Substitutions of Asn-726 in the active site of yeast DNA topoisomerase I define novel mechanisms of stabilizing the covalent enzyme-DNA intermediate.

Eukaryotic DNA topoisomerase I (Top1p) catalyzes changes in DNA topology and is the cellular target of camptothecin. Recent reports of enzyme structure highlight the importance of conserved amino acids N-terminal to the active site tyrosine and the involvement of Asn-726 in mediating Top1p sensitivity to camptothecin. To investigate the contribution of this residue to enzyme catalysis, we evaluated the effect of substituting His, Asp, or Ser for Asn-726 on yeast Top1p. Top1N726S and Top1N726D mutant proteins were resistant to camptothecin, although the Ser mutant was distinguished by a lack of detectable changes in activity. Thus, a basic residue immediately N-terminal to the active site tyrosine is required for camptothecin cytotoxicity. However, replacing Asn-726 with Asp or His interfered with distinct aspects of the catalytic cycle, resulting in cell lethality. In contrast to camptothecin, which inhibits enzyme-catalyzed religation of DNA, the His substituent enhanced the rate of DNA scission, whereas the Asp mutation diminished the enzyme binding of DNA. Yet, these effects on enzyme catalysis were not mutually exclusive as the His mutant was hypersensitive to camptothecin. These results suggest distinct mechanisms of poisoning DNA topoisomerase I may be explored in the development of antitumor agents capable of targeting different aspects of the Top1p catalytic cycle.

Eukaryotic DNA topoisomerase I is a highly conserved enzyme that constitutes the cellular target of camptothecin, a potent antineoplastic agent with broad spectrum antitumor activity (1)(2)(3)(4)(5). The catalytic cycle of this enzyme consists of successive transesterifications involving the concerted cleavage and religation of DNA (1, 6 -8). DNA topoisomerase I binds double-stranded DNA and transiently cleaves one strand of the duplex via the nucleophilic attack of the active site tyrosine on the phosphodiester backbone bond of the DNA. A phosphotyrosyl linkage in the covalent enzyme-DNA intermediate tethers the enzyme to the 3Ј-phosphate end of the cleaved DNA strand.
Biochemical and crystallographic data suggest that the free DNA end rotates around the intact complementary strand, affecting alterations in the linkage of the two DNA strands (1, 6 -9).
Camptothecin specifically interferes with the catalytic cycle of DNA topoisomerase I by reversibly stabilizing the covalent intermediate or cleavable complex (3,4,10,11). During Sphase, this ternary drug-enzyme-DNA complex produces an irreversible inhibition of DNA synthesis, double-stranded DNA breaks, cell cycle arrest in G2 phase, and cell death via a mechanism of apoptosis. The cytotoxic action of the drug has been attributed to replication fork stalling or breakage when the advancing polymerase complex collides with the drug-stabilized enzyme-DNA intermediate (12)(13)(14)(15)(16). This model is supported by the fact that inhibitors of DNA synthesis abrogate camptothecin-induced lethality. Moreover, rad52⌬ yeast cells, defective in the recombinational repair of double-stranded DNA breaks and collapsed replication forks, are hypersensitive to the drug (17)(18)(19).
Significant advances have been made in defining the structural features of camptothecin important for the productive interaction of this agent with DNA topoisomerase I and DNA (3, 20 -22). Recent crystallographic data on the structures of reconstituted and truncated versions of human DNA topoisomerase I in complex with DNA provide further insights into mechanistic aspects of enzyme function (6,7,23). Most of the mutations known to affect enzyme sensitivity to the drug are clustered along one face of the DNA in the Top1p-DNA cocrystal structure, close to the active site tyrosine (2,3,6). However, despite extensive investigation of camptothecin-resistant yeast and mammalian enzymes, the specific interactions required for the effective binding of the drug to the covalent intermediate remain largely unknown.
The residues preceding the active site tyrosine are highly conserved among the C-terminal domains of eukaryotic cellular enzymes ( Fig. 1) (24,25). A conservation of function is also supported by several studies involving specific amino acid substitutions. For example, mutation of these residues in either yeast or human Top1 to match the Ser-Lys-Arg-Ala-Tyr sequence found at the corresponding position in the camptothecin-resistant vaccinia virus enzyme rendered the mutant Top1vac enzymes resistant to the drug (26,27). Conversely, substitution of Ala for Thr-722 enhanced the stability of the covalent enzyme-DNA complex in the absence of camptothecin, which proved lethal when top1T722A was expressed in yeast or mammalian cells (28 -30). Similar effects on human Top1p function have recently been reported for the analogous replace-ment of Thr-718 with Ala (31). The characterization of second site mutations in yeast top1T722A or human top1T718A, which suppress the lethal phenotype, provides evidence of functionally conserved domainal interactions (29,31).
