA 90024. Tel.: 310-825-2545; Fax: 310-206-7286. Permanent address: Inst. of Molecular Genetics, the Russian Academy of Sciences, Moscow, Russia. required for the catalytic activity of these enzymes. Analysis of the DNA cleavage site specificity and gene sequences indicates that group A topoisomerases form two evolutionary distinct classes ( Table I). The first class, consisting of the enzymes named prokaryotic topoisomerase I, is represented by reverse gyrase in hyperthermophiles and by a DNA-relaxing enzyme in other prokaryotes (2,5,7,14).
Group B topoisomerases differ from group A in forming a covalent intermediate with the 3' end of DNA in their ability to relax both positive and negative supercoils and in their activity in the absence of divalent cations (Table I). Until recently, group B enzymes had been found only in eukaryotes and poxviruses. Topoisomerase I species purified from different eukaryotic organisms share a consensus sequence for DNA cleavage sites, are inhibited by camptothecin, and are immunologically cross-reactive (6, 28-31).
Previous attempts to find a magnesium-independent topoisomerase in prokaryotes, including thermophiles, were unsuccessful (32). There are, however, the enzymes of site-specific recombination which belong to the A integrase family and posses very weak topoisomerase activity (one cycle per 100 min) in the absence of a divalent cation (33-36). These enzymes show biochemical properties common to type I-group B topoisomerases, but phylogenetically and immunologically they are not related to eukaryotic topoisomerase I (33-42).
Most recently we discovered the first prokaryotic DNA topoisomerase which does not require divalent cations and named it topoisomerase V (42). It was found in Methanopyrus kandleri, a hyperthermophilic methanogen which grows up to 110 "C and represents a separate lineage distinct from other methanogens (43-46). Our preliminary findings showed that topoisomerase V relaxes positive and negative supercoils with a rate similar to other DNA topoisomerases and binds to the 3' end of the cleaved DNA strand, i.e. belongs to group B (42). Furthermore, the enzyme is recognized by a polyclonal antibody against human topoisomerase I. The one cleavage site mapped so far resembles the consensus site for DNA cleavage by eukaryotic topoisomerase I in the absence of camptothecin. In this paper we describe the purification of topoisomerase V from M. kandleri. We present a comparative study of effects of temperature, ionic conditions, and initial DNA superhelicity on the relaxation and unlinking reactions.

EXPERIMENTAL PROCEDURES
Cells". kandleri strain AV-19 (DSM 6324) has been described (45). Batch cultures were grown in "BSM" medium at 100 "C in a 300-liter enamel-protected fermentor (HTE Bioengineering, Wald, Switzerland) with stirring and continuous gassing (H2/C02, 80:20) as described (45). Exponentially growing cells were rapidly cooled, harvested with a separator (Westfalia, Germany), and stored at -70 "C. Materials-The protease inhibitors PMSF,' TPCK, TLCK, pepstatin A, leupeptin, and benzamidine were from Calbiochem. P-Mercaptoethanol was obtained from Fisher, Polymin P from British Drug House, and agarose SeaKem GTG from FMC. Reagents for SDS-PAGE were from Bio-Rad, CAPS from Boehringer Mannheim, and chloroquine and camptothecin from Sigma. Wheat germ topoisomerase I and pGEM3 DNA were from Promega, human topoisomerase I was from TopoGEN, proteinase K was obtained from Boehringer Mannheim, gel filtration and SDS-PAGE molecular weight protein standards were from Bio-Rad. Negatively supercoiled pBR322 DNA was a gift from Mary ODea, and pUC19 plasmid was from New England Biolabs. Positively supercoiled and relaxed DNA substrates were prepared by incubation with M. kandleri reverse gyrasez or wheat germ topoisomerase I, respectively.

3295
Ilbpoisomerase Assay-The standard conditions for the topoisomerase V assay have been described previously (42). For the topoisomerase assay in crude extracts, the reactions terminated by SDS were treated with proteinase K (400 pg/ml) at 37 "C for 1 h and then heated at 80 "C for 2 min. The topoisomerization products were analyzed by 1.5% agarose gel electrophoresis with 1.6 pg/ml chloroquine at 3 V/cm for 10 h.
One unit of activity was defined as the amount of enzyme required to relax 50% of form I pBR322 DNA (0.2 pg) in the standard assay reaction.
