ICANCER RESEARCH 58. 1260-1267. Mardi 15.

Mutations of the Bacteriophage T4 Type II DNA Topoisomerase That AlterSensitivity to Antitumor Agent 4'-(9-Acridinylamino)methanesulfon-m-anisidide and an Antibacterial Quinolone1

Catherine H. Freudenreich,2 Christine Chang,3 and Kenneth N. Kreuzer4

Department of Micriibiiilitgy. Duke University Medical Center. Durham. North Carolina 27710

ABSTRACT

Various antitumor and antibacterial agents target type II DNA topoi-

somerases, stabilizing a cleaved DNA reaction intermediate and therebyconverting topoisomerase into a cellular poison. Two 4'-<9-acridinylami-

no)methanesulfon-m-anisidide (m-AMSA 1-resistant hacteriophage T4 to-

poisomerases have previously been characterized biochemically, and wehave now determined the sequence of the causative mutations. In one case,a mutation (K457K) in a conserved domain of gp39 (ATPase subunit)causes resistance to antitumor agent m-AMSA but hypersensitivity to the

quinolone oxolinic acid. In the second case, a combination of two aminoacid substitutions (S79F and G269V) in gp52 (DNA-cleaving subunit)causes resistance to both m-AMSA and oxolinic acid. The S79F mutation

is responsible for drug resistance, whereas the G269V mutation suppresses a topoisomerase deficiency caused by S79K. Surprisingly, the

(•269Vmutation by itself causes a dramatic hypersensitivity to bothinhibitors, defining a new class of topoisomerase mutants. Because S79and the adjacent N78 are homologous to two key residues of DNA gyrasethat affect quinolone sensitivity, we generated additional amino acidsubstitutions at these two positions. The substitutions alter sensitivity tom-AMSA and to oxolinic acid, sometimes in opposite directions. Further

more, the quinolone sensitivities of the various mutants paralleled those ofcorresponding gyrase mutants. These results support models in whichboth quinolones and antitumor agents bind to a conserved site thatoverlaps the active site of the enzyme.

INTRODUCTION

Type II DNA topoisomerases catalyze DNA interconversions bymaking a double-strand break in one segment of DNA, passing

another segment through the break, and religating the cleaved DNA(reviewed in Refs. 1-3). A key intermediate in these reactions, the

cleavage complex, consists of topoisomerase covalently attached viaphosphotyrosine bonds to the 5' phosphates of the broken DNA. The

cleavage complex is stabilized by numerous antitumor and antibacterial agents that inhibit type II topoisomerases. Cytotoxicity of thesechemotherapeutic agents is critically dependent on the stabilizedcleavage complex, indicating that these agents convert a normallybeneficial cellular enzyme into a destructive poison (4-6).

Recent results strongly suggest that the topoisomerase inhibitorsstabilize the cleavage complex by binding to DNA at the active site ofthe enzyme. For example, the identity of the base pair that immediately flank the scissile phosphodiester bonds was shown to determinewhich chemical families of inhibitors can stabilize a cleavage complex at a particular sequence (7, 8). More directly, a photoactivatible

Received 9/16/97; accepted 1/16/98.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 with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported by Grant CA60836 from the National Cancer Institute." Present address: Department of Molecular Biology, Princeton University. Princeton.

NJ (1X544.' Present address: Department of Medicine. University of Pittsburgh Medical Center.

Pittsburgh. PA 15213.4 To whom requests for reprints should be addressed, at Department of Microbiology.

Duke University Medical Center, Box 3020. Durham. NC 27710.

analogue of the antitumor agent m-AMSA? was cross-linked to the bp

adjacent to the scissile bond within a cleavage complex, but cross-

linking was not detected to other bp within the substrate (9).Although the above results indicate that the inhibitors are in close

contact with the DNA. the enzyme also clearly plays an important rolein drug binding: (a) in the cross-linking study just mentioned, them-AMSA derivative reacted with DNA only when topoisomerase was

present; (b) a direct analysis using radiolabeled quinolone demonstrated drug binding to a gyrase-DNA complex but not to either gyrase

or DNA alone (10, 11); (c) the spectrum of sensitivities to particularclasses of topoisomerase inhibitors is species dependent (e.g.. quinolones preferentially inhibit bacterial DNA gyrase whereas variousantitumor agents inhibit the human topoisomerase); and (d) muta-

tional alteration of a type II topoisomerase can lead to drug resistance.In Escherichia coli DNA gyrase. a high level of resistance is conferredby amino acid substitutions in the "quinolone-resistance determiningregion" of the GyrA subunit (residues 81-87. particularly S83: Refs.

12 and 13). Antitumor drug resistance mutations have also beenanalyzed in eukaryotic topoisomerases. and some are within or verynear this same conserved region (see "Discussion").

The bacteriophage T4 type II topoisomerase is sensitive to many ofthe same antitumor agents that inhibit the mammalian enzyme, including m-AMSA, ellipticines, mitoxantrone, and the epipodophyllo-toxin VP-16 (14, 15). At least in the case of m-AMSA, the topoi

somerase is the physiological target in this model system because amutation in the topoisomerase is sufficient to provide drug resistance(16, 17). The phage-encoded enzyme is also weakly sensitive to

oxolinic acid, a quinolone that inhibits bacterial DNA gyrase (18, 19).The facile genetics, biochemistry, and molecular biology of the phageT4 system have made this a productive model system for analyzingthe mechanism of action of these topoisomerase inhibitors (20).

Two m-AMSA-resistant T4 mutants have been previously isolated

and analyzed by genetic mapping. One was shown to carry a mutationin gene 39. which encodes the ATPase subunit of the enzyme (homologous to gyrase subunit B), whereas the other was mapped to gene52, which encodes the breakage-resealing subunit (homologous to

gyrase subunit A; Refs. 16 and 17). Topoisomerase purified from eachof the resistant phage mutants exhibited m-AMSA-insensitive DNA

relaxation and cleavage activities ( 16, 17). Furthermore, both purifiedenzymes demonstrated cross-resistance to other antitumor agents,

arguing for a common mechanism of inhibition and overlappingdrug-binding sites (15). Interestingly, the gene J9-mutant enzymedisplayed hypersensitivity to oxolinic acid and VP-16, suggesting a

link between the mechanism of action of antibacterial and antitumoragents (15). This and other results also argue that the drug-resistance

mutations subtly alter the drug binding site rather than simply destroying it.

In this communication, we report the precise amino acid substitutions in the two mutant enzymes and demonstrate that these substitutions are sufficient for m-AMSA-resistant phage growth. In addition,

5 The abbreviations used are: m-AMSA. 4'-(9-acridinylamino)melhanesulfon-m-anisi-

dide: VP-16. etoposide; I/S. insertion/substitution; MDE. mutation detection enhancement.