The structures of human Top1p in covalent and noncovalent complexes with DNA demonstrate remarkably few contacts between the DNA substrate and the C-terminal domain (6,23). Among residues near the active site tyrosine, only those immediately N-terminal to Tyr (Thr, Lys, and Asn in Fig. 1) interact with phosphate groups near the site of DNA scission. Mutation of this Asn to Leu in yeast Top1N726Lp diminished the catalytic activity and camptothecin sensitivity of the enzyme (26). Substitution of Ser for the same Asn residue in human DNA topoisomerase I also conferred camptothecin resistance, without a concomitant decrease in activity (32).
To directly address the contribution of this residue to DNA topoisomerase I catalysis, we investigated the consequences of substituting His, Ser, and Asp for Asn-726 in yeast Top1p. Here we report that these mutations had dramatically different effects on enzyme binding to DNA, cleavage of DNA, and camptothecin sensitivity. Our studies indicate a basic residue immediately N-terminal to the active site tyrosine is essential for the camptothecin sensitivity of the enzyme. Further, the presence of histidine enhanced cleavage of the DNA substrate. In contrast, replacement of Asn with an acidic residue diminished the ability of the enzyme to bind DNA. As a consequence, noncovalent binding of DNA 3Ј to the cleavage site was selectively destabilized in the covalent complex. These data suggest novel mechanisms of stabilizing the covalent enzyme-DNA intermediate, which may be exploited in the development of new antitumor drugs that target DNA topoisomerase I.

EXPERIMENTAL PROCEDURES
Chemicals, Yeast Strains, and Plasmids-Camptothecin was obtained from Sigma and dissolved at 4 mg/ml in Me 2 SO. Aliquots were stored at Ϫ20°C.
Expression of yeast TOP1 from the pGAL1 promoter in plasmids YCpGAL1-TOP1-L and YCpGAL1-TOP1 has been described (18, 26 -28, 34). In top1N726L, amino acid residue Asn-726 (adjacent to active site tyrosine 727) was mutated to Leu (25, 26) (see Fig. 1). Substitution of Ser for Asn-726 (top1N726S) was achieved by oligonucleotide-directed mutagenesis using a kit from Amersham Pharmacia Biotech. A 445base pair SalI-PstI DNA fragment encoding the active tyrosine region of yeast TOP1 was cloned into M13mp19 RF DNA. The single-stranded phage DNA was mutagenized with the oligonucleotide 5Ј-GCACTTC-CAAAATCAGTTATATAGACCCTAG-3Ј. The mutated sequences were excised with SalI/PstI and used to replace the wild-type sequences in YCpGAL1-TOP1-L to yield plasmid YCpGAL1-top1N726S-L. L designates the selectable marker LEU2. Other ARS/CEN vectors used in this study contain the URA3 marker. Plasmid YCpGAL1-top1N726S was constructed by replacing the Bsu36I-XbaI DNA fragment of YCpGAL1-TOP1 with the corresponding sequences from YCpGAL1-top1N726S-L.
Mutation of Asn-726 to Asp and His, in mutants top1N726D and top1N726H, respectively, was accomplished by degenerate oligonucleotide mutagenesis of TOP1 sequences spanning the active site tyrosine (described in Ref. 28). Briefly, the mutated sequences in the replicative form of M13mp18 were restricted with SpeI and PstI and successively cloned into the expression vectors ptacTOP1 (35) and YCpGAL1-TOP1. The presence and orientation of all mutations were confirmed by DNA sequencing.
Yeast strains were transformed by treatment with lithium acetate (36). Plasmids were maintained by growing cells in synthetic complete medium lacking uracil (S.C. 1 -ura) or leucine (S.C.-leu) and supplemented with 2% dextrose or galactose, as indicated.