Methods of Protein Characterization-Protein concentrations were determined spectrophotometrically. Protein composition of the fractions was analyzed by SDS-PAGE. Gels were stained using a Bio-Rad Silver stain kit or Coomassie G250.
Protein Purification-Unless indicated otherwise, the purification steps were performed at 4 "C. Chromatography was done on an LKB liquid chromatography system. 120 g (wet weight) of cells were thawed in a water bath at room temperature in 60 ml of lysis buffer: 100 m~ Tris-HC1, pH 8.0 at 25 "C, 0.5 M NaC1, 10 I" P-mercaptoethanol, 50 pg/ml each of PMSF, TPCK, TLCK, pepstatin A, and leupeptin, 1 I" benzamidine and passed through a French pressure cell (American Instrument) at 16,000 p. 8.i. The recovered solution was diluted with lysis buffer to a final volume of 300 ml and centrifuged for 2 h at 40,000 rpm in a Beckman Ti-50 rotor (fraction I, 259 ml). 5% polyethleneimine P, pH 7.0 After mixing at 0 "C for 30 min, the solution was centrifuged at 12,000 rpm for 40 min in Sorvall RC-5B. The supernatant was saved (fraction 11,265 ml), and 245 ml of 4 M ammonium sulfate was added while stirring. Then solid ammonium sulfate was added to saturation, and the solution was stirred overnight at 4 "C. The supernatant (545 ml) was decanted and centrifuged at 11,000 rpm in Sorvall RC-5B for 2 h. The pellet was dissolved in a 400 ml of buffer A (30 m~ sodium phosphate, pH 7.4 at 25 "C, 10 m~ P-mercaptoethanol, 10% glycerol, 25 pg/ml each of PMSF, TPCK, and TLCK, 5 pg/ml pepstatin A, 1 pg/ml leupeptin, 1 m~ benzamidine) plus 0.2 M NaCl and dialyzed against two 2-liter changes of the same buffer (fraction 111,438 ml). After centrifugation the solution was loaded onto a phosphocellulose P11 column (2.6 x 30 cm, Whatman) equilibrated with the same buffer. After loading, the column was washed with 3 column volumes of this buffer. Topoisomerase V was eluted with a 600-min linear gradient of 0.2-1.0 M NaCl in buffer A at 1.5 d m i n , followed by a 400-min linear gradient of 1.0-2.0 M NaCl in buffer A. 15-ml fractions were collected and assayed for DNA relaxation activity. Active fractions were combined into two pools: 0.55-0.73 M NaCl pool (135 ml) and 0.73-1.45 M NaCl pool (620 ml). The latter was concentrated in Centriprep 30 cartridges (Amicon) to a final volume 100 ml.
Fraction IVa was loaded onto a 5-ml HiTrap heparin column (Pharmacia Biotechnology Inc.) equilibrated with buffer B + 0.5 M NaC1. After washing with 5 volumes of the same buffer, a 100-min linear gradient of 0.5-1.5 M NaCl in buffer B at 0.5 d m i n was applied. 1-ml active fractions between 1.0 and 1.25 M NaCl were pooled (fraction Va, 13 m l ) and concentrated on a 1-ml HiTrap heparin column, ie. the NaCl concentration was decreased to 0.5 M by dilution with buffer B, the sample was loaded on the small column and eluted by a steeper gradient of NaCl. The concentrate was passed through a HiLoad 16/60 Superdex 200 F G column (Pharmacia Biotechnology Inc.), equilibrated with 30 m~ Tris-HC1, pH 8.0 at 25 "C, 1 M NaCI, 5% glycerol, 2 m~ P-mercaptoethanol (fraction VIa, 13 ml).
Fraction IVb was chromatographed on a 5-ml HiTrap heparin column in the same way as fraction IVa. The active fractions eluted between 0.95 and 1.25 M NaCl were combined (fraction Vb, 15 ml) and concentrated on a 5-ml HiTrap heparin column. 90% of the total activity eluted between 1.07 and 1.17 M NaCl. It was collected into three separate pools of 3, 2, and 3 ml. Each of the pools was subjected to gel filtration as above. This resulted in fractions VIb-d (13 ml each). Fractions VIa-d were stored at 4 "C or -80 "C. in Table 11. The enzyme was purified by a procedure involving three chromatographic steps. The assay was based on its ability to relax positively and negatively supercoiled DNA without magnesium ions. DNA degradation by nucleases, a common problem in the assays of magnesium-dependent topoisomerases, was not observed in our standard assay condition. We noticed, however, a substantial change in the electrophoretic mobility of supercoiled or open circular DNA after incubation with early fractions unless the samples were heated to 80 "C in the presence of SDS or treated with proteinase K (not shown).