1260

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOIlìlSOMIiRASE MUTATIONS

we generate and analyze additional mutations in the quinolone-resist-

ance determining region that affect sensitivity to both the antitumordrug m-AMSA and the quinolone oxolinic acid. Particularly in conjunction with recent progress in elucidating the three-dimensional

structure of type II topoisomerases (21, 22), these results provideimportant insight into inhibitor-enzyme interactions relevant to both

bacterial and eukaryotic type II topoisomerases.

MATERIALS AND METHODS

Strains. E. coli strains MCS1 (supD araD139 A(ara-/e«)7697A(/ac)X74ga/E galK rpsL pro lisdK), BE-BS (nonsuppressing). and CR63 (supD) are

described by Kreuzer et al. (23). E. coli strain ABI 1571. which is a derivativeof AB 1157 (24), has a transposon insertion in the acrA gene that causesincreased permeability to m-AMSA.6 E. coli strain 1643 (F~ ihr-1 leuB6 lhi-1

lacYl supE44fliuA21 gyML83 parCK84 TnlO-tel Tn\Q-kan) is extremely re

sistant to quinolones (25) and thus allows visualization of the T4 topoisomer-ase-dependent oxolinic acid sensitivity of a phage T4 infection. E. coli NapIV(supD). which was used to test the topoisomerase-negative phenotype, was

obtained from L. Gold (University of Colorado, Boulder, CO; Ref. 26). T4strain T4+D (wild-type) was originally from the collection of B. M. Alberts(University of California, San Francisco, CA). T4 J9-AR (m-AMSA resistance

mutation in gene 39: otherwise wild type) is described by Huff el al. (16), andT4 52-AR (m-AMSA resistance mutation in gene 52; otherwise wild type) is

described by Huff et al. (17). T4 strain KIO carries the following mutations:amB262 (gene 38), amS29 (gene 5/), nd28 (gene denA), and rllPTS (rll-denB

deletion; Ref. 27).Media and Inhibitors. L broth contained 10 g/l NaCl, 10 g/1 Bacto-

Tryptone, and 5 g/l yeast extract. Hershey agar plates contained 13 g/l Bacto-

Tryptone. 8 g/l NaCl. 2 g/l sodium citrate, 1.3 g/l glucose, and 10 g/l agar.m-AMSA (NSC 249992) was provided by the Drug Synthesis and Chemistry

Branch, National Cancer Institute; it was dissolved (50 mg/ml) in DMSO anddiluted in water just prior to use. Oxolinic acid (Sigma Chemical Co.) wasdissolved (10 mg/ml) in 100 niM NaOH and diluted in water just prior to use.

Detection of Mutations by MDE Gels. Four overlapping gene 39 fragments were amplified from either T4+D or 39-AR phage by PCR (see Fig. 1).The PCR products from T4+D, 39-AR, or a mixture of the two were purified

over a CL6B column and then subjected to electrophoresis through an MDEgel (AT Biochem) with 0.6X TBE running buffer (1X = 89 min Tris base. 89

mM boric acid, and 2 mM Na, EDTA) at 20 V/cm for about 25 h.Introducing Mutations into T4 Phage. Mutations were generated in plas-

mids and then introduced into the phage genome by homologous recombination using the 1/S system (27. 28). To introduce the E457K mutation into gene39 and the G269V mutation into gene 52, an appropriate PCR fragment wasgenerated with the desired mutation near the center. Each fragment was clonedinto the polylinker of pBSPLO" and sequenced to confirm that the fragment

contained the desired mutation and no others. The mutation was then transferred to the phage genome as follows. T4 strain KIO (which has two ambermutations) was grown on plasmid-bearing cells, and recombinant phagescarrying integrated plasmid were selected by virtue of the plasmid-borne sitpF

gene. The inserted plasmid was allowed to segregate back out of the phagegenome by homologous recombination, leaving behind the engineered topoi-somerase mutation in a fraction of the segregants (—50% contain the desired

mutation when the mutation does not adversely affect phage growth; Refs. 27and 28). Segregants carrying the gene 39 E457K mutation were identified bytheir resistance to m-AMSA. Segregants carrying the gene 52 G269V mutation

were identified by PCR amplification of the mutated region followed byrestriction enzyme treatment; the G269V mutant fragment contains a Bbsl sitethat is not present in the wild-type fragment.

For the mutations in codons 78 and 79 of gene 52. the intact 5' end of the

gene 52 coding sequence was first PCR amplified and cloned between thepolylinker Clal and Kpn\ sites of pBSPLO . The inserted fragment containsthe first 220 bp of gene 52 coding sequence, with the last 6 bp being a newlyintroduced Kpnl site (replaces 5'-GATATC-3' with 5'-GGTACC-3'; the two

mutations are silent with respect to the coded protein). The sequence of thisinsert was confirmed by DNA sequencing. A second round of PCR amplifi-

6 J. George and K. N. Kreuzer, unpublished observations.

cation generated a fragment with the same Kpn\ site followed by the next 128bp of gene 52. The lasi 6 bp of this fragment was a newly introduced Xha\ site(replaces 5'-AGTCGT-3' with 5'-TCTAGA-3'; the mutations are silent with

respect to the coded protein). The newly introduced Kpn\ and Xhn\ sites weredesigned to facilitate this and future efforts to introduce mutations in gene 52.The second round of amplification was performed with each of three differentprimers overlapping the Kpnl site: FKpn\ ( 1), which incorporates the S79Wmutation (TCT to TGG at the DNA level); FKpnl (2), with mixed bases atcodon 79 (TCT to (T/C)(T/C)T) to generate S79 (wild type), S79F, S79L andS79P; and FKpnl (3), with mixed bases at codon 78 (AAC to T(G/C)(T/G)) togenerate N78S. N78W. and N78C. After cloning these second PCR fragmentsinto the plasmid containing the first fragment, the individual mutant plasmidswere identified by DNA sequencing (no other mutations were present in any ofthe cloned fragments). The mutations were then substituted into the genome ofT4 strain KIO using the I/S system essentially as described above. Segregantswere first tested for the possible presence of the desired mutations by screeningfor a change in the level of sensitivity to either m-AMSA or oxolinic acid. This

resulted in the identification of N78S. N78W. N78C. and S79L but none of theother S79 substitutions. Approximately 60 segregants from each of the constructions with the other three S79 substitutions were then screened for atopoisomerase-negative phenotype (failure to grow at 25°C). We found one

topoisomerase-negative segregan! with the S79F construction but not with the

S79P or S79W constructions. The S79W mutation eliminates a Bsgl sitepresent in the wild-type sequence, and no segregan! phages (of 20) were found

without this site (tested by PCR). Because S79W mutants were not found usingthis direct physical assay, it is very likely that S79W results in a topoisomerase-negative phenotype (topoisomerase-negative mutants are selectively lost in

the plasmid segregation step due to poor growth). Every isolated mutant phagewas confirmed either by DNA sequencing or by detecting the gain of anexpected restriction site in an appropriate PCR fragment.

