A quantitative decatenation assay for type II topoisomerases

ANALYTICAL

BIOCHEMISTRY

A Quantitative

156, 364-379

( 1986)

Decatenation BENI

M.

SAHAI

Received

Assay for Type II Topoisomerases J.

AND

February

CORDIN

KAPLAN

4, 1986

Type 11 topoisomerases catalyze decatenation of the catenated network of kinetoplast DNA [J. C. Marini. K. G. Miller, and P. T. Englund (1980) J. Bid. (%em. 255, 4976-49791. The individual DNA circles and small catenanes produced during the decatenation reaction can be separated from the large network of substrate DNA by 5 min centrifugation at 13,000~ and quantitated. The appearance of these decatenated DNA molecules which appear in the supematant tirst showed a lag. whose duration depended on the enzyme concentration, and then increased linearly with time until it reached a plateau. The slope of the linear part of the kinetic curve was directly proportional to the enzyme concentration, whether a purified or crude preparation of type II topoisomerase from mammalian cells was used. These findings led us to a rapid quantitative assay of type II topoisomerases not involving electrophoresis. The method was developed with purified enzyme but was also useful for assay of the activity in crude extracts. Surprisingly, the type I topoisomerase. even when present in large excess, failed to decatenate the nicked DNA circles often present in the kinetoplast DNA. This renders the assay virtually free from interference by type I enzyme. The method is sensitive and allowed quantitative estimation of the enzyme activity present in the crude extracts corresponding to that derived from 500 to 700 cultured mammalian cells. Since various type II topoisomerases from procaryotic. eucaryotic. and viral sources decatenate kinetoplast DNA and generate similar DNA products. the assay method is likely to be generally applicable. ,c 1986 Academx Press. Inc. KEY WORDS: topoisomerases: kinetopiast DNA: decatenation; microcentrifugation: electron microscopy: fluorometry.

On the basis of reaction mechanism, DNA topoisomerases have been divided into two classes ( 1,2). Type I enzymes introduce a transient single strand break and change the linking number of DNA in steps of one; examples are w protein of lWwrickia roli (3) and topoisomerase I from HeLa cell nuclei (4). The type I enzymes catalyze various topoisomerization reactions in vitro such as relaxation of supercoiled DNA, formation and resolution of knots in single-stranded DNA circles, linking of single-stranded DNA circles of complementary sequences [for a review see Ref. (5)], and also the catenation, decatenation (6,7), and knotting (7) of nicked double-stranded DNA circles. The latter reactions with doublestranded DNA circles, which require eventual topological passing of both the strands, occur at low frequencies and require a much higher 0003-2697/86

$3.00

Copyright 0 1986 by Academic Press. Inc All ri?,bts of reproductmn in any form reserved.

364

enzyme concentration than that necessary for a single strand passing reaction, namely relaxation of supercoils (7). The type II enzymes, on the other hand, produce a transient double strand break and change the linking number of DNA in steps of two; examples are DNA gyrase (8) bacteriophage T4 topoisomerase (1,9). and topoisomerase II from HeLa cell nuclei (10). In vitro these enzymes relax supercoiled DNA (10-l 2), introduce knots in double-stranded DNA circles and resolve these ( 1,13), and reversibly catenate double-stranded DNA circles ( 10,12,14). The procaryotic type II enzyme, DNA gyrase, unlike its eucaryotic counterpart, can also supercoil the relaxed DNA circles (8). With the exception of relaxation of supercoiled DNA by gyrase ( 11) and knotting of duplex DNA by T4 topoisomerase (1). all known reactions of type II topoisomer-

ASSAY

FOR

TYPE

ases require ATP hydrolysis [for a review see Ref. (15)]. Despite recent advances in our knowledge of type II topoisomerases ( 16). there is pressing need for a convenient quantitative assay for these enzymes. The various methods that have been used to assay these enzymes are by following ATP-dependent (i) relaxation of supercoiled DNA (8,12,17,18) and, in the case of gyrase. supercoiling of relaxed DNA (19). (ii) unknotting of knotted double-stranded circles of bacteriophage P4 head DNA (20), (iii) catenation of double-stranded DNA circles ( 12.2 1,22), and (iv) decatenation of catenated double-stranded DNA circles such as the network of kinetoplast DNA (k-DNA)’ ( 10). As the type I topoisomerases also relax the supercoiled DNA and since a specific inhibitor of these enzymes is not known, the type II enzyme assays which are based on relaxation or supercoiling of DNA cannot be reliably used for preparations also containing the former enzyme. With preparations containing both type I and II enzymes the unknotting, catenation, and decatenation assays have been successfully used to determine specifically the type II enzyme activity since in these assays no type I enzyme-catalyzed reactions, measured in absence of ATP. were detectable (10,20-22). However, the reaction products (or the remaining substrate) in unknotting, catenation, and decatenation assays are quantitated after electrophoretic separation on agarose gels by scanning the negative of the gel photograph or. if a radioactive substrate was used, by measuring the radioactivity in the corresponding DNA band. These methods are time-consuming and are difficult to quantitate. They are not suitable for assays of a large number of samples as is often required during enzyme purification. studies of kinetic properties and regulation. and screening of various cell types for the activity. In this report we describe a I Abbreviations used: k-DNA. kinetoplast DNA: CV- 1S cells. CV- 1 monkey kidney cells grown in suspension; SDS. sodium dodecyl sulfate; kb. kilobase. PEG, polyethylene glycol.

365

II TOPOISOMERASES

highly quantitative, sensitive, specific, and rapid method for the assay of type II topoisomerases based on the decatenation of kDNA (23). In the assay, nanogram amounts of single DNA circles and small catenanes produced during the reaction are separated from the large catenated network of k-DNA by a brief centrifugation and then quantitated fluorometrically or by scintillometry. The assay method was characterized with purified enzyme but was found equally applicable to assay of enzyme in crude extracts.’ MATERIALS

AND

METHODS

[3H]k--DNA. C’rithidiu .fasciculata (generously provided by Dr. L. Simpson of the University of California, Los Angeles) were grown for 72 h at 27°C in brain heart infusion medium (Difco) supplemented with hemin and 10 pCi/ml of [Inethl&‘H]thymidine (78 Ci/ mmol; NEN). Cells were harvested and kDNA wasisolated asdescribed(24) except that prior to banding on CsCl the DNA preparation was incubated at 37°C for 20 min each with DNase-free RNase A (10 wg/ml) and subsequently with partially self-digestedproteinasek (20 pg/ml). Unlessstated otherwise, the final preparation consisted of a catenated network of nearly equal proportion of nicked and covalently closed DNA circles as determined by fluorometry of DNA before and after denaturation. and it had a specific activity of approximately 10.000 cpm/yg DNA. Culls. CV-IS monkey kidney cells and mouse L. cells were collected from log phase cultures. The number of cells was determined by counting in a hemacytometer. Eqww preparatiom. The extract containing topoisomerase activity was prepared by a previously described procedure (10) with minor modifications. Briefly, the cells were suspended at a density of 0.5 to 1 X lO’/ml in buffer A (5 mM potassium phosphate, pH 7.5, ’ A brief account of the method was published as an abstract (B. M. Sahai and J. G. Kaplan (1984) Fed. Proc. 43, 1543) and has been reviewed ( 16).