Purification of DNA Topoisomerase
We presume the aberrant migration is caused by DNA-binding proteins, and we do not know how they affect the quantitation of topoisomerase V activity during the first steps of purification.
Approximately one-half of the protein of M. kandleri was precipitated by saturated ammonium sulfate. No DNArelaxing activity was found in the supernatant. Phosphocellulose column chromatography resulted in the elimination of the majority of proteins (Table 11). Topoisomerase V began to elute at 0.55 M NaCl and formed a broad trailing peak up to 1.45 M NaCl. The activity was pooled into two fractions Iva and IVb that differ in protein composition (Fig. 1). Fraction IVb was concentrated 6-fold using Centriprep 30 cartridges. DNA relaxing activity eluted from heparin Sepharose column at about 1 M NaCl (Fig. 2). Fraction Iva was purified 50-fold and gave a nearly pure preparation of the enzyme ( Fig. 2A, elution fractions 30-32). Topoisomerase V comprises 60% of the total protein in fraction Va as judged from the data on its specific activity relative to pure enzyme (Table 11). Gel filtration chromatography of fraction Va on a Superdex 200 column (Fig. 3A, profile a ) gave only one peak of DNA relaxing activity (not shown). This peak exactly coincided with the major protein peak of one detectable polypeptide with M, 110,000 as determined by SDS-PAGE (Fig. 3). According to the gel filtration calibration standards, however, this protein peak corresponds to the chromatographic behavior of a globular protein with M, 142,000.
Heparin chromatography of fraction IVb resulted in a 12-fold purification of topoisomerase V (Table 11). The protein composition, however, remained complex (Fig. 2B). During concentration of fraction Vb on a heparin column, the activity was collected into three pools which were applied to a Superdex 200 column. All three gel filtrations gave a protein peak in the position where pure topoisomerase V elutes (Fig. 3A, profiles b-d 1. In all cases DNA relaxation activity was centered around This material was purified by gel filtration 10-fold to 50% purity (fraction VIc, Table I1 and Fig. 3B). of relaxation of the composite pBR322 DNA preparation containing negatively and positively supercoiled topoisomers by 1.5 units of topoisomerase V at 88 "C at different salt conditions.

Relaxation of Supercoiled
Topoisomerase V is active in a wide range of K-Glu concentration from 75 m to 3.1 M (Fig. 4A). The highest activity is at 1.5 M K-Glu (Fig. 4A, lane 12). The activity is reduced by about an order of magnitude at lower salt (0.075-0.55 M, lanes 2-7). In K-Glu, topoisomerase V acts somewhat processively, as demonstrated by the appearance of fully relaxed topoisomers before the disappearance of negatively supercoiled substrate. Replacement of the Glu-anion by C1-narrows the range of Na+ and K+ concentrations for enzyme activity (Fig. 4, B and C and Table  111). No activity is detected above 0.65 M KC1 or NaCl, and the optimal salt concentrations are decreased from 1.5 M K-Glu to 0.35 M NaCl and 0.45 KC1 (Table 111) without substantial change in the maximal activity of the enzyme (Fig. 4A, lane 12 versus B, lane 6 and C, lane 5). In NaCl, DNA relaxation is rather more distributive.
Essentially the same results as above were obtained at lower temperature (Table 111). The use of MgCl, instead of EDTA at 80 "C (not shown) does not stimulate or inhibit DNA relaxation by topoisomerase V (Table 111).
Previously we have shown that topoisomerase V relaxes positively and negatively supercoiled substrates with nearly equal efficiency in 1 M K-Glu at 88 "C ( Fig. 2   chloroquine gives distinct bands for positively and negatively supercoiled substrates (42). Inspection of the results shows that topoisomerase V prefers positively over negatively supercoiled DNA (Fig. 4A, lanes 9-13, and B, lanes 4 -8 , C, lanes 4-8).