Yeast Cell Sensitivity to Camptothecin-The camptothecin sensitivity of yeast cells transformed with various TOP1 constructs was determined as described (2,15,27). Individual transformants were grown in selective medium plus dextrose and serially 10-fold diluted, and 5-l volumes were spotted onto selective plates supplemented with 25 mM HEPES, pH 7.2, 2% dextrose or galactose, and 10 g/ml camptothecin in a final 0.25% Me 2 SO. No drug control plates contained 0.25% Me 2 SO. Cell viability was scored following incubation at 30°C.
DNA Topoisomerase I Activity Assays and DNA Cleavage Reactions-DNA topoisomerase I was purified from EKY3 cells transformed with various pGAL1-TOP1 constructs as described (26,27,33). Briefly, extracts from 1.5 liters of galactose-induced cultures were fractionated by successive 30 and 75% saturation ammonium sulfate precipitations and phosphocellulose chromatography. Proteins bound to the phosphocellulose column were eluted with successive 2-column volumes of TEEG buffer (50 mM Tris, pH 7.4, 1 mM EDTA, 1 mM EGTA, 10% glycerol) containing 0.2, 0.4, 0.6, 0.8, and 1.0 M KCl. Top1 proteins, in the 0.6 M KCl fractions, were adjusted to a final 50% glycerol and stored at Ϫ80°C. Protein concentrations were determined with the Bio-Rad protein assay. Top1 protein integrity was assessed in immunoblots probed with rabbit antibodies specific for yeast DNA topoisomerase I and stained with alkaline-phosphatase-coupled secondary antibodies (26,27).
DNA topoisomerase I activity was assayed by the relaxation of supercoiled plasmid DNA, as described (26 -28, 33). Briefly, serial 10-fold dilution of proteins corrected for concentration were incubated in 20-l reaction volumes with 0.3 g of negatively supercoiled pHC624 DNA in 20 mM Tris-HCl (pH 7.5), 0.1 mM EDTA, 10 mM MgCl 2 , 50 g/ml gelatin, and 150 mM KCl. Where indicated, the final KCl concentration was adjusted from 50 to 200 mM, in 25 mM increments. Following incubation at 30°C, the reactions were terminated by the addition of 1% SDS, and the extent of plasmid DNA relaxation was assessed following electrophoresis in agarose gels. Topoisomer bands were stained with ethidium bromide and photographed.
Enzyme sensitivity to camptothecin was assayed in DNA cleavage reactions (26 -28). A single 3Ј-end-labeled DNA substrate was prepared by cleaving plasmid pHCAK3-1 with BglII, repairing the ends with [ 32 P]dATP, dNTPs, and Sequenase, and restricting with BamHI. Following purification in a 5% polyacrylamide gel, ϳ7.5 ng (8000 cpm) of the 480-base pair DNA fragment was incubated with equal concentrations of Top1p in 50-l reaction volumes containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 0.1 mM EDTA, 50 mM KCl, 50 g/ml gelatin, and a final 50 M camptothecin. After 30 min at 30°C, the covalent complexes were trapped with 1% SDS at 75°C for 10 min. The samples were digested with 0.2 mg/ml proteinase K, ethanol precipitated, and electrophoresed in an 8% polyacrylamide, 7 M urea gel in 0.1 M Tris-Borate buffer at 20 V/cm. The cleavage products were visualized by autoradiography.
Cleavage reactions were performed at room temperature by incubating equal concentrations of enzyme with 20 fmol of DNA in 20 mM Tris-HCl (pH 7.5), 5 mM MgCl 2 , 0.1 mM EDTA, 50 mM KCl, and 50 g/ml gelatin. Reactions were terminated with SDS (0.5% final) followed by the addition of 3.3 volumes of loading buffer (1 mg/ml bromphenol blue, 1 mg/ml xylene cyanol in either 95% formamide or 80% formamide, 10 mM NaOH). Samples were loaded on 16% polyacrylamide, 7 M urea gels and electrophoresed at 40 V/cm at 50°C.