366

SAHAI

AND

1 mM MgC&, 0.5 mM dithioerythritol, 0.1 mM Na7EDTA. 0.005% Triton X-100, and 1 mM freshly added phenylmethylsulfonyl fluoride) and, after 20 min of swelling, lysed by addition of equal volume of buffer B (2 M NaCl, 0.1 M Tris-HCl, pH 7.5, 0.5 mM dithioerythritol, 0.005% Triton X- 100, and 1 mM freshly added phenylmethylsulfonyl fluoride). About 20 min later, the DNA in the lysate was precipitated by slow addition of one-half volume of solution C [ 18% polyethylene glycol (PEG), nfr 8000. 1 M NaCl. 0.5 mM dithioerythritol, and 0.005% Triton X-100] and allowed to stand for 30 min. (This step is not absolutely essential when radioactive k-DNA is used in the assay but is recommended to facilitate accurate pipetting.) After centrifugation at 13,OOOg for 5 min the supernatant, called PEG supernatant, was removed and used for the assay. The preparation contained both topoisomerase I and topoisomerase II activities. A typical preparation from CV- 1S cells (60 ~1 PEG supernatant per 2 X lo6 cells or 300 ~1 PEG supernatant per 1 X 10’ cells) contained approximately 800 pg protein/ml which was determined by the BioRad microassay procedure. Both topoisomerase I (4) and topoisomerase II (10) were purified from the nuclei of log phase CV- 1s cells to near homogeneity as described previously except that in case of topoisomerase II the fractionation on Sepharose 4B column was replaced by affinity chromatography on a novobiocin-Sepharose column (17). Topoisomerase I was purified approximately 600-fold and topoisomerase II about lOOO-fold. The approximate protein concentration was determined from the intensity of the protein band in polyacrylamide gels. A gradual loss of topoisomerase II activity occurred while stored at -20°C in presence of 50% glycerol. Decatenation of k-DNA and assay for type II topoisomerase activity. Typically, the decatenation reaction mixture in 40 ~1 contained 50 mM Tris-HCl, pH 7.9, 10 mM MgC12. 0.5 mM Na2EDTA, 1 mM ATP, pH 7.7, 0.5 mM dithioerythritol, 0.005% Triton X- 100, 30 pg/

KAPLAN

ml [3H]k-DNA, and 100 mM NaCl contributed by 4 ~1 of enzyme solution appropriately diluted in buffer D [mixture of equal volumes of buffer A, buffer B, and solution C supplemented with 100 pg/ml bovine serum albumin (BRL)] which contains 6% (w/v) polyethylene glycol, or in buffer E [20 mM Tris-HCl, pH 7.5, 0.5 mM dithioerythritol, 25% glycerol, 1 M NaCl, 100 pg/ml bovine serum albumin (BRL), and 0.005% Triton X-100]. An additional 0.05 mM dithioerythritol was also contributed to the assay by the enzyme sample. The incubation was carried out in 1.5-ml plastic Eppendorf tubes at 30°C and the reaction was stopped by addition of 5 ~1 of 2.25% sodium dodecyl sulfate (SDS) solution. The reaction mixture was centrifuged at 13,OOOg for 5 min in a table top Fisher microcentrifuge. Thirty-five microliters of supernatant was carefully removed and the DNA was quantitated by determining radioactivity by scintillation counting with 10 ml aqueous counting scintillant (Amersham). In assays where DNA was to be quantitated fluorometrically. the reaction was stopped by addition of 5 ~1 of 0.25 M Na,EDTA (instead of 2.25% SDS) and the reaction mixture was centrifuged as above. Thirty-five microliters of supernatant was removed and added to 1.6 ml ethidium bromide solution at pH 8 (5 mM Tris-HCl. 0.5 mM Na*EDTA, 0.5 pg/mI ethidium bromide) or at pH 11.9 (5 mM potassium phosphate, 0.5 mM Na2EDTA, 0.5 pg/ ml ethidium bromide). Fluorescence was measured with a Gilson Spectra/G10 Filter fluorometer equipped with filters for a peak excitation at 520 nm and a peak emission at 600 nm. Relaxation assays of topoisomerases. Topoisomerase I activity was assayed by following the relaxation of supercoiled PM2 DNA as described (25). One unit activity completely relaxed 0.1 pg PM2 DNA in 10 min at 37°C. Topoisomerase II-catalyzed relaxation of supercoiled PM2 DNA was carried out in 20 ~1 of reaction mixture whose composition was as described for the decatenation assay mixture except that k-DNA was replaced by 12.5 pg/

ASSAY

FOR

TYPE

ml PM2 DNA (85% form I) and that 2 ~1 enzyme solution was used. Reactions were stopped and DNA samples electrophoresed on 0.7% agarose gel as described below. DNA electrophoresis. For electrophoretic analysis. the decatenation and relaxation reactions were carried out in a final volume of 20 ~1 and stopped by addition of 5 ~1 of 5% (w/v) SDS, 25% (w/v) Ficoll, 0.25 mg/ml of bromphenol blue. Electrophoresis of DNA samples was performed on 3 mm horizontal agarose gels, either 1% (for 12 h at 1 V/cm in 90 mM Tris-borate, pH 8.3, and 2.5 mM Na?EDTA) or 0.7% agarose gels (for 15 h at 1 V/cm in 50 mM Tris, 20 tttM sodium acetate, 20 mM sodium chloride, and 2 mM Na2EDTA, pH 8.0). Gels were stained with ethidium bromide (1 pg/ml) and photographed under uv illumination. Electron tnicroscopj~. DNA samples were prepared for electron microscopy by the formamide spreading procedure described by Davis cz al. (26). Circumference of circular molecules was measured on prints (nominal magnification 54,000X) with a Hewlett-Packard 9874A digitizer coupled to a Tektronix 4051 graphic computer and number of base pairs calculated with pBR322 DNA (27’) as internal standard. RESULTS