The decrease in positive superhelicity (Fig. 4C, lane 8 ) or ita complete relaxation (Fig. 4B, lanes 6 and 7) can take place prior to the use of negatively supercoiled topoisomers. From the results of Fig. 4, which were obtained at a 1:5:5 ratio of enzyme molecules to positively and negatively supercoiled DNA molecules, we conclude that topoisomerase V prefers cycling on positively supercoiled molecules.
The results of topoisomerase V action at 95 "C are shown on Fig. 40. At this temperature, DNA melting is favored even at the highest salt concentration. As we showed previously, DNA melting forces DNA-relaxing enzyme to decrease linking number (17, 42, 47). This reaction, termed unlinking, results in a highly unwound form which migrates faster than negatively supercoiled substrate (Fig. 40, lanes 1-12). Fig . 40 shows that the unlinking activity of topoisomerase V at 95 "C increases with the increase in K-Glu concentration. Thus, the ionic effects on relaxing and unlinking activities are similar (Fig. 4 and D). There is, however, a difference between the two reactions catalyzed by topoisomerase V. The enzyme prefers to relax positive supercoils at 88 "C ( Fig. 4A), but we have not found conditions at 95 "C, where it would discriminate between positively and negatively supercoiled substrate during the unlinking reaction. lbpoisornerase V Activity on Relaxed DNA-It is difficult to follow the unlinking reaction on the supercoiled substrate because the products migrate close t~, or with, the substrate on the gel (Fig. 4 A 4 , lanes 2 and 3, and D, lanes 1-12). A simpler assay uses a relaxed substrate and allows the detection of topoisomerase V activity at very low salt concentration (Fig. 5).
At 80 "C in the range 1-11 IMI NaCl, the enzyme unlinks DNA (Fig. 5A, lanes 2 and 3). The extent of conversion increases when the amount of enzyme is increased 5-fold (Fig. 5C,  melt, and topoisomerase V does not unlink DNA (Fig. 5A, lanes [5][6][7] , and C, lanes 3, 4 ) . Instead, it re-relaxes DNA to a state in accordance with the incubation temperature of 80 "C. It is known that at low NaCl concentration the addition of Mg2' stabilizes the DNA duplex. As can be seen from Fig. 5, Mg2' suppresses the unlinking reaction in favor of DNA rerelaxation (A, lanes [9][10][11] and C, lanes 6 and 7). NaCl concentrations below 155 m do not affect the topoisomerase V reaction in the presence of 5 m MgC12 (Fig. 5A, lanes 9-12 versus lanes 2-5, and C, lanes 6-8 versus lanes 1 3 ) .
Results on DNA unlinking at 95 "C are presented on Fig. 5, B and D. At this temperature, at low salt concentration, nonenzymatic strand separation of open circular DNA yields single-stranded species (Fig. 5B, lanes 1 and 8). At 0.5 M NaCl, incomplete DNA melting prevents strand separation (Fig. 5D,   lanes 5 and IO). The products of DNA unlinking catalyzed by topoisomerase V are seen as a band or a doublet (marked by U).