DNA Topoisomerase I Activity in vivo-JCW28 cells were co-transformed with plasmid YEptopA-PGPD, which constitutively expresses bacterial DNA topoisomerase I (28,38), and the yeast YCpGAL1-top1-L constructs. Cells were grown at the permissive temperature (25°C) in S.C.-tryptophan, -leucine dextrose medium, diluted into selective medium containing raffinose, and induced with 2% galactose at A 595 ϭ 1.3-1.4. After 5 h, half of each culture was shifted to the nonpermissive 1 The abbreviations used are: S.C., synthetic complete; Cpt, camptothecin. temperature (35°C) for an additional 3 h. To prevent changes in plasmid DNA topology during cell lysis, the cultures were mixed with an equal volume of buffered ethanol/toluene (20 mM Tris, pH 8.0, 95% ethanol, 3% toluene, and 10 mM EDTA) prechilled to Ϫ20°C (28). Following cell lysis, plasmid topoisomers were resolved by two-dimensional agarose gel electrophoresis, where the 0.6 g/ml chloroquine of the first dimension was increased to 3 g/ml in the second dimension (26,28). Plasmid topoisomers were visualized with a random-primed 32 P-labeled probe containing the 2-m plasmid origin of replication followed by autoradiography.

Mutation of Asn-726 in Yeast DNA Topoisomerase I Affects
Cell Viability and Sensitivity to Camptothecin-The residues preceding the active site Tyr-727 in yeast DNA topoisomerase I ( Fig. 1) are highly conserved among eukaryotic cellular enzymes (24,25). Mutation of these residues in yeast and human Top1p has previously been reported to effect changes in enzyme activity or camptothecin sensitivity (15,(25)(26)(27)(28)(29)(30)(31)(32). For instance, in either enzyme, substitution of Ala for the conserved Thr, five residues N-terminal to the active site Tyr, enhanced the stability of the covalent intermediate, which proved lethal in yeast and mammalian cells (28 -31). However, mutations involving the Asn residue adjacent to the active site Tyr (yeast and human top1vac, yeast top1N726L, and human top1N722S) diminished the camptothecin sensitivity of the enzymes, with varying effects on DNA relaxation activity (25)(26)(27)32). These findings underscore the importance of the conserved sequence N-terminal to the active site tyrosine of eukaryotic DNA topoisomerase I and further support the direct involvement of Asn-726 in mediating the camptothecin sensitivity of Top1p. To more fully investigate the contribution of this residue to DNA topoisomerase I catalysis, these studies were expanded to evaluate the effect of substituting His, Asp, or Ser for Asn-726 in the yeast enzyme.
Replacing Asn with Ser rendered the mammalian enzyme resistant to camptothecin (32). The analogous Asn-726 to Ser mutation was engineered in yeast top1N726S by oligonucleotide-directed mutagenesis. Substitution of Asn-726 with Asp or His was accomplished with degenerate oligonucleotides that spanned the sequences flanking the active site tyrosine. The lethal top1T722A mutant was previously isolated with this mutagenesis strategy (28).
To assess the phenotypic consequences of these mutations, the top1 alleles were expressed from the galactose-inducible GAL1 promoter on a single copy ARS/CEN plasmid in a yeast strain deleted for TOP1 (top1⌬). As shown in Fig. 2A, yeast cells expressing either top1N726H or top1N726D exhibited at least a 3-log drop in cell viability in comparison to cells expressing wild-type TOP1. Whereas galactose-induced expression of top1N726S had no demonstrable effect on cell viability, the cells were resistant to camptothecin (Fig. 2B). This was in contrast to the dramatic decrease in viability induced by camptothecin in cells expressing comparable levels of wild-type Top1p.
The Top1N726S Protein Is Catalytically Active in Yeast-Because resistance to camptothecin may be a consequence of diminished catalytic activity (e.g. Top1N726Lp (26)), the ability of Top1N726Sp to catalyze the relaxation of plasmid DNA was evaluated in vivo. As reported by Giaever and Wang (38), transcription from divergent plasmid promoters generates local domains of negatively supercoiled DNA behind the transcription complexes and positively supercoiled DNA ahead. In a top1⌬, top2ts yeast strain, constitutively expressing the bacterial topA gene and shifted to the nonpermissive temperature, the only DNA topoisomerase activity detected is that of the bacterial enzyme. Unlike eukaryotic Top1p and Top2p, the bacterial enzyme will only catalyze the relaxation of negatively supercoiled DNA, resulting in the net accumulation of positively supercoiled plasmid DNAs (indicated by (ϩ) in Fig. 3). However, as previously reported (26,28), when wild-type TOP1 is reintroduced on a plasmid and expressed at the nonpermissive temperature, Top1 activity produces a quantitative shift in the topoisomer distribution from positive to negative (indicated as (Ϫ) in Fig. 3). The Top1N726S mutant protein was also active in vivo, producing a similar redistribution of plasmid DNA topoisomers (Fig. 3). Thus, unlike Leu (26), substitution of Ser for Asn-726 did not produce an appreciable decrease in catalytic activity.