AND

DISCUSSION

Decaterlation Qfk-DNA b!, Type II Topoisomerases k-DNA is mitochondrial DNA from trypanosomes such as Crithidia jasciculata. It consists of a large catenated network of two types of DNA circles, called minicircles and maxicircles. and is a complex of approximately 1 X 10” daltons (38,29). About 95-97% of its massconsistsof minicircles and the remaining 335% of maxicircles. In Crithidia the size of the minicircles is about 2.5 kilobases(kb) and of the maxicircles about 40 kb (29). Type II topoisomerasessuch asT4 topoisomerase and DNA gyrase decatenate the large network to produce the individual constituent DNA circles and smaller catenated complex as inter-

367

II TOPOISOMERASES

mediates (23). Type I topoisomerasesdo not catalyze decatenation of k-DNA (23). We have used both purified and crude preparations of topoisomerase II from log phase (Y-1 monkey kidney cells grown in suspension (CV- 1s cells) to characterize all aspectsof the requirements and the kinetics of the decatenation reaction. The crude preparation (the PEG supernatant) contained topoisomeraseI and DNase activities which were absent in the purified preparation. Figure 1 showsthe requirements and the kinetics of decatenation of k-DNA with crude preparations of topoisomerase II. The reaction required ATP (lane 3) and MgZt (lane 4), and it was partially inhibited if the dithioerythritol concentration was reduced from 0.55 to 0.05 mM (lane 5). The reaction proceeded in a timeA) I

2

3

4

5

6

8) 7

8

9

IO

0 II

12

13 14

I5 I6

FIG. I. Decatenation of k-DNA by topoisomerase 11. The decatenation reaction was carried out in 20-p] aliquots. Samples were subjected to electrophoresis on I % agarose gel and the gel was photographed as described under Materials and Methods. A crude preparation of topoisomerase II, the PEG supematant (see Matetials and Methods) from CV-IS cells, was used as the source of enzyme. (A) Reaction mixture was incubated for 60 min in absence of enzyme (lane 1) or in presence of 50 ng protein of the PEG supernatant (lanes 2-6). The reaction mixture was either complete (lane 2). or contained no ATP (lane 3). no M&i* (lane 4). 0.05 mM instead of 0.55 mM dithioerythritol ((lane 5). or no k-DNA (lane 6). (B) The reaction mixture, containing 50 ng protein of PEG supematant. was incubated for 0 min (lane 7). 15 min (lane 8) 30 min (lane 9). or 45 min (lane 10). Essentially similar results with respect to reaction requirements and kinetics were obtained with 0. I ng protein of purified topoisomerase II preparation as enzyme (results not shown). (C) The reaction mixture was incubated for 60 min with 3.12 ng (lane I I). 6.25 ng (lane 12). 12.5 ng (lane 13). 25 ng (lane 14). 50 ng (lane 15). or 100 ng (lane 1 I) protein of the PEG supernatant. K and M indicate catenated network of k-DNA and free minicircles of k-DNA. respectively.

368

SAHAI

AND

dependent manner (Fig. 1B) and its rate increased with increasing enzyme concentrations (Fig. 1C). Similar reaction requirements and a time-dependent decatenation were observed whether whole cell extract, nuclear extract, or purified topoisomerase II preparation was used as enzyme. The results with a crude enzyme preparation have been presented so as to demonstrate that the activities of topoisomerase I and DNase did not decatenate the k-DNA (lanes 3, 4). Figure 2A shows the dependence on enzyme concentration of decatenation of k-DNA in the presence of purified topoisomerase Il. The decatenation reaction proceeds through the production of intermediate-size catenanes seen at suboptimal enzyme concentration (lanes 7-9). These intermediates were barely detectable in reactions with crude enzyme preparation (Fig. 1, lanes 9, 10, and 12). Type I topoisomerases from E. co/i (6,7) and rat liver (7) have been shown to catalyze catenation and decatenation of double-stranded DNA circles provided at least one of the two circles was nicked, as well as knotting of nicked

A

TOPOISOMERASE

II

B

KAPLAN

double-stranded DNA circles (7). However, these reactions, involving topological passing of both the strands, require 10 to 20 times more enzyme than that required for a single strand passing reaction (7). Results in Fig. 1 (lanes 3 and 4) show that despite the presence of a sizable portion (about half) of nicked circles in k-DNA (see Materials and Methods), the topoisomerase I activity in the crude extract (which completely relaxed 0.6 /*g of supercoiled PM2 DNA in 10 min at 37°C) failed to catalyze detectable decatenation of k-DNA. This observation is in agreement with the previously published results (23) and consistent with the observations that topoisomerase I activity present in the crude mammalian ceil extracts fails to catalyze detectable (i) unknotting of knotted double-stranded circles of bacteriophage P4 head DNA which naturally contains nicks (20) and (ii) catenation of nicked PM2 DNA circles (22). under conditions in which topoisomerase II activity in the same extract efficiently catalyzed these reactions in the presence of ATP.

TOPOISOMERASE

I UN ITS

0.6

II52

4 8163264

0

6

I 152

4 8163264

FIG. 2. Actions of purified topoisomerase I and II on k-DNA. Decatenation reaction mixture (20 ~1) containing k-DNA with 7 1% nicked DNA circles was incubated at 30°C in absence or presence of increasing amounts of purified topoisomerases. The reaction was terminated at 30 min and samples were electrophoresed on 0.7% agarose gel as described under Materials and Methods. K, M. IC, I, I’, and II indicate catenated network of k-DNA, free minicircles. intermediate-size catenanes, supercoiled PM2 DNA, relaxed PM2 DNA. and nicked PM2 DNA circles, respectively. (A) Reaction mixture contained (lanes I-I I) 0, 3. 6, 12, 24, 36. 48. 60, 72, and 84 pg of purified topoisomerase II, respectively. (B) Reaction mixture contained the indicated units of purified topoisomerase I. The reaction mixtures on right panel also contained 0.25 pg of supercoiled PM2 DNA (95% form I).