In the presence of 5 m MgC12, the yield of products is insensitive to NaCl concentration in the range 1-91 mM, while it depends on the amount of enzyme (Fig. 5B, lanes 9-12, and D,  lanes 6 and 7). Without MgC12, the yield of unwound topoisomers increases with NaCl concentration in the range 1-191 m and then drops at 0.5 M (Fig. 5B, lanes 2-7, and D, lanes 1 4 ) . The unlinking reaction is highly processive in low NaCl conditions (e.g. Fig. 5A, lanes 2 and 3; B, lanes 9-12; C, lane 1, and 0, lanes 6 and 7) and becomes less processive a t higher salt concentration (Fig. 5B, lanes 6 and 13, and 0, lanes 3,4,8, and Reversibility of Topoisomerase V Binding to DNA-The ability of M. kandleri topoisomerase V to dissociate from a product topoisomer and to bind to another substrate was tested using a competition assay (Fig. 6). The scheme of the experiments was as follows. Topoisomerase V was preincubated with substrate DNA a t 300 mM NaCl at room temperature. Then the sample was diluted, and the NaCl concentration was adjusted either to 50 or to 300 m NaCl. Competitor DNA was added, and the reaction was performed at 90 "C, with products analyzed using two-dimensional gel electrophoresis. In contrast to previously described experiments, in competition assays topoisomerase V was in molar excess over plasmid DNA. Fig. 6 shows that at 300 l l l~ NaCl the enzyme utilizes both substrate and competitor DNA, giving unwound topoisomers (indicated by arrows). The result does not depend on which DNA was premixed with the enzyme or whether premixing was with one or both plasmids ( Fig. 6, A and B, lanes [6][7][8]. This means that at these salt conditions topoisomerase V effectively cycles between DNA molecules. At 50 m NaCl, topoisomerase V produces a highly unwound form that migrates in the second dimension faster than control negatively supercoiled DNA (Fig. 6, A and B, lane 1 ). If the ,enzyme is premixed with both plasmids, it yields unwound products of both substrates (lane 1). If only one plasmid is premixed with topoisomerase V, prior to the decrease in salt concentration and addition of the competitor, then the enzyme unwinds only one plasmid and does not act on the other (Fig. 6,   A and B, lanes 2 and 3). Only a few or no products of competitor DNA are seen if both substrate and competitor are relaxed DNA (Fig. SA, lanes 2 and 3 ) or if one is relaxed and the other is supercoiled (Fig. 6B, lanes 2 and 3). We conclude that topoisomerase V is tightly bound to DNA at 50 m NaCl and that its cycling between DNA molecules is very inefficient.

Concentration at which the activity is maximal
Highest concentration at which the activity was detected K-Glu KC1 NaCl K-Glu KC1 NaCl  5) and up to 100 p~ in 10% MezSO  (lanes 8-10). Higher concentrations of camptothecin (100 p~ in 2% MezSO and 1 m~ in 10% Me2SO) are required to decrease the extent of DNA relaxation (lanes 6 and 11 ). From this data we conclude that camptothecin is not nearly as good an inhibitor of M. kandleri topoisomerase V as it is of eukaryotic topoisomerase I.

Comparison of M. kandleri DNA 'Ibpoisomerase V with Other Group B Topoisomerases-Common
properties of the group B DNA topoisomerases, including eukaryotic topoisomerase I, poxviral topoisomerase, and Methanopyrus topoisomerase V are listed in Table I. Topoisomerase V has a molecular mass of 110 kDa (Fig. 3) which is within the 62-165-kDa size range of topoisomerase I purified from different eukaryotic sources (8,12,48). The poxviral enzyme is only 32 kDa (49). Many eukaryotic topoisomerases are immunologically cross-reactive, although in Xenopus laevis two forms of topoisomerase I are not mutually cross-reactive (31,50). Interestingly, topoisomerase V from Methanopyrus is recognized by antibody against human topoisomerase I (42). A major biochemical difference between these enzymes is the conditions in which they are enzymatically active. Eukaryotic topoisomerases are active below 41 "C and rapidly lose activity at 60 "C. Poxviral topoisomerase is slightly more active at 55 "C than at 30 "C and retains some activity at 75 "C (51). Methanopyrus topoisomerase V is active up to at least 100 "C (42). It is active in an unusually wide range of ionic strengths from 10 m~ to 3.1 M (Fig. 4). The highest activity is at 0.35 M NaCl, 0.5 M KCI, or 1.5 M K-Glu (Table 111), well above the optima for other topoisomerases (8,12,47,49,52). This result is not unexpected due to the high osmolarity inside M. kandleri cells. Cyclic 2,3-diphosphoglycerate, a trivalent anion, is present at 1.1 M (45). Presumably, the concentration of potassium counterion is 3 M (53).
Mg2' is not required for group B DNA topoisomerases, although it can stimulate eukaryotic and poxviral enzymes by 2-25-fold (12,49,52). We have not found conditions where MgClz will influence DNA relaxation by topoisomerase V.
Camptothecin inhibits topoisomerase I from many eukaryotes (28-30), but it is not nearly as good an inhibitor for poxviral topoisomerase (49) and Methanopyrus topoisomerase V (Fig. 7). Interestingly, a single point mutation in the naturally resistant Vaccinia enzyme makes it sensitive to the drug (54).