Gradient Plate Tests of Drug Sensitivity. Gradient plates were made bypouring 25 ml of drug-containing Hershey agar into a square Petri plate with

one edge of the plate resting on a pencil. After the agar hardened, the plate wasleveled, and 25 ml of drug-free Hershey agar were poured on top of the first

layer. Thus, one edge of the plate had a drug concentration near zero, and theopposite edge had a concentration close to that of the bottom layer of agar. Asuspension of —5X 10" bacterial cells, either ABI 157I (for m-AMSA plates)

or E. coli 1643 (for oxolinic acid plates), was mixed with 5 ml of Hershey topagar and poured onto the surface of the gradient plate. After the top agarsolidified, a diluted suspension of the indicated phage strain was spotted sixtimes across the gradient and once onto a drug-free control plate. Each spotcontained 3 /xl with —160 plaque-forming units (dilutions were made and

titered on the previous day and then adjusted to accurately match the desiredphage concentration). The plates were incubated overnight at 37°C.

RESULTS

Mapping and Reintroduction of the 39-AK Mutation. Genetic

analyses indicated that one mutation conferring m-AMSA resistance

maps within gene 39, which encodes the ATPase subunit of T4topoisomerase (16). To narrow down the region containing this 39-A.R

mutation, we used a high-resolution, nondenaturing MDE gel system.

MDE gels are often able to resolve two duplex DNA fragments thatdiffer by a single bp and can often separate a DNA homoduplex froma heteroduplex (wild-type/mutant) containing a single mismatch. PCR

primers were designed to amplify four overlapping fragments thatspanned the length of gene 39 (Fig. 1, top). For each of the fourfragments, the PCR product amplified from wild-type DNA, from the39-AR mutant phage DNA. and a mixture of the two were subjected

to MDE gel electrophoresis. In addition, a mixture of each pair ofPCR fragments was heat denatured and renatured to look for possibleheteroduplexes with altered migration. None of the four mixturesproduced heteroduplexes with altered migration, and there was nodetectable difference in the migration of wild-type versus mutantfragment A. B. or C (Fig. 1). However, the wild-type and ^9-AR

fragment D showed a clear migration difference, revealing the presence of a mutation near the 3' end of the gene (Fig. 1). MDE gels do

1261

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOPOISOMI-.RASI-. MUTATIONS

lili1 iiiili1 i !!gene

39ACBD

34 1 34 1234 1234

500 -

400 -

300 -

Fig. l. Localization of the 39-A" mutation. As indicated in the gene diagram at the lop,four PCR fragments of gene 39 were amplified from the wild-type and 39-AK phage.

MDE gels of each fragment are labeled with the fragment name (A-D). In each case. Lane1 contains the wild-type fragment. Lane 2 contains a mix of wild-type and .Õ9-AR

fragments. Lane 3 contains a heal-dcnatured/reannealed mix of the two fragments, andLane 4 contains the .?9-AR fragment. Lt'ft, size markers (low molecular weight biomarkers

from Bio Ventures). A photograph of the clhidium bromide-stained gel is shown.

not detect all possible mutations, and therefore we cannot concludethat 39-AR fragments A, B, and C are free of mutations.

DNA sequence analysis of PCR fragment D from the ,ÃŽ9-ARmutantphage revealed a single mutation within the gene 39 coding region.7

An A-to-G transition at nucleotide position 1369 would substitutelysine for glutamic acid-457 (E457K) in gp39 of the mutant phage.

This amino acid substitution should cause a net +2 charge alterationin gp39, consistent with previous biochemical characterization of thepurified enzyme (16).

To determine whether the E457K substitution is responsible for them-AMSA resistance phenotype of the .?9-AR phage, the mutation was

substituted back into an essentially wild-type T4 using the I/S system(27, 28). A 223-bp fragment containing either completely wild-typesequence or the single A-to-G transition (near the middle of the

fragment) was cloned into the 1/S plasmid. This plasmid contains amarker (supF) that can be used to select for the phage-plasmid

cointegrates that are formed by homologous recombination in gene39. Several cointegrates with either the wild-type or mutant plasmid

were then propagated under nonselective conditions, which allowssurvival of phages that have lost the integrated plasmid by means ofa second homologous recombination event (between the duplicatedgene 39 sequences of the cointegrate). Because the mutation was nearthe center of the 223-bp fragment, approximately one-half of the

7 In addition, an error in the published sequence (48) was detected. The reported C at

position 1539 was absent in both the mutant and wild-type fragment D, and therefore IheCOOH terminus of gp.W should be RKNL rather than RKKSIMT. Additional sequencechanges in other parts of gene 3V are incorporated in the revised T4 sequence database (E.Kutler. personal communication).

resulting segregan! phage from the mutant plasmid should carry theE457K substitution. Indeed. 41 of 80 segregants from the mutantplasmid displayed m-AMSA-resistant phage growth, whereas none of80 segregants from the wild-type plasmid grew in the presence of thedrug. The drug-resistant segregant phage clones had the same pheno-types as the original .?9-AR mutant, i.e.. a strong m-AMSA resistance

along with a marked hypersensitivity to the quinolone oxolinic acid(the phenotypes are shown in Fig. 4; see Refs. 15-16 for descriptionof original :?9-AR mutant). We conclude that the E457K substitution

is both necessary and sufficient for resistance to m-AMSA and hy

persensitivity to oxolinic acid.Analysis of Sequence Changes in 52-AR Mutant. Genetic anal

ysis of a second m-AMSA-resistant phage mutant indicated that drug

resistance is caused by a mutation within gene 52, which encodes thecleavage-resealing subunit (17). MDE gels were again used to survey

four overlapping PCR fragments from the gene for sequence changes;two of the fragments showed migration differences (data not shown).We decided to sequence the entire gene 52. using either 52-AR-mutant

or wild-type DNA as the template. In total, four nucleotide substitutions were found in the 52-AR-mutant DNA, and several mistakes inthe published sequence were uncovered.8 Two of the four mutationsin the 52-AR DNA (nucleotide G1083 to A and C1305 to A) were in

third positions of codons and silent with respect to amino acid coding.The other two mutations, nucleotide C236 to T and G806 to T, wouldcause S79F and G269V amino acid substitutions in gp52. The alteration S79F was particularly provocative, because this substitution fallswithin the quinolone-resistance determining region defined for bacterial DNA gyrase subunit A (see "Introduction"). As shown in Fig. 2,

the S79 residue of gp52 aligns with A84 of E. coli DNA gyrasesubunit A, a residue that is found mutated to proline in severalquinolone-resistant bacterial mutants (12, 29, 30). The immediately

adjacent residue in gyrase, S83, is often found mutated to leucine ortryptophan in bacterial mutants that are highly resistant to quinolones(13).