Camptothecin sensitivity is greatly enhanced in yeast cells deficient in recombinational repair because of the deletion of the RAD52 gene (17)(18)(19)). Yet, galactose-induced expression of the camptothecin-resistant mutants, top1N726L and top1vac, in cells lacking functional Rad52p, is lethal in the absence of camptothecin (26). These results indicated that the alterations in catalytic activity necessary for drug resistance induce sufficient DNA damage to cause cell death in the absence of Rad52mediated repair. As both topN726L and top1vac involve Asn-726 substituents, we asked if the camptothecin-resistant phenotype of the Ser for Asn-726 mutant would also correlate with adverse affects on rad52⌬ cell viability.
As shown in Fig. 4, repair-proficient RAD52 cells expressing wild-type TOP1 were sensitive to camptothecin, whereas cells expressing top1N726L or top1N726S were resistant. Consistent with previous findings (26), galactose-induced top1N726L expression was lethal in a repair-deficient rad52⌬ strain in the absence of camptothecin. Expression of TOP1 and top1N726S slightly diminished cell growth relative to the top1⌬ controls. However, viable cells expressing top1N726S were evident even at high camptothecin concentrations. Thus, in contrast to other Asn-726 substitutions, replacement of this residue with Ser conferred a high degree of drug resistance, even in the absence of recombinational repair.
Substitution of Asn-726 Produces Differential Stability of the Covalent Enzyme-DNA Complex and Altered Enzyme Sensitivity Toward Camptothecin-To further investigate the effects of replacing Asn-726 with Ser, Asp, or His on enzyme mechanism and camptothecin sensitivity, the wild-type and mutant proteins were purified from galactose-induced top1⌬ cells expressing plasmid borne pGAL1-top1 constructs. The proteins were corrected for concentration, serially diluted, and assayed in plasmid DNA relaxation assays. At the salt concentrations optimal for wild-type Top1p (150 mM) the specific activity of Top1N726Sp and Top1N726Hp was comparable to that of wildtype Top1p (Fig. 5.). Thus, consistent with in vivo data (Fig. 3), the camptothecin resistance of Top1N726Sp-expressing cells could not be attributed to a loss in enzyme function. However, under the same conditions, Top1N726Dp-catalyzed relaxation of plasmid DNA was reduced by ϳ20-fold. Repeated assays indicated a consistent 20 -50-fold reduction in specific activity. The catalytically inactive Top1Y727F mutant (25,29,30) served as a negative control.
The effects of the Asn-726 substituents on covalent enzyme-DNA complex stability were assessed in DNA cleavage assays in the presence and absence of camptothecin. As described (27,28,30), enzyme preparations were incubated with a single 3Ј-end-labeled DNA fragment containing a high affinity binding site (marked with an asterisk). Covalent complexes were irreversibly trapped with SDS, and the extent of DNA cleavage was determined following electrophoresis in DNA sequencing gels.
As shown in Fig. 6, replacement of Asn-726 with Ser in Top1N726Sp abrogated the ability of camptothecin to stabilize the covalent complex. The low level of DNA cleavage by Top1N726Sp was unaffected by drug addition. In contrast, camptothecin induced a dramatic increase in cleavable complexes formed with equivalent amounts of wild-type Top1p, at the high affinity site as well as other sites.