DNA Unlinking Catalyzed by Thermostable 'Ibpoisomeruses-At high temperatures DNA melting drives the catalysis of DNA unlinking by DNA-relaxing topoisomerases (Fig. 8). It can proceed to a very low linking number which is determined by the extent of the double helical regions which remain after  6 ) were incubated with 100 units of topoisomerase V a t room temperature for 10 min in 30 IILM 'his-HC1, pH 8.0, at 25 "C, 300 IILM NaCl, 5 m~ NazEDTA followed by dilution in the same buffer (lanes 6 8 ) or reducing NaCl concentration to 50 m~ and addition of pGEM3 (lanes 2 and 7) or pBR322 (lanes 3 and 8) and incubation a t 90 "C for 15 min. Relaxed pGEM3 (lane 4 ) and pBR322 (lane 5) were used for the incubations; negatively supercoiled pGEM3 (lane 9) and pBR322 (lane 10) are included as markers. B, the same as in A but a full set of topoisomers of pGEM3 from relaxed to supercoiled was used instead of relaxed pGEM3 (lane 4 ) , and incubation at 90 "C was for 10 min.
Two-dimensional agarose gel electrophoresis was without chloroquine in the first dimension and with 4 pg/ml chloroquine in the second. The positions of highly unwound pBR322 (middle) and pGEM3 (bottom) which run in the second dimension faster than native -sc DNA are indicated by arrows. relaxation. The latter is regulated by the temperature and comficiently unlinks circular substrates with any topology (Figs. position of the buffer. [4][5][6] and Fig. 1  group A, requires a single-stranded substrate, and therefore As expected, the lower the ionic strength the lower the temuses positively supercoiled DNA only inefficiently as a sub-perature at which unlinking reaction occurs (Figs. 4 and 5). The strate. Methanopyrus topoisomerase V, a group B enzyme, ef-mobility of the products, which is a measure of the extent of lbpoisomerase V RG. 7 unwinding, increases as the salt concentration decreases (Figs.  4 0 , 5, and 6). These results are in complete agreement with known ionic effects on the DNA helix-coil transition (55-57).
We found that topoisomerase V preferentially relaxes positive supercoils (Fig. 4, A X ) . However, when the unlinking reaction takes place, the enzyme does not discriminate between positively and negatively supercoiled DNA (Fig. 40). The difference between the two reactions is illustrated in Fig. 8. In conditions where the relaxation of double helix occurs, substrates retain the sign of the initial superhelicity. At higher temperatures, DNA melting generates positive superhelicity in the remaining duplex regions, regardless of the initial linking number of the DNA. Therefore substrates which were initially positively or negatively supercoiled become indistinguishable for the enzyme at those temperatures. We previously observed a similar effect for Desulfurococcus topoisomerase I11 which unlinks DNA using single-stranded regions (17). From these two examples we conclude that: (i) DNA unlinking driven by melting is a common reaction for thermostable relaxing topoisomerases and (ii) at temperatures above the melting temperature of linear DNA, the differences in initial supercoiling of substrates no longer affect the mechanism of this process.
DNA Ilbpokomemses in Hyperthermophiles-Three type I DNA topoisomerases have been purified from hyperthermophiles. Reverse gyrase was detected in many representatives of the hyperthermophiles (32, 58-60). It positively supercoils DNA at the expense of ATP hydrolysis. Its functional and phylogenetic counterpart in prokaryotes with lower growth temperatures is the ATP-independent relaxing DNA topoisomerase I. Me-dependent DNA topoisomerase I11 was purified from Desulfurococcus amylolyticus (17). It relaxes negative supercoils and catalyzes DNA unlinking at higher temperatures. Similar activity was detected in other thermophiles (61). Its bacterial and eukaryotic analogs are thought to be Escherichia coli and yeast topoisomerase 111. Its functional role remains obscure (11,17,26,27,62-64). So far DNAtopoisomerase V has been found only in M. kandleri. It relaxes positive and negative supercoils and unlinks DNA at higher temperature. The estimated abundance of topoisomerase V is at least 1,500 copied cell. Its function may be similar to that of eukaryotic topoisomerase I: relaxation of superhelical tension generated by replication or transcription (18, 20, 65).
In addition, topoisomerase V, together with reverse gyrase, may regulate the superhelicity of chromosomal DNA.
Recent discoveries of eukaryotic-like enzymes in hyperthermophiles provide additional support for the hypothesis that the prokaryotic ancestor of eukaryotes was a hyperthermophile