Mutational Analysis of T4 gp52 Residues N78, S79, and G269.Given the interesting position of the S79F mutation in the 52-AR

strain, we substituted this mutation back into a clean genetic background using the I/S system, essentially as described above for theE457K mutation in gene 39. In addition, we introduced other potentially informative amino acid substitutions at positions 78 and 79(equivalent to E. coli GyrA positions 83 and 84). The genome substitutions are described in detail in "Materials and Methods."

The phenotypes of the mutant phage strains were determined bygrowth on gradient plates containing an appropriate bacterial host andeither no drug or increasing concentrations of m-AMSA or the quin

olone oxolinic acid (Fig. 3). Equal volumes of a dilute phage suspension were spotted in six drops across the gradient plate, and growthwas allowed overnight at 37°C.The bacterial host for the tests of

m-AMSA sensitivity was an acrA mutant, which apparently accumulates much higher intracellular levels of m-AMSA due to a defect ina multidrug efflux pump (31, 32).6 The host for oxolinic acid testing

was the bacterial strain 1643. which contains quinolone-resistance

mutations in both GyrA (S83L) and ParC (E84K; ParC is the GyrAhomologue of bacterial topoisomerase IV; Refs. 25 and 33). Theoxolinic acid sensitivity of phage strains could be determined with thisvery quinolone-resistant bacterial strain because the bacterial lawn

was unaffected by high levels of the quinolone.

8 Sequencing of thà w¡Id-typegene 52 revealed three insertions and two base changes

from the published sequence (49). These changes result in a correction of the amino acidsequence from the published TLHKSQC (positions 78-84) to NSAQDAGA (positions78-85). Note that this change improves Ihe match of gp52 to the family of type II

lopoisomerases by aligning S79 with a highly conserved position occupied by either S orA in other members of the family (see Fig. 2).

1262

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOPOISOMÃœRASE MUTATIONS

T4 gp52 (75-88): H G E

EC GyrA (80-93): H G D

Se TopII (737-750): H G E

Hs TopII (759-772): H G E M71

S9 AQDAGALMA

A84 VYDTIVRMA

S741 LAQTIIGLA

S763 LMMTIINLA

Fig. 2. Sequence alignment of type II topoisomerases. The indicated segments of T4gp52. £.culi gyrase subunit A (£<•GyrA), Saccharmnyces cerevisiue type II topoisomer-

ase (Sc TopII). and human type II a topoisomerase (Hs TopII) were aligned according toCarónand Wang (36). except that the corrected sequence of T4 gp52 was used. Note thata different alignment was proposed by Liu et al. (44), with S74I of the yeast enzymealigning with S79 of bacterial gyrase. This requires misaligning the two enzymes such thattwo nearby single amino acids are deleted, whereas the alignmeni shown in this figurerequires no misalignment in this region.

The phage strains with mutations at position 78 of gp52 will bediscussed first. One mutation replaced the asparagine with serine(N78S). thereby mimicking the bacterial GyrA residue (serine 83)known to be important in quinolone sensitivity. In addition, a trypto-

phan substitution (N78W) was generated because the bacterial GyrAsubstitution S83W leads to a high level of quinolone resistance.Finally, a cysteine substitution (N78C), which has no precedent in theliterature, was easily obtained due to the method used in the construction.

Growth of the wild-type control phage was strongly inhibited, butnot abolished, at the high concentration end of the m-AMSA ( 1

/xg/ml) gradient plate (Fig. 3A). All three N78 mutants were hypersensitive to m-AMSA, with N78W being the most sensitive (Fig. 3/4).

The oxolinic acid phenotypes of the N78 mutant phage were evenmore interesting (Fig. 3ß).The N78S mutant showed a markedhypersensitivity to the quinolone. Thus, by substituting the corresponding residue from DNA gyrase, we have converted the T4 topoisomerase into a more gyrase-like (i.e., quinolone-sensitive) enzyme.

Furthermore, substitution at this position with tryptophan. whichcauses high-level quinolone resistance in gyrase. also caused high-

level oxolinic acid resistance in T4 (Fig. 3ß).The equivalent of agyrase S83W mutant can be imagined by comparing the N78S withthe N78W mutant of T4, and the increase in oxolinic acid resistanceis very dramatic. Finally, the T4 gp52 N78C substitution also showeda high level of oxolinic acid resistance.

Turning to position 79, we attempted to generate phage carryingphenylalanine (S79F; found in the 52-AR phage), proline (S79P;

proline substitutions at the corresponding A84 of bacterial gyrasecause quinolone resistance; see above), leucine (S79L). and tryptophan (S79W). Unlike the above experiments with position 78 substitutions, we had difficulty isolating phages with substitutions at position 79, although the plasmids used in the I/S system had the desiredmutations by DNA sequence analysis (see "Materials and Methods").

The only substitution that we obtained easily was S79L, which causedresistance to both m-AMSA and oxolinic acid (Fig. 4, A and Q.

From the S79F construction, we did not find phage segregants withnormal growth and an obvious drug-sensitivity or -resistance pheno-type. We therefore searched for putative topoisomerase-negative mu

tants among the segregan! phage by preferentially picking smallplaques. We found one segregant (of 60) that behaved like a topoisomerase-negative mutant (see below). The S79F mutation shouldeliminate a Bsgl restriction site in the wild-type DNA sequence, and

digestion of a PCR product from this segregant revealed the expectedloss of the site. Furthermore, DNA sequencing confirmed that thisphage had the S79F mutation (and no other mutations in the regionsequenced). Topoisomerase-negative mutants of phage T4 have adistinct phenotype of growing poorly at 37°Cand not at all at 25°C

(34). The S79F mutant phage displayed this characteristic phenotype(Fig. 5), arguing strongly that this mutation greatly reduces or abolishes topoisomerase activity in vivo.

Because the original 52-AR mutant phage did not show a topoi

somerase-negative phenotype, we then suspected that the G269V

mutation also present in that phage suppresses the defect caused bythe S79F mutation. To analyze this possibility, we first substituted theG269V mutation back into a clean genetic background. The desiredsegregant was recognized by means of a new Bbsl restriction sitecaused by the mutation. The single G269V mutant itself has a slightgrowth defect at 25°C(but normal growth at 37°C),suggesting that

this mutation reduces (but does not abolish) topoisomerase activity(Fig. 5). We next used the relatively healthy G269V mutant to rescuethe S79F mutation from a plasmid. As expected from the suppressionmodel, the reconstituted double mutant phage had little or no growthdefect, actually growing better at 25°Cthan either single mutant (Fig.

5). Clearly, the severe growth defect caused by the S79F mutation isstrongly suppressed by the G269V mutation (see "Discussion").