Of the lethal Asn-726 substituents, only His enhanced the stability of the covalent complex in the absence of camptothecin (Fig. 6). Whereas cleavage was most pronounced at the high affinity site, stable complexes were also formed at other sequences. Relative to wild-type Top1p, the His mutant was hypersensitive to camptothecin as evidenced by the increase in the intensity and distribution of DNA cleavages in the presence of the drug. Although similar results were obtained when Thr-722 was mutated to Ala in Top1T722Ap (28), the pattern of DNA cleavage in the presence of camptothecin reverted to that of wild-type Top1p. In contrast, Top1N726Hp, produced a composite of DNA scissions; some were unique to the mutant enzyme, and others were coincident with drug-treated wild-type Top1p. Interestingly, insertion of Asp at residue 726 also conferred camptothecin resistance as cleavage was only slightly  JCW28 (top1⌬, top2ts), co-transformed with plasmid YEptopA-PGPD and either YCpGAL1-TOP1-L, YCpGAL1-top1N726S-L, or YCpGAL1-L (negative control), was grown in S.C.-tryptophan, -leucine dextrose medium, diluted into medium containing raffinose, and induced with galactose for 5 h at 25°C. Cells were then shifted to the nonpermissive temperature (35°C) for an additional 3 h before harvesting and lysis as described under "Experimental Procedures." The topological state of the endogenous 2-m plasmid was evaluated by two-dimensional gel electrophoresis, blotting onto a nylon membrane, and hybridization with a enhanced at the high affinity site. Reactions containing Top1Y727Fp, which is able to bind but not cleave DNA, appeared identical to the DNA control.
Taken together, these data suggest that efficient stabilization of the covalent complex by camptothecin requires a basic residue, such as Asn or His, immediately N-terminal to the active site Tyr. Replacement of Asn-726 with Ser or Asp abolishes camptothecin sensitivity. Yet, these same substitutions have profoundly different effects on enzyme activity: Ser had no demonstrable effect; His enhanced covalent complex stability; and Asp reduced activity by ϳ20-fold. The increment in covalent complexes formed with the His mutant accounts for the top1N726H lethal phenotype. Less clear is the basis for top1N726D-induced lethality.
Mutation of Asn-726 to His Enhances the Rate of DNA Cleavage, whereas Substitution with Asp Diminishes DNA Binding-To investigate the mechanism of enhanced stabilization of the Top1N726Hp-DNA covalent intermediate, rates of DNA cleavage were evaluated using a suicide DNA substrate. As diagrammed in Fig. 7, the partially double-stranded DNA consisted of a 19-mer scissile strand labeled at the 3Ј-end (*A) and annealed to a 37-mer complementary nonscissile strand. This substrate contains the high affinity site described above, except G was substituted for A at the ϩ1 position (downstream from the cleavage site) on the scissile strand. The suicide DNA substrate allows for an uncoupling of the DNA cleavage and religation reactions, as the 6-mer (5Ј-2GGAAA*A-3Ј) liberated upon cleavage readily dissociates from the complex and is unavailable for religation. As shown in Fig. 7, substitution of Ser or Asp for Asn-726 had little effect on the rate of enzymemediated DNA cleavage. Similar results were obtained with the lethal Top1T722Ap mutant. 2 In contrast, the rate of DNA cleavage was stimulated by the His mutation. Thus, the increased steady-state levels of covalent complexes obtained with Top1N726Hp in Fig. 6 could be attributed to enhanced DNA cleavage. This represents a novel mechanism of Top1 poisoning that contrasts with the inhibition of DNA religation ascribed to camptothecin (39).
Noncovalent binding of DNA topoisomerase I to DNA involves bipartite interaction of the enzyme with the DNA phosphodiester backbone upstream and downstream of the cleavage site. To assess the contribution of specific enzyme-DNA contacts that enhance or suppress the cleavage reaction, nicks were introduced at different sites along the nonscissile strand. As diagrammed in Fig. 8, the resulting double-stranded DNA substrates consist of a 3Ј-end-labeled 37-mer scissile strand (containing the same cleavage site in the suicide substrate) paired to complementary nonscissile strands containing nicks between positions Ϫ4 and ϩ3 opposite the site of scission. Cleavage of these substrates was evaluated in the absence of camptothecin (Fig. 8, lane 1), following salt addition at the end of the incubation period to reverse covalent complex formation (lane 2), as well as in the presence of camptothecin (lane 3) with salt reversal (lanes 4). A truncated version of human DNA topoisomerase I (hTop1) was also included.
As a basis for comparison, enzyme-mediated cleavage was first evaluated on the fully double-stranded DNA substrate without nicks. In Fig. 8, the behavior of wild-type Top1p, Top1N726Sp, Top1N726Hp, and Top1Y727Fp mirrored that observed with longer DNA substrates (compare Figs. 6 and 8). Further, all covalent complexes, whether drug-or mutationinduced, were salt reversible. However, the levels of DNA cleavage by Top1N726Dp in the absence of camptothecin exceeded that obtained at the high affinity site in the longer DNA substrate. This may reflect the difference in base composition at ϩ1 (CTT2GGG versus CTT2AGG). Nevertheless, covalent complex formation remained unaffected by the addition of camptothecin.