ASSAY

FOR

TYPE

To determine whether or not higher concentrations of topoisomerase I would lead to decatenation of nicked circles from k-DNA, the decatenation reaction was carried out in the presence of various amounts of purified topoisomerase I ranging from 0.6 to 64 units (one unit of activity relaxed 0.1 pg supercoiled PM2 DNA in 10 min at 37°C). Surprisingly, no decatenation was detectable at either enzyme concentration (Fig. 2B, left). Prolonged incubation (up to 2 h) produced the same results. To ascertain that topoisomerase I was active, the decatenation reaction was carried out in the presence of 0.25 pg of super-coiled PM2 DNA. As seen in Fig. 2B (right). 1 unit of enzyme activity relaxed all the 0.25 pg supercoiled DNA in 30 min at 30°C but produced no decatenated DNA. There was no detectable decatenation even when the enzyme was present at 64 times greater concentration. On the other hand, consistent with published results (6,7), in a parallel experiment 16 units of purified topoisomerase I activity did catalyze a detectable catenation of nicked PM2 DNA circles (a double strand passing reaction), and the catenanes increased with increase in enzyme concatenation (results not shown). To ensure that the failure of topoisomerase I to decatenate was not due to presence of some impurity in a particular k-DNA preparation, a number of preparations of k-DNA with varying ratios of nicked and covalently closed circles (ranging from 1 to 3) were used as substrate. None of the k-DNA preparations was detectably decatenated by as high as 96 units of purified topoisomerase I activity in 2 h at 30 or 37°C. To test the limit, we used up to 400 units of purified topoisomerase I in a 40~1 decatenation assay mixture but still observed no decatenation (see Table 1B). Thus, we conclude that. contrary to expectation, the nicked DNA circles in k-DNA cannot be decatenated by topoisomerase I under these reaction conditions. This is probably due to (i) a much greater topological complexity of the k-DNA, such that it may be resolved by only a type II, not by a type I enzyme, and/or (ii) an exceptionally slow reaction in presence of

II TOPOISOMERASES

369

type I enzyme because of the presence of a single or a very limited number of, rather than multiple (6.7), nick(s) in the DNA circles.

Since the products of the decatenation reaction are considerably smaller in size than the large complex of k-DNA. we attempted to separate them by centrifugation. In the absence of enzyme. about 95% of the radioactivity of k-DNA present in the decatenation assay mixture was sedimented by a brief centrifugation at 13.OOOg (Table 1A). Centrifugation at 15,600g produced the same result whereas at 8700~ only 85% of the radioactivity of k-DNA sedimented (data not shown). However. after incubation in presence of a crude or purified preparation of topoisomerase II, a significant proportion of the radioactivity of k-DNA did not sediment and appeared in the supernatant after centrifugation (Table I). Heat-inactivated topoisomerase II failed to release the radioactivity of k-DNA to the supernatant. Furthermore, reaction conditions which do not allow decatenation of k-DNA, such as omission of ATP or Mg’+ (see Fig. 1A). prevented the appearance of radioactivity in the supernatant (Table IA). These results show that topoisomerase II generates DNA products which do not sediment with k-DNA and suggest that they resulted from decatenation. In the absence of ATP, the activities of topoisomerase I and ATP-independent nucleases, present in the crude preparations of topoisomerase II. released no DNA products from k-DNA that appeared in the supernatant after centrifugation (Table IA). This observation is in agreement with the results in Fig. 1 (lane 3). About 5% of the DNA that does not sediment with k-DNA after a 5-min centrifugation at 13,OOOg (Table 1A) consists of linear DNA often derived from maxicircles as revealed by electron microscopy (see below). This is essentially generated during the isolation procedure but can also occur as a result of prolonged (6 months or more) storage in deep freeze of high specific activity ‘H-labeled k-

370

SAHAI

AND TABLE

SEDIMENTATION

OFII-DNA

KAPLAN 1 BY CENTRIFUGATION

Centrifugation (min)

Incubation

time

Radioactivity

in supernatant (cm)

A. Crude enzymea With no enzyme With no enzyme With enzyme With heat-inactivated enzymeb With enzyme, -ATP With enzyme, -MgQ With enzyme. 0.05 mM dithioerythtitol

0 5 5 5 5 5 5

10.086 470 7,320 458 512 478 1.982

B. Purified With With With With With

0 5 5 5 5

56,077 503 42,980 491 589

enzymes’ no enzyme no enzyme topoisomerase topoisomerase topoisomerase

II (0. I5 ng) I (96 units) I (400 units)

’ In a final volume of 40 ~1, the decatenation reaction mixture containing [3H]k-DNA (see Materials and Methods) was incubated for 30 min in the presence or absence of 100 ng protein of PEG supernatant (the crude enzyme) from CV-IS cells. Where indicated the reaction mixture contained no ATP, no MgC12. or contained 0.05 IIIM instead of 0.55 mM dithioerythtitol. After incubation. 5 ~1 2.25% SDS was added and samples were centrifuged at 13.OOOg. 35 ~1 supernatant was removed and radioactivity determined. Similar results were obtained when 0.2 ng protein of a purified topoisomerase II preparation instead of 100 ng protein of PEG supernatant was used as enzyme (data not shown). ’ The enzyme preparation was incubated for 5 min in boiling water. ’ In a final volume of 40 ~1, decatenation reaction mixture containing [3H]k-DNA (54.000 cpm/pg DNA) that was further purified by banding on a second CsCl gradient (24) and that contained 55% nicked DNA circles was incubated for 30 min in the absence or presence of the indicated amounts of purified topoisomerase. After the incubation the samples were processed as described in footnote a.

DNA. However, most of the linear DNA molecules can be separated by rebanding of kDNA preparations on CsCl(24). Results with k-DNA thus purified as substrate are depicted in Table 1B, which show that DNA in the supematant without enzyme (which would serve as a blank in an assay) can be reduced to 1% or even lower as seen in some other experiments. The results also show that under these reaction conditions even 400 units of purified topoisomerase I activity failed to decatenate k-DNA that contained more than half the DNA circles as nicked circles.

Characterization of the DNA Products in Supernatant After electrophoresis on 1% agarose gel, the DNA molecules in the 13.OOOg supernatant

resolved into a fast-migrating prominent species which was preceeded in the gel by a faint smear of larger DNA and a small proportion of very large DNA that barely entered the gel (Fig. 3A). The fast-migrating species constituted 40-500/n of the total DNA as measured by radioactivity. It consisted of nearly equal proportions of covalently closed and nicked double strand DNA circles as revealed by electrophoresis in presence of ethidium bromide (Fig. 3B), before (lane 2) and after (lane 3) denaturation of DNA. The nicking of the DNA did not occur during the decatenation reaction since a similar proportion of nicked DNA circles existed in the untreated k-DNA (see Materials and Methods). The nicked DNA circles migrated near the region of 2.5 kb (Fig. 3B). These circles consisted of 2570 t 40 base

ASSAY

FOR

TYPE

371

II TOPOISOMERASES B)