The drug sensitivities of both single mutants and the double S79FG269V mutant were compared using m-AMSA and oxolinic acid

gradient plates (Fig. 4). The double mutant was strongly resistant toboth drugs, which is the same phenotype as the original 52-AK mutant

(15, 17). Thus, the phenotype of the original mutant is reconstituted bythe combination of G269V and S79F. As described above, the S79Fsingle mutant has a strong growth defect even in the absence ofinhibitors. Mutants that lack topoisomerase activity display a drug-

resistant phenotype, presumably because they do not accumulate

A 1

oxolinic acid, ug/inl

Fig. 3. Drug sensitivities of N78 substitutions. The sensitivity of T4 strain KIO, whichis wild-type for topoisomerase (wt), is compared with that of KIO derivatives carrying theN78S, N78C, and N78W substitutions. The gradient plate in A contains m-AMSA(maximum concentration, I ¿ig/ml),whereas the plate in H contains oxolinic acid (maximum concentration. l(X) jxg/ml). The strip on the righi of each panel is a drug-freecontrol. As described in "Materials and Methods," each of the six spots across the gradient

plates contain —160 plaque-forming units, and the effectiveness of ihe inhibitor can be

judged by the inhibition of phage growth as the drug concentration increases across thegradient.

1263

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

14 TOl'OISOM! RASI NUTATIONS

m-AMSA, ng/ml oxolinic acid, fig/ml

I 39:1 E457K

wt

"52:] S79L

S79F-G269V

S79F

G269V

52: S79L

S79F-G269V

S79F

G269V

B m-AMSA, ng/ml

*

G269V

wt

oxolinic acid, ng/ml

I«

Fig. 4. Drug sensitivities of gene 39 E457K substitution and gene 52 substitutions. The sensitivity of T4 strain K10. which is wild-type for topoisomerase Of/), is compared withthat of K10 derivatives carrying the E457K substitution in gene JV or the S79L. S79F. G269V. or S79F-G269V substitution in gene 52. The gradient plate in A contains m-AMSA(maximum concentration. 3 /¿g/ml).and the plate in B contains a shallower gradient of m-AMSA (maximum concentration. 1 /ig/ml) to highlight the sensitivity of the G269V mutant.The gradient plate in C contains contains oxolinic acid (maximum concentration. 100/ig/ml), and the plate in D contains a shallower gradient of oxolinic acid (maximum concentration.20 /ig/ml) to highlight the sensitivity of the G269V mutant. The strip on (he rialti of each panel is a drug-free control. For unknown reasons, the S79F mutant grows slightly betteron the bacterial host for oxolinic acid tests (strain 1643) than it does on the host for m-AMSA tests (strain AB! 157I).

drug-induced cleavage complexes (35) and, therefore, the apparent

drug resistance of the S79F mutant (Fig. 4) could simply be the resultof the topoisomerase defect. Surprisingly, the G269V single mutantwas extremely hypersensitive to both m-AMSA and oxolinic acid

(Fig. 4). By comparing the G269V single mutant and the G269V S79Fdouble mutant, it is clear that the S79F mutation causes a very strongdrug resistance in the context of a topoisomerase with the G269Vmutation.

25°C 37°C

§oe o oo o pò

wt í¿

G269V

S79F

S79F/G269V

Fig. 5. Topoisomerase-negative phenolype of S79F mutant. Aliquots (3 /ni) containing~ 1000, I(K),or 10 plaque-forming units of either K10 (u7) or the indicated mutant of K10were spotted on two (drug-free) Hershey agar plates with lawns of NapIV (mipD} bacterialcells. The plates were incubated at the indicated temperature. Topoisomerase-negativemutants fail to grow at 25°Cand grow poorly at 37°C.Control platings with topoisomer

ase amber mutants (and a NapIV nonsuppressing host) showed a phenotype identical tothat of the S79F mutant (data not shown).

DISCUSSION

We will use the recently published crystal structures of the yeasttype II topoisomerase (21) and bacterial gyrase subunit A (22) as aframework for discussing drug-resistance mutations. This approach is

speculative, because the T4 enzyme could have significantly differentstructural features. Nonetheless, comparison with the published crystal structures seems warranted: (a) the overall features of the yeast andbacterial topoisomerase structures are quite similar, although they doappear to represent two different enzyme conformations; and (h) alltype II topoisomerases share extensive homologies throughout mostof the protein sequence (36), arguing for a generally conserved structure for this family of enzymes.

Both bacterial and eukaryotic drug-resistance mutations are

clustered in two distinct regions of the type II topoisomerases(roughly, residues 439-492 and 726-780 with respect to the yeast

coordinates), leading to the suggestion that these regions togetherform a drug-binding pocket (37, 38). The T4 ,?9-AR (E457K) and52-AR (S79F) mutations each fall within one of these regions.

These two regions roughly face each other, but are quite far apart,in the crystal structure of the yeast enzyme (21). One region iswithin the A' domain, very near the active site tyrosine (yeast

Y783) and within a CAP-like structure thought to bind the cleaved

DNA strand. The other region is within a different domain of theprotein (B') that is connected by a flexible linker to the A' domain.Thus, a reasonable (but untested) scenario is that the B' domain

folds down upon the DNA-binding region during the reactioncycle, juxtaposing the two regions of drug-resistance mutations

1264

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOmiSOMI.KASE MUTATIONS

around the bound DNA molecule, where the corresponding residues could interact with the drug molecule.

The mutation cluster in the B' domain contains two conserved

oligopeptide motifs. EGDSA and PLRGKMLN. which are altered bymany of the known drug-resistance mutations (38). The mutation inthe J9-AR phage (E457K; E495 in the yeast enzyme) is 14 amino

acids downstream from the second motif. This position has a highlyconserved glutamic acid residue in nearly all type II topoisomerases.To our knowledge, no other drug-resistance mutations have been

reported at this position, but mutations in nearby residues have beendetected previously (39-42).

Several results suggest that this region of the B' domain interacts

directly with the topoisomerase inhibitors. Mutations in this region ofE. coli DNA gyrase can differentially affect sensitivity to differentmembers of the quinolone family (37. 43), and the E457K mutation inT4 topoisomerase likewise differentially affects sensitivity to members of the ellipticine family of inhibitors (15). Furthermore, theE457K mutation causes resistance to some antitumor agents buthypersensitivity to oxolinic acid and VP-16 (15). These results indicate that the drug-resistance mutations subtly alter the drug binding

site to interfere with the binding of some inhibitors but favor thebinding of others. It is conceivable that these differential effects couldbe obtained by conformation changes that alter a more remote drugbinding site, but it seems more likely that the relevant interactions aredirect and local.

The second region important for inhibitor sensitivity is the putativeDNA-binding region in the A' domain close to the active site tyrosine.

Essentially every quinolone-resistance mutation in bacterial GyrA or

Part maps within this region, with the most commonly found mutations altering GyrA residues G81, S83, A84, D87, and Q106 (yeastresidues G738, 740, 741, 744, and A767; Fig. 2). The S83W mutationblocks the binding of radiolabeled quinolone to the enzyme-DNA

complex (11), and, therefore, mutations in this region almost certainlycause drug resistance by directly destabilizing the interaction betweendrug and enzyme.