As was previously reported for human Top1 (37), the introduction of a nick between Ϫ4 and Ϫ3 suppressed DNA cleavage by yeast Top1p, Top1N726Sp, Top1N726Hp, and Top1N726Dp. This is consistent with a requirement for a phosphate at this position for efficient binding of the yeast enzyme to DNA. Further, with the exception of the catalytically inactive Top1Y727Fp, a nick at the Ϫ2/Ϫ1 position produced a 2-base pair shift upstream in the site of cleavage resulting in the liberation of a 25-mer product (indicated by the open arrow).
Shifting the nick between Ϫ1 and ϩ1 directly opposite the site of cleavage strongly enhanced cleavage even in the absence of camptothecin. Cleavage opposite the nick would produce a blunt-ended double strand break. Unless the untethered duplex DNA was tightly held by the protein, the noncovalently bound DNA ends could dissociate from the complex, effectively acting as a suicide substrate. Indeed, the lack of salt reversal attests to the formation of such abortive double-stranded breaks. Cleavage was most pronounced with the His mutant, followed by wild-type Top1p, and then the Ser and Asp mutants. Camptothecin did augment complex formation with wild-type Top1p, but not Top1N726Sp or Top1N726Dp, suggesting some reversal of the cleavage reaction that was drug inhibitable.
The effect of moving the nick to position ϩ2/ϩ3 differed significantly for Top1N726Dp compared the other enzymes tested, as the covalent Top1N726Dp-DNA complexes alone were not reversed upon salt addition. Moreover, relative to other DNA substrates examined, covalent complex stability was enhanced. These findings suggest that the Asp mutant enzyme may be defective in binding DNA or may alter DNA structure when bound. However, once the covalent intermediate was formed, any reduction in DNA binding would be localized to noncovalent interactions downstream from the cleavage site. The selective loss of the noncovalently bound DNA from the reaction intermediate would preclude efficient religation of the nicked DNA.
Because noncovalent binding of DNA topoisomerase I to DNA is dependent on the ionic strength of the reaction, enzyme catalytic activity was re-examined over a range of salt concentrations. As previously reported (26,29), wild-type Top1p  1 and 3) or NaCl (final concentration 0.5 M) for 30 min. followed by SDS (lanes 2 and 4). Lanes 3 and 4 also contained 10 M camptothecin. C contains DNA substrate alone. Arrows and numbers correspond to the position and size of the cleavage products, respectively. achieved maximal activity at ϳ150 mM KCl and retained activity at concentrations as high as 175 mM (Fig. 9). In stark contrast, Top1N726Dp catalytic activity was optimal in 75-100 mM KCl and rapidly decreased at salt concentrations in excess of 150 mM. Moreover, even at lower salt concentrations (Ͻ100 mM), Top1N726Dp exhibited a distributive mode of plasmid DNA relaxation, as evidenced by a continuum of bands extending from the negatively supercoiled DNAs to the top of the topoisomer ladder. Under the same conditions, wild-type Top1 and Top1N726Hp exhibit a distinctively processive mode of plasmid DNA relaxation. These data support the idea that replacing Asn-726 with Asp diminishes the affinity of the enzyme for DNA, with potentially dire consequences on the stability of the covalent intermediate (see "Discussion" for alternative models). Top1N726Hp activity, on the other hand, was optimal at higher salt concentrations than wild-type Top1p (175 mM) and was less active at lower salt concentrations. This may reflect a higher affinity of the His mutant enzyme for DNA. DISCUSSION DNA Topoisomerase I Sensitivity to Camptothecin-A shift in equilibrium between the concerted DNA cleavage and religation reactions catalyzed by DNA topoisomerase I can have dire consequences on cell viability. Indeed, numerous studies attribute the cytotoxic action of camptothecin to its ability to reversibly inhibit the religation reaction (Fig. 10, step III (reviewed in Refs. [2][3][4]). The resultant stabilization of the covalent enzyme-DNA complexes increases the probability of catastrophic collisions with advancing replication forks (step V) producing the DNA lesions that trigger cell cycle arrest and cell death. Previous studies have shown that substitution of Ala for the conserved Thr, five residues N-terminal to the active site tyrosine of yeast or human Top1p, also enhances the stability of the covalent complex