I23

FIG. 3. Electrophoretic analysis of DNA in the supernatant. Decatenation of k-DNA was carried out in a volume of 120 @I in the presence of 300 ng protein of the PEG supernatant (the crude enzyme) from CV1s cells. After 30 min incubation. 15 ~1 of 2.25% SDS was added and the reaction mixture centrifuged at 13.OOOg for 5 min. To the 105~~1 supernatant , 22.5 ~1 of 5%) (w/v) SDS. 25% (N/V) Ficoll, and 0.25 mg/ml bromphenol blue were added, 30-~1 aliquots were subjected to electrophoresis on 1% agarose gel and the gel was photographed as described under Materials and Methods. ‘(A) Distribution of DNA and radioactivity after electrophoresis: The gel in the lane was sliced into 9-mm sections. Each section was minced and incubated for 4 h at 80°C with 1 ml NCS tissue solubilizer (Amersham), and the radioactivity determined by scintillation counting. Origin is at the left. To determine the amount of radioactivity applied to the gel. 30 ~1 of the sample was incubated with I ml NCS as above and then counted. Of the 3300 cpm applied on the gel. a total of 3250 cpm was recovered in the gel sections after electrophoresis. (B) Distribution of DNA after electrophoresis in presence of ethidium bromide: Both the gel and electrophoresis buffer contained 0.24 pg/ml ethidium bromide. The sample of the supernatant applied in lane 2 received no treatment whereas that in lane 3 was heat denatured in presence of 10 mM KOH prior to electrophoresis. Sample in lane 1 contained, in 25 ~1, Hind111 restriction digest of 2 pg phage lambda DNA. I% (w/v) SDS. 5%’ (w/v) Ficoll. and 0.05 mg/ml bromphenol blue. The DNA fragments served as size markers (30). N, L, and C indicate nicked. linear. and covalently closed circular DNA, respectively.

pairs as calculated from the circumference which was measured by electron microscopy, using pBR322 DNA (27) as internal standard. Since under the reaction conditions DNA circles of this size (other than minicircles of kDNA) are not likely to be produced, we conclude that the DNA circles of the fast migrating DNA band are minicircles. The linear form of the circular molecule was not detected. Nearly all (96%) of the radioactivity applied to the gel was recovered after electrophoresis and the distribution of radioactivity coincided with that of DNA on the gel (Fig. 3A). This and the fact that all the radioactivity in the supernatant could be precipitated by 10% trichloroacetic acid showed that the 13,OOOg supernatants did not contain significant amounts of nucleotides or oligonucleotides even when

we used a crude preparation of topoisomerase II that showed significant levels of nuclease activity under different assay conditions. Electron microscopy of untreated k-DNA shows a complex network, somewhat smaller in diameter (4-6 pm) than that reported previously (23) (Fig. 4A); the difference may be due to the presence of polyethylene glycol (0.53%) in our DNA samples. The supematant obtained after centrifugation (13,OOOg) of untreated k-DNA contained only a small quantity (2-5%) of DNA mainly consisting of large linear molecules (photograph not shown) presumably derived as a result of linearization of maxicircles. The supernatant obtained after the treatment of k-DNA with topoisomerase II contained much higher amounts of DNA which consisted mainly of double strand cir-

372

SAHAI

AND

KAPLAN

FIG. 4. Electron micrographs of DNA. Section A shows the network of k-DNA after incubation at 30°C for 60 min in decatenation reaction mixture in the absence of enzyme which was substituted with an equal volume ofbuffer D containing 6% polyethylene glyeol. The inset shows a portion of the network at a twofold higher magnification. The decatenation of k-DNA was carried out for 60 min with either 100 ng protein of the PEG supernatant (the crude enzyme) or 0.2 ng protein of purified topoisomerase II from CV- 1S cells; the reaction was stopped by addition of SDS solution and the reaction mixture centrifuged as described. The supernatant was examined under the electron microscope as described under Materials and Methods. The forms and sizes of the DNA molecules present in the supematants after decatenation with crude or purified enzyme preparation were similar. Electron micrographs in B. C (with crude enzyme) and D (with purified enzyme) are selected to show the types of DNA molecules present in the supernatant. Arrows show di- and trimeric catenanes. (Electron micrographs by courtesy of Dr. D. G. Scraba.)

ASSAY FOR TYPE II TOPOISOMERASES

cles of 2570 f 40 basepairs (Fig. 4C and D), corresponding to the size of minicircles of kDNA. The supernatant also contained a small proportion of two- or three-molecule catenanes (Fig. 4C) and relatively few large catenated complexes (Fig. 4B) and maxicircles (not shown). Linear DNA moleculescorresponding to the size of minicircles were seenonly rarely. These observations are in agreement with the results of electrophoretic analysis of DNA in the supernatant (seeFig. 3). Becauseof(i) the releasein the supernatant of covalently closed DNA circles in a proportion similar to that presented in the untreated k-DNA substrate (Fig. 3B), and (ii) failure of activities of topoisomerase I and nucleases present in the crude enzyme preparation to releasedetectable DNA products (Table 1A), we conclude that the releasefrom k-DNA of all the various types of DNA products that appear in the supernatant is essentially due to decatenation by topoisomerase II activity.

373

are converted into smaller catenanes and free DNA circles which then appear in the supernatant. This view was supported by two observations: (i) the decatenation reaction does proceed through the formation of intermediate-size catenanes(Fig. 2A, lanes 7-9, also see below Fig. 7B, lanes 5- 11) and (ii) at the same enzyme concentration, the duration of the lag period with a partially decatenatedpreparation of k-DNA as substrate was much shorter than with the intact k-DNA (results not shown). After the initial lag, the reaction progressed linearly with time until it reached a plateau. At the plateau, 70 to 90% of the mass of kDNA has been converted into smaller decatenated DNA products (Figs. 5A and B). Reactions using a concentration of k-DNA four times greater resulted in a substantial elevation of the plateau level but without altering the rate of reaction (results not shown), suggesting that the concentration of k-DNA in the normal reaction was saturating. Because of the presence in the reaction mixture of 100 mM NaCl which is in excess of the concentration of Kinetics of‘ Deecatenation of’k-DNA monovalent cation required to prevent the agThe kinetics of decatenation was followed gregation of DNA (3 1), the aggregation of deby measuringthe amount of decatenated DNA catenated DNA circles would be at a minimolecules that appeared in the 13,OOOgsu- mum. This would prevent their recatenation pernatant. The DNA was quantitated either by topoisomeraseII since aggregation of DNA by scintillometry (Figs. 5A and B) or by fluo- circles is required for their catenation (3 1). rometry in presence of ethidium bromide An advantage of the fluorometric method (Figs. 5C and D). In caseof both purified and is that it allows one to check for possible oncrude preparations of topoisomerase II, a lag going nucleaseaction during the decatenation occurred in the early phase of the reaction of k-DNA with crude enzyme preparations. during which the amount of decatenated DNA From the ratio of the fluorescence intensities in the supernatant did not increase in direct measuredbefore (b) and after (a) denaturation proportion to the time of incubation. The lag of the product DNA, it is possibleto monitor usually lasted until about 190 ng DNA prod- and quantitate DNA molecules released due ucts were releasedin the supernatant. The du- to the action of nucleaseson the k-DNA since ration of the lag decreasedwith increasing en- such molecules would lose fluorescence after zyme concentration and at high enzyme con- denaturation. This method also allows the centrations the lag period was considerably specific quantitation of covalently closedDNA diminished or absent. This suggestedthat the products (32). Results in Fig. 5D show that lag was not related to activation of the enzyme. even with a crude enzyme preparation such The most likely explanation for this lag is that as PEG supernatant, which contains signifiduring the initial reaction most of the reaction cant nucleaseactivity assayableunder different products are large catenaneswhich cosediment assay conditions, there was no detectable rewith k-DNA and asthe reaction proceedsthese leaseof DNA due to cutting or nicking of the