Several antitumor-drug resistance mutations in human or yeast type

II topoisomerase also map to this general region of the enzyme (41,42, 44-46). although the mutations do not cluster as intensely as withthe bacterial enzymes. To our knowledge, only two well-characterizedeukaryotic drug-resistance mutations map to the region correspondingto the very strong hotspot for bacterial gyrase drug-resistance mutations (residues 83-87). These are G738D (GyrA position G81) and

S741W (GyrA position A84; see Fig. 2 legend) in the yeast topoisomerase, both of which alter the sensitivities to fluoroquinoloneCP-115,953 and etoposide (but not m-AMSA; see Refs 42, 44, and

45).Our results have demonstrated the importance of T4 gp52 residues

N78 and S79 in sensitivity to antitumor drugs and quinolones. TheN78S substitution caused an increased sensitivity to m-AMSA and to

oxolinic acid, whereas the N78C and N78W substitutions causedhypersensitivity to m-AMSA but marked resistance to oxolinic acid(Fig. 3). The S79L substitution caused resistance to both m-AMSA

and oxolinic acid, as did the S79F substitution when tested in thecontext of the G269V mutation (see below).

The crystal structures of the yeast type II topoisomerase and E. coligyrase subunit A greatly clarify the importance of this region of theprotein in drug sensitivity (21, 22). The yeast residues equivalent toT4 gp52 N78 and S79 (Q740 and S741) are within A'a4, the second

helix of a helix-turn-helix structure that is similar to the DNA-binding

region of CAP and histone H5. Although there are no cocrystals of atype II topoisomerase and DNA. these structural homologies allowedBerger et al. (21) to model the topoisomerase-cleaved DNA into thisregion. In the topoisomerase-DNA model. A'a4 contacts the major

groove of the segment of DNA just outside the cleaved phosphodi-

ester bond. According to this model, residues Q740 and S741 of theyeast enzyme would be extremely close to the 3' hydroxyl end of the

cleaved DNA. The structure by Berger et al. (21) shows an openconfiguration of the en/yme, with the two active site tyrosines (andthe nearby CAP-like helix-turn-helix motifs) far enough apart to allow

the passage of a segment of duplex DNA through the gap. The recentstructure of a large fragment of E. coli gyrase subunit A. however,shows a closed configuration that is thought to be similar to that justprior to DNA cleavage (22). Once again, modeling of DNA into thisactive site region places the critical residues for drug resistance inclose contact with the DNA. Thus, although the two crystal structuresappear to reflect different enzyme conformations, both are entirelyconsistent with the quinolone-resistance determining region binding

to DNA very near the site of DNA cleavage.Drug cross-linking studies with a photoactivated m-AMSA deriv

ative have placed the drug-binding site precisely at the cleaved phos-phodiester bonds (9). Thus, assuming that the models of the protein-

DNA complexes are accurate and (hat the T4 enzyme has a similarstructure. T4 residues N78 and S79 are located very near or at (hedrug-binding site. It seems highly likely that these residues interact

directly with inhibitors at the active site. Based on this inference, wecan also make a strong argument that quinolones and intercalatingantitumor agents such as m-AMSA bind to common or closely over

lapping sites: mutations of N78 and S79 alter sensitivity to bothcompounds.

Our results with N78 substitutions reveal a striking parallel betweenthe T4 topoisomerase and bacterial DNA gyrase. In gyrase. thehighest level ot quinolone resistance is obtained when the small polarS83 is substituted with bulky nonpolar residues (L. F, or W; Refs. 13,29, 30, and 47). The wild-type T4 topoisomerase is much less sensi

tive to oxolinic acid than is DNA gyrase. arguably because the N78residue of the T4 enzyme is bulkier than the serine of gyrase.'' We

found that substitution of N78 with serine causes hypersensitivity tooxolinic acid (i.e., making the T4 topoisomerase more like gyrase).Furthermore, substitution of W at residue 78 causes dramatic resistance to oxolinic acid. These results strongly suggest that oxolinic acidinteracts with this residue in the T4 topoisomerase and E. coli gyrasedrug-binding pockets in a fundamentally similar manner. Interest

ingly, the N78C substitution in gp52 also causes resistance to oxolinicacid, but as far as we know, an S83C mutant has never been tested inbacterial gyrase."1

The S79F substitution by itself caused a topoisomerase-negative

phenotype in vivo, implying that the mutant protein is either veryunstable or catalytically inactive. The S79L substitution also causedpoorer than normal growth at low temperature (data not shown),suggesting reduced activity, and the failure to isolate S79P and S79Wsuggested that these substitutions have negative consequences as well(see "Materials and Methods"). These results argue that residue 79 is

critical for enzyme activity. In general, when considering whichamino acid substitutions lead to drug resistance in various topoisomerases, it is thus important to realize that some substitutions may not beconsistent with enzyme function.

One of the most interesting results we obtained was the suppressionof the topoisomerase-negative phenotype of S79F by the G269V

mutation. In the recent crystal structures, the residues correspondingto G269 in yeast topoisomerase (D939) and bacterial gyrase (M301 )

'' Mutation of S83 to N in E. coli GyrA would require a Ihree-nucleolide substitution:

hence, this substitution would not be expected in a spontaneous quinolone-resistant

mutarli, even if it causes quinolone resistance.10Mutation of SX3 lo C in E. culi GyrA would require a two-nucleotide substitution:

hence, this substitution would not be expected in a spontaneous quinolone-resistant

mutant, even if il causes quinolonc resistance.

1265

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOPOISOMI-RASK MUTATIONS

are within the "tower domain," which sits atop the CAP-like domain

discussed above (21. 22). However, these residues are too far from theresidues corresponding to S79 to argue for a direct interaction. Therefore, the crystal structures do not provide a simple explanation of themechanism of suppression. Perhaps the two residues are in closeproximity in some conformation of the enzyme that has not beencaptured in either of the crystal structures.

Surprisingly, the G269V substitution caused a dramatic hypersen-sitivity to both oxolinic acid and w-AMSA. This mutant appears to

define a category of topoisomerase mutants that has not been isolatedpreviously, i.e., those that are generally hypersensitive to all inhibitorsthat stabilize the cleavage complex. As with the question of suppression, it is not obvious how an alteration of G269 affects drug sensitivity so dramatically when the residue appears to be distant from thecleavage site. To our knowledge, no mutation that alters drug sensitivity has ever been reported in the region corresponding to the towerdomain of a type II topoisomerase, but hypersensitive mutants havenever been systematically sought. It should be very interesting toexplore the nature of the drug hypersensitivity in purified enzymewith the G269V substitution.

In summary, this analysis of altered drug sensitivity provides additional insight into the mechanism of action of topoisomerase inhibitors and into drug resistance. It is very likely that substitutions of N78and S79 directly affect the binding of both the quinolone oxolinic acidand the antitumor agent w-AMSA, providing another strong link

between the mechanism of action of these inhibitors. Our results,coupled with the recent crystal structures of topoisomerases, supportmodels in which the inhibitors bind to the internucleotide space at thecleaved phosphodiester bonds and interact with residues of topoisomerase that are in close contact with the DNA at that location. Wealso showed that mutations in two regions quite distant from thescissile phosphodiester bond in the available crystal structures markedly affect drug action, and one of these regions has not been implicated previously in drug sensitivity. Additional studies are needed toclarify the precise chemical nature of drug-enzyme interactions at the

active site and to explain the altered drug sensitivity in mutants withsubstitutions that are distant from the cleaved DNA.

ACKNOWLEDGMENTS

We thank Laura Wieslo for technical assistance; Jim George and LynnZechiedrich for bacterial strains; and Steve White. James Berger. and TonyMaxwell for helpful discussions of topoisomerase structure.

REFERENCES

1. Recce. R. J., and Maxwell. A. DNA gyrase: structure and function. CRC Crii. Rev.Biochem. Mol. Biol.. 26: 335-375. 1991.

2. Chen. A. Y.. and Liu. L. F. DNA topoisomerases: essential enzymes and lethaltargets. Annu. Rev. Pharmacol. Toxico!., 34: 191-218, 1994.

3. Wigley. D. B. Structure und mechanism of DNA topoisomerases. Annu. Rev. Bio-phys. Biomol. Struct.. 24: 185-208. 1995.

4. Kreuzer. K. N., and Co/./.arelli. N. R. Escherichia culi mutants thermosensitive fordeoxyrihonucleic acid gyrase suhunit A: effects on deoxyrihonucleic acid replication,transcription, and hacteriophage growth. J. Bacterio!., 140: 424-435. 1979.

5. (ìlisson.B. S-. and Ross, W. E. DNA topoisomerase II: a primer on the en/yme andits unique role as a mullidrug target in cancer chemotherapy. Pharmacol. Then. 32:89-106. 1987.

6. Drlica. K.. and Franco, R. J. Inhibitors of DNA lopoisomerases. Biochemistry. 27:2253-2259. 1988.

7. Capranico. G., and Zunino. F. DNA lopoisomerase-trapping amitumour drugs. Eur. J.Cancer, 28A: 2055-2060. 1992.

8. Freudenreich. C. H., and Kreuzer, K. N. Mulationul analysis of a type I! topoisomerase cleavage site: distinct requirements for enzyme and inhibitors. EMBO J.. 12:2085-2097, 1993.

9. Freudenreich. C. H., and Kreuzer. K. N. Localization of an aminoacridine antitumoragent in a type II topoisomerase-DNA complex. Proc. Nati. Acad. Sci. USA, 91:11007-11011. 1994.

10.

19.

20.

21.

22

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

Shen. L. L.. Kohlbrenner, W. E.. Weigl. D., and Baranowski. J. Mechanism ofquinolone inhibition of DNA gyrase. Appearance of unique norfloxacin binding sitesin enzyme-DNA complexes. J. Biol. Chem., 264: 2973-2978. 1989.

Willmott. C. J. R.. and Maxwell. A. A single point mutation in the DNA gyrase Aprotein greatly reduces binding of fluoroquinolones to the gyrase-DNA complex.Anlimicrob. Agents Chemother., 37: 126-127, 1993.Yoshida. H.. Bogaki. M.. Nakamura. M.. and Nakamura. S. Quinolone resistance-determining region in the DNA gyrase K.vM gene of Escherichia coli. Antimicrob.Agents Chemolher., 34: 1271-1272. 1990.Maxwell. A. The molecular basis of quinolone action. J. Antimicrob. Chemother.. 30:409-416. 1992.Rowe, T. C.. Tewey. K. M.. and Liu. L. F. Identification of the breakage-reunionsubunit of T4 DNA topoisomerase. J. Biol. Chem., 259: 9177-9181, 1984.

Huff, A. C.. and Kreuzer. K. N. Evidence for a common mechanism of action forantitumor and antibacterial agents that inhibit type II DNA topoisomerases. J. Biol.Chem., 265: 20496-20505, 1990.

Huff. A. C., Leatherwood, J. K.. and Kreuzer. K. N. Bacteriophage T4 DNAtopoisomerase is the target of amitumor agent 4'-(9-acridinylamino)methanesulfon-

m-anisidide (w-AMSA) in T4-infected Escherichia coli. Proc. Nati. Acad. Sci. USA,86: 1307-1311. 1989.

Huff. A. C.. Ward. R. E., IV, and Kreuzer. K. N. Mutalional alteration of thebreakage/resealing suhunit of bacteriophage T4 DNA topoisomerase confers resistance to anlitumor agent m-AMSA. Mol. Gen. Genet., 221: 27-32, 1990.Liu. L. F.. Liu, C. C.. and Alberts, B. M. T4 DNA topoisomerase: a new ATP-dependent enzyme essential for initiation of T4 bacteriophage DNA replication.Nature (Lond.), 281: 456-461, 1979.Kreuzer. K. N., and Alberts. B. M. Site-specific recognition of bacteriophage T4DNA by T4 type II DNA topoisomerase and Escherichia coli DNA gyrase. J. Biol.Chem.. 259: 5339-5346. 1984.

Kreuzer. K. N. A bacteriophage model system for studying lopoisomerase inhibitors.In: L. F. Liu (ed.). DNA Topoisomerases: Topoisomerase-Targeting Drugs, pp.171-186. San Diego: Academic Press. 1994.Berger, J. M.. Gamblin. S. J.. Harrison. S. C.. and Wang. J. C. Structure andmechanism of DNA topoisomerase II. Nature (Lond.), 379: 225-232, 1996.

MoráisCabrai. J. H.. Jackson. A. P.. Smith. C. V.. Shikotra, N.. Maxwell. A., andLiddington, R. C'. Crystal structure of the breakage-reunion domain of DNA gyrase.

Nature (Lond.). 388: 903-905. 1997.Kreuzer. K. N.. Engman. H. W.. and Yap. W. Y. Tertiary initiation of replication inbacteriophage T4. Deletion of the overlapping nvsY promoter/replication origin fromthe phage genome. J. Biol. Chem.. 2f>3: 11348-11357. 1988.Bachmann. B. J. Derivations and genotypes of some mutant derivatives of Esche-richia culi K-12. In: F. C. Neidhardt (ed.) Escherichia coli and Salmonella, pp.2460-2488. Washington. DC: ASM Press, 1996.