374

SAHAI I N

8

I

1

A ,~

AND

KAPLAN

I 0.

- --o-------n --.------a

---.

-*-A--

I O

30

60

90 INCUBATION

I

I

30 TIME

------.

-06

*-o---L-*---~

40

o/b b

0" 2 20

x-e

I' ti L” ,' -0 /,i'-A 0

30

, 60

, 90

, I,$ 120

INCUBATION

I ’ I’ , 30 60

TIME

I 90

(min)

o/b

(j

,

60

-04

4-o , 90

-02 , 120

(min)

FIG. 5. Kinetics of decatenation of k-DNA. The decatenation of k-DNA was carried out and the decatenated DNA products were separated by centrifugation and quantitated either by measuring radioactivity (A. B) or fluorometrically (C, D). as described under Materials and Methods. In the case of A and B the quantity of DNA products released was calculated from the specific radioactivity of k-DNA. (A) The reactions were carried out with 0.2 ng (o), 0.1 ng (0). and 0.05 ng (a) protein of purified topoisomerase II preparation from CV-1s cells. (B) The reactions were carried out with 32 ng (0). 24 ng (0). and 16 ng (A) protein of the PEG supernatant derived from CV-1 S cells. (C) The reaction was carried out with 200-fold diluted PEG supernatant derived from the log phase mouse L cells. This corresponded to the total enzyme from approximately 660 cells in the assay. Fluorometric measurment of DNA was done at pH 8.0. (D) The reaction was carried out with 32 ng protein of the PEG supernatant derived from CV- 1S cells and DNA fluorometry was done at pH I 1.9 before (0) and after (0) heat denaturation of DNA, also indicated as b and a, respectively. A, Ratio of a and b. Inset, fluorometric measurement of k-DNA at pH 8 (0) and at pH I 1.9 before (0) and after (A) heat denaturation.

strands. The a/b ratio of the DNA products was similar to that of untreated k-DNA throughout the reaction. It should be noted that results so far have consistently shown an absence of detectable nicking (Fig. 5D), cutting

(Table IA, Figs. 3B, 4B-D, and 5D) or extensive digestion (Table IA and Fig. 3A) of kDNA by nucleases during its decatenation by crude preparations of topoisomerase II from CV- 1S cells. A similar lack of nuclease action

ASSAY

FOK

TYPE

II TOPOISOMERASES

was also observed during decatenation with crude extracts from HeLa and mouse L cells. This is probably due to (i) the presence of a smaller proportion of DNase activity relative to the topoisomerase II activity in the extracts, the former diminishing or becoming negligible after dilution of these extracts before and after their addition to the assay, and (ii) the presence of 100 mM NaCl. a concentration which significantly inhibits several mammalian DNases. A DNase activity in mammalian cells not inhibited by such NaCl concentration is maximally active at pH 5.0-5.5 but almost inactive at the pH of 7.9 that is used in our reaction (Sahai. Khan, and Stambrook. unpublished results).

Quantitative Assay sf Type II Topoisonzrrases Figure 5 shows that the amount of topoisomerase II-catalyzed decatenation products

RELATIVE

ENZYME

375

appearing in the supematant increased linearly with time after the initial appearance of 190 ng DNA and remained linear (mean linear regression coefficient, r = 0.999) until about 600 ng of DNA had been released. The slope of this linear part of the kinetic curve was a linear function of the enzyme concentration (linear regression coefficient, r = 0.999) (Fig. 6A). The plots of enzyme concentration vs slope intercept the y axis at zero or at a negative value, apparently depending on the purity of the enzyme preparation. At present we do not fully understand the cause of this difference between enzyme preparations. It is possible that at lower concentrations the enzyme in the crude state is unstable or certain constituents (such as DNA binding proteins) of the crude enzyme preparation inhibit the decatenation reaction. From the plot in Fig. 6A the activity of type II topoisomerase in a given preparation of comparable purity can be quantitated. For

CONCENTRATION

FIG. 6. Quantitation of topoisomerase II Activity. (A) Relationship of enzyme concentration to slope: Slope of the linear part of the kinetic curve of decatenation reaction (i.e., during the release of 190 to 600 ng DNA in the supernatant) was calculated. A relative enzyme concentration of 1 corresponds to 0.05 and 8 ng protein of purified (0) and crude (PEG supernatant) (0) preparation of topoisomerase II, respectively, from CV- I S cells. 16 ng protein of the PEG supernatant corresponds to the amount obtained from 660 CVIS cells. (B) Relationship of enzyme concentration to the rate of release of decatenated DNA: The decatenation reaction was carried out for 15 min with the purified topoisomerase II and the amount of decatenated DNA released was quantitated as in Figs. 5A and B. Dashed line. rate of product formation measured during the lag, i.e., prior to the release of 190 ng DNA.

376

SAHAI

AND

simplification an alternative method was also considered. Figure 6B depicts the rate of production of decatenated DNA molecules measured during the linear part of the kinetic curve (i.e., during the release of 190 to 600 ng DNA) as a function of enzyme concentration. The linear regression coefficient, Y, of this curve is 0.99 (P < O.OOl), which increased slightly in a log-log plot. This relationship can also be used for measuring the enzyme activity. In order to determine the activity one must choose a concentration of enzyme sample that yields decatenation products in the supernatant in the range of 190 to 600 ng DNA at a given time. We define one unit of type II topoisomerase activity as the amount of enzyme required to release 400 ng of decatenated DNA in the supernatant in 30 min at 30°C under these reaction conditions. As mentioned above, use of a higher concentration of k-DNA in the reaction elevates the plateau without affecting the rate of reaction, and therefore allows a substantial increase in the range of linearity with enzyme concentration. However, we have consistently observed that high concentrations of crude cell or nuclear extracts inhibit various type II topoisomerase-catalyzed reactions including catenation and decatenation reactions; this makes it advisable to use such extracts at lower concentrations. At high concentrations of crude extract, proteins such as DNA binding proteins, DNases, and also proteases may cause interference.

Advantages and Limitations

of the Assay

(i) The greatest advantage of this assay method is its rapidity and convenience. Twenty to thirty assays can be carried out in 1 h. This is due mainly to utilization of a 5min microcentrifugation instead of gel electrophoresis to separate the reaction products and to a rapid quantitation of these products. (ii) The assay is highly sensitive. Figure 6A shows the relative activity of topoisomerase II in as little as 16 ng protein of PEG supematant derived from the log phase CV-IS cells. This corresponds to enzyme from approximately 660 cells. Similar crude enzyme preparations

KAPLAN

from log phase mouse L cells (Fig. 5C) and HeLa cells (data not shown) showed comparable topoisomerase II activity. By measuring the DNA fluorometrically, it was possible to quantitate the enzyme activity in crude whole cell extracts that had been diluted to correspond to an extract from as few as 500 cells. (iii) The assay is virtually free from interference by topoisomerase I activity (Fig. 1, lanes 3 and 4; Fig. 2B, Table 1A). As much as 400 units of purified topoisomerase I activity failed to result in any detectable decatenation (Table 1B). However, the assay can be made immune to such interference by using exclusively as substrate the form I k-DNA, which contains only covalently closed circles and can easily be purified (33). (iv) The presence of 100 mM NaCl in the reaction mixture presumably offers two advantages in addition to preventing the recatenation of decatenated DNA circles (3 1): it allows an optimum turnover of the enzyme and prevents action of DNases. At lower NaCl concentrations, the rate of decatenation is reduced (unpublished results), probably because of a decrease in turnover of the enzyme rather than to increase in recatenation of the decatenated DNA circles. Although our assay is free from interference by the levels of DNase activity normally present in crude enzyme extracts (such as PEG supernatant) of mammalian cells (Figs. 3, 4B-D, and 5D), it is not immune to high levels of DNases that may be present in certain extracts. Since the fluorometric method of product quantitation allows a simultaneous detection of DNA molecules released due to nuclease action (Fig. 5D). such interference could be easily detected. One way to overcome a marginal interference would be to raise the NaCl concentration from 100 to 150 mM; this does not affect topoisomerase II activity. However, if the nuclease interference persists then the only course would be to purify the type II topoisomerase away from the interfering nuclease activity. (v) Our assay method is based upon quantitation of a majority of the reaction products (such as free DNA circles, di-, tri-, and certain multimeric catenanes), unlike the gel assays, in which usually only one of the product topoisomers

ASSAY

FOR

TYPE

is accounted for and those produced in relatively smaller proportions are ignored. (vi) Because of the nonuniformity of the topological complexity of k-DNA substrate and of the decatenation products, it is not possible to determine the absolute number of strand breaking and rejoining events by this assay. Presently none of the commonly used assays of type II topoisomerase provide an extact measure of strand breaking and rejoining events that may ideally be desirable. (vii) Since other type II topoisomerases, e.g., the DNA gyrase and T4 topoisomerase, have been shown to decatenate k-DNA and produce similar products (23) the assay method should be universally applicable to this class of enzymes.

Comparison

377

II TOPOISOMERASES A

B.

I 2 3 4 5 6 7 8 9 1011 121314

I

2 3 4 5 6 7 8 9 IO II I2 13 14

with Other ‘Assays

Other methods that have been used to assay the type 11 topoisomerase activity are (i) catenation of duplex DNA circles, (ii) unknotting of bacteriophage P4 DNA, and (iii) relaxation of supercoiled DNA which is applicable for samples free from type I enzyme activity. The lowest amount of PEG supernatant from CV1S cells or nuclei that catalyzed a dectectable catenation of PM2 DNA circles [see Ref. (22) for conditions] was about twice that required for a detectable unknotting of P4 DNA [see Ref. (20) for conditions] or decatenation of kDNA (unpublished results). This difference may be due to a lower turnover of enzyme in catenation reaction which is carried out in presence 50 mM NaCl instead of 100 mM. Although the detectable levels of these reactions occur at a comparable enzyme concentration, the quantitation of activity by catenation and unknotting assays is difficult, especially at low levels of product formation, since these depend on the scanning of the negative of gel photographs. In addition, the results of catenation assay can be influenced by the level of DNA condensing agents (such as histone H 1) usually present in crude extracts. Figure 7 presents a comparison between the relaxation and decatenation assays for the detection of purified

pg

TOPOISOMERASE

II

3 x I@

FIG. 7. Comparison of the relaxation and decatenation assays. The relaxation (A) or the decatenation (B) reaction was carried out at 30°C in 20 ~1 reaction mixture (see Materials and Methods) in the absence (lane I) or presence of (lanes 2-14) 6.25, 12.5. 25. 37.5, 50, 62.5. 75. 87.5, 100. 125. 150. 225, and 300 pg purified topoisomerase II, respectively. The reaction was terminated at 30 min and samples were electrophoresed on 0.7% agarose gel as described. The symbols indicate the form of DNA as described in Fig. 2. (C) 40 ~1 decatenation reaction mixture containing [3H]k-DNA was incubated with the indicated amount ‘of purified topoisomerase II. The reaction was terminated at 30 min. the reaction mixture was centrifuged, and decatenated molecules separating in the supernatant were quantitated as described.

topoisomerase I1 activity. The minimum amount of enzyme detected by relaxation assay is roughly two times lower than that de-

378

SAHAI

AND

tected by the decatenation assay. While 25 pg enzyme in 20 ~1 relaxation reaction mixture relaxed all the 0.25 pg PM2 DNA (Fig. 7A, lane 3) the same amount of enzyme, in 40 yl decatenation reaction mixture, releasedonly a little (52 ng) decatenated DNA (Fig. 7C). This difference may be explained by the hypothesisthat at low enzyme concentration (i.e., at early phaseof decatenation reaction) most of the decatenated DNA molecules produced are large catenaneswhich do not separatefrom the substrate by microcentrifugation and thus remain undetected. However, in addition to not being suitable for preparations containing type I enzyme, the quantitation in relaxation assay is also dependent on scanning; these limitations offset the advantage of slightly greater sensitivity of this assay.A comparison of Figs. 7B and C shows that the collective quantitation of various forms of decatenated DNA molecules by scintillometry (or by fluorometry), as used in our assay, is a more sensitive method for monitoring the early phase of the decatenation reaction than the quantitation of individual bands of product DNA distributed in the lane in gel. Type II topoisomerases have been implicated in DNA replication (9,17,34). The synthesis and consequently the activity of the procaryotic enzyme, DNA gyrase, is regulated by supercoiling of DNA in the cell (35). The activity of the eucaryotic enzyme, on the other hand, is regulated during the transition of physiologically quiescent (noncycling) cells such asperipheral lymphocytes to proliferating (cycling) cells and presumably vice versa, e.g., maturation of thymocytes into immunocompetent lymphocytes [Sahai and Kowalski, manuscript in preparation and Ref. (22)]. Becauseof the involvement of type II topoisomerasein eucaryotic DNA replication (17,34), we are studying the regulation of the activity of this enzyme and its role in structural remodeling of chromatin and nuclear matrix during mitogenic activation of lymphocytes (36). Such studies require a reliable, quantitative and sensitiveenzyme assay.The method described in this paper fulfills these requirements.