Khodursky, A. B.. 7-echiedrich, E. L., and Cozzarelli, N. R. Topoisomerase IV is atarget of quinolones in Escherichia coli. Proc. Nati. Acad. Sci. USA. 92: 11801-

11805. 1995.Nelson. M. A.. Ericson. M.. Gold. L.. and Pulitzer. J. F. The isolation and characterization of TabR bacteria: hosts that restrict bacteriophage T4 rll mutants. Mol. Gen.Genet.. 188: 60-68. 1982.Selick. H. E.. Kreuzer. K. N., and Alberts. B. M. The bacteriophage T4 insertion/substitution vector system. A method for introducing site-specific mutations into thevirus chromosome. J. Biol. Chem.. 26.?: 11336-11347. 1988.Kreuzer. K. N.. and Selick. H. I:. Directed insertion/substitution mutagenesis. In: 1. D.Karam (ed.). Molecular Biology of Bacteriophage T4. pp. 452-454. Washington. DC":

ASM Press. 1994.Munakata. N.. Morohoshi. F.. Saitou. M.. Yamazaki. N., and Hayashi. K. Molecularcharacterization of thirteen gyrA mutations conferring nulidixic acid resistance inBacillus subtilis. Mol. Gen. Genet.. 244: 97-103, 1994.Ilo. H.. Yoshida. H.. Bogaki-Shonai, M., Niga. T.. Hattori, H.. and Nakamura. S.Quinolone resistance mutations in the DNA gyrase gyrA and gyrB genes of Siapliy-locm-cus áureas. Antimicrob. Agents Chemother.. 38: 2014-2023, 1994.

Nikaido. H. Multidrug efflux pumps of gram-negative bacteria. J. Bacteriol., 178:5853-5859. 1996.

Ma. D.. Cook. D. N.. Alberti. M.. Pon, N. G.. Nikaido, H., and Hearst. J. E. GenesacrA and acrB encode a stress-induced efflux system of Escherichia coli. Mol.Microbio!.. 16: 45-55. 1996.Yoshida. H.. Kojima. T.. Yamagishi. J.. and Nakamura. S. Quinolone-resistantmutations of the gyrA gene of Escherichia coli. Mo!. Gen. Genet.. 211: 1-7. 1988.Mufti. S., and Bernstein. H. The DNA-delay mutants of bacteriophage T4. J. Virol..14: 860-871. 1974.Neece. S. H., Carles-Kinch. K., Tomso. D. J.. and Kreuzer. K. N. Role of recombi-national repair in sensitivity to an antitumor agent that inhibits bacteriophage T4 typeII DNA topoisomerase. Mol. Microbiol., 20: 1145-1154. 1996.Carón. P. R., and Wang. J. C. Appendix II: alignment of primary sequences of DNAtopoisomerases. In: L. F. Liu (ed.). DNA Topoisomerases: Topoisomerase-TargetingDrugs, pp. 271-297. San Diego. CA: Academic Press. 1994.

Yoshida. H.. Bogaki. M.. Nakamura. M.. Yamanaka. L. M.. and Nakamura. S.Quinolone resistance-determining region in the DNA gyrase a\rB gene of Escherichia co/i. Antimicrob. Agents Chemother.. 35: 1647-1650. 1991.

Carón, P. R.. and Wang. J. C. DNA topoisomerases as targets of therapeutics: astructural overview. In: T. Andoh. H. Ikeda. and M. Oguro (eds.l. Molecular Biologyof DNA Topoisomerases and Its Application to Chemotherapy, pp. 1-18. Boca Raton.FL: CRC Press. 1993.Gensberg. K.. Jin. Y. F.. and Piddock. L. J. V. A novel gyrB mutation in afluoroquinolone-resislant clinical isolate of SalnnnieUti l\phinuiriinn. Fed. Eur. Microbiol. Soc. Microbiol. Lett.. 132: 57-60. 1995.

1266

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

T4 TOPOISOMBRASE MUTATIONS

40. Wasserman, R. A-, and Wang. J. C. Mechanistic studies of amsacrine-resistantderivatives of DNA topoisomerase II. Implications in resistance to multiple antitumordrugs targeting the enzyme. J. Biol. Chem., 269; 20943-20951. 1994.

41. Beck. W. T., Danks. M. K.. Woverton, J. S., Chen. M.. Granzen. B., Kim, R., andSuttle, D. P. Resistance of mammalian tumor cells to inhibitors of DNA topoisomerase II. In: L. F. Liu (ed.). DNA Topoisomerases: Topoisomerase-Targeting Drugs, pp.145-169. San Diego, CA: Academic Press, 1994.

42. Nitiss. J. L. Yeast as a genetic model system for studying topoisomerase inhibitors.Adv. Pharmacol., 29B: 201-226, 1994.

43. Yamagishi, J.. Yoshida. H., Yamayoshi. M.. and Nakamura. S. Nalidixic acid-resistantmutations of the gyrB gene of Escherichia coli. Mol. Gen. Genet., 204: 367-373. 1986.

44. Liu. Y-X., Hsiung, Y.. Jannatipour. M.. Yeh. Y., and Niliss. J. L. Yeast topoisomeraseII mutants resistant to anti-topoisomerase agents: identification and characterization

of new yeast topoisomerase II mutants selected for resistance to etoposide. CancerRes., 54: 2943-2951. 1994.

45. Hsiung. Y. C., Elsea, S. H., Osheroff. N.. and Nitiss, 1. L. A mutation in yeast TOP2homologous to a quinolone-resistant mutation in bacteria. J. Biol. Chem., 270:20359-20364, 1995.

46. Withoff, S., De Jong, S., De Vries, E. G. E.. and Mulder. N. H. Human DNAtopoisomerase II: biochemistry and role in chemotherapy resistance. Anticancer Res.,16: 1867-1880, 19%.

47. Sreedharan, S., Oram. M.. Jensen. B., Peterson, L. R.. and Fisher. L. M. DNA gyraseg\rA mutations in ciprofloxacin-resistant strains of Staphvlococcux aureua: closesimilarity with quinolone resistance mutations in E.scherichia coli. J. Bacterio!.. 772.'

7260-7262. 1990.48. Huang. W. M. Nucleotide sequence of a type II DNA topoisomerase gene. Bacteri-

ophage T4 gene 39. Nucleic Acids Res., 14: 7751-7765, 1986.49. Huang. W. M. The 52-protein subunit of T4 DNA topoisomerase is homologous to

the gyrA-prolein of gyrase. Nucleic Acids Res.. 14: 7379-7390, 1986.

1267

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

1998;58:1260-1267. Cancer Res   Catherine H. Freudenreich, Christine Chang and Kenneth N. Kreuzer  Antibacterial Quinolone

-anisidide and anm-(9-Acridinylamino)methanesulfon-′That Alter Sensitivity to Antitumor Agent 4

Mutations of the Bacteriophage T4 Type II DNA Topoisomerase

  Updated version

  http://cancerres.aacrjournals.org/content/58/6/1260

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/6/1260To request permission to re-use all or part of this article, use this link

Research. on January 18, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from