KAPLAN

ACKNOWLEDGMENTS We thank Dr. D. G. Scraba for the electron microscopy, Dr. A. R. Morgan and Dr. D. Kowalski for their valuable criticism and suggestions. and Dr. L. Simpson for providing Crifhidia .fa.sciculuta. We gratefully acknowledge the excellent technical assistance of Ms. Kathy Semple. This work was supported by grants from the Medical Research Council, Natural Science and Engineering Research Council, and an equipment grant from the Alberta Heritage Foundation for Medical Research.

REFERENCES I 2. 3. 4. 5.

6. 7. 8.

9.

Liu, L. F.. Liu, C.-C.. and Alberts. B. M. (1980) Cell 19,697-707. Cozzarelli, N. R. (1980) Cell 22, 327-328. Wang, J. C. (1971) J. Mol. Biol. 55, 523-533. Liu. L. F., and Miller, K. G. (198 I) Proc. &‘u//. .4cud. Sri USA 78, 3487-349 I. Wang, J. C., and Liu, L. F. (1979) in Molecular Genetics. Part III, (Taylor. J. M., ed.), pp. 65-88, Academic Press, New York. Tse, Yuk-thing, and Wang, J. C. (1980) Cell 22,269276. Brown, P. 0.. and Cozzarelli, N. R. (198 I) Proc. Natl. .-lead. Sci. USA 78, 843-847. Gellert, M., Mizuuchi. K.. O’Dea, M. H., and Nash. H. H. (1976) Pm. Natl. Acad. Sri. US.4 73, 38723876. Liu, L. F., Liu, C.-C., and Albcrts. B. M. ( 1979) Nature (London) 281,456-46 1.

10. Miller, K. G., Liu. L. F.. and Englund, P. T. (1981) J. Biol. Chem. 256, 9334-9339. I I. Gellert, M., Mizuuchi. K.. O’Dea, M. H., Itoh, T., and Tomizawa, J. (1977) Proc. Nat/. .4cad. Sci. USA 74,4772-4766. 12. Hsieh. T.-S., and Burtlag. D. (1980) CeN21. 115-125. 13. Hsieh, T.-S. (1983) J. Biol. Chem. 258, 84 13-8420. 14. Kreuzer. K. N.. and Cozzarelli, N. R. ( 1980) Cell 20, 245-254. 15. Gellert, M. ( I98 I) in The Enzymes (Boyer. P. D., ed.). Vol. XIV. pp. 345-364, Academic Press, New York. 16. Wang, J. C. (1985) Anmc. Rev Biochenz. 54, 665697. 17. Colwill, R. W., and Sheinin. R. (1983) Pruc. Nat/. .4cad. Sci. US.4 80. 4644-4648. 18. Osheroff. N., Shelton, E. R., and Brutlag. D. L. (1983) J. Biol. Chem. 258,9536-9543. 19. Mizuuchi, K.. O’Dea, M. H.. and Gellert, M. (1978) Proc. Natl. Acad. Sri. USA 75, 5960-5963. 20. Liu. L. F.. Davis. J. L.. and Calender. R. ( I98 I ) Nucleic Acids Res. 9, 3979-3989. 21. Baldi. M. I., Benedetti. P.. Mattoccia. E., and TocchiniValentini, G. P. (1980) Cell 20,461-467.

ASSAY

FOR

TYPE

22. Tricoli, J. V.. Sahai. B. M., McCormick. P. J.. Jarlinski, S. J., Bertram, J. S.. and Kowalski. D. ( 1985) Exp. Cell Res. 158, 1-14. 23. Marini, J. C.. Miller, K. G.. and Enghmd. P. T. ( 1980)

J. Biol. <‘hem. 255, 4916-4979. 24. Simpson, L. (1979) Proc Natl. Acud. Sci. l.iS.4 76, 1585-1588. 25. Kowalski, D. (1979) ,4nai. Biochern. 93, 346-354. 26. Davis, R. W., Simon. M.. and Davidson, N. (197 I) in Methods in Enzymology (Grossman. L.. and Moldave. K.. eds.), Vol. 2 1, pp. 4 13-428. Academic Press, New York. 27. Sutcliffe, J. G. (1979) Cold Spring Hararhor S)mp.

Qttant B&i. 43, 77-90. 28. Simpson, L. (1972) Inr. Rev Cyiol. 32, 139-207. 29. Borst. P., Fairlamb. A. H., Fase-Fowler. F.. Hoeijmakers. J. H. J.. and Weislogel. P. 0. ( 1976) in

379

II TOPOISOMERASES

The Genetic Function of Mitochondrial DNA (Saccone. C.. and Kroon. A. M., eds.), pp. 59-69, Elsevier/North-Holland. Amsterdam. 30. Sanger. F.. Coulson, A. R., Hong, G. F., Hill, D. F., and Petersen, G. B. (1982) J. Mol. Biol. 162, 729773. 3 I. Krasnow, M. A.. and Cozzarelh, N. R. ( 1982) J. Biol.

Chem. 257.2687-2693. 32. Morgan,

A.

R.,

and

Pulleyblank,

D.

E. (1974)

Biochern. Bioplzvs. Res. Commun. 61, 396-403. 33. Englund, 34. DiNardo.

P. T. (1978) S., Voelkel.

Cell 14, 157-168. K.. and Sterngianz. R. (1984) Proc. Natl. Acad. Sci. US.4 81, 26 16-2620. 35. Menzel, R., and Gellert, M. (1983) CeN34, 105-I 13. 36. Setterfield, G., Hall. R., Bladon, T., Little, J.. and Kaplan, J. G. (1983) J. C’lfrastruct. Rex 82, 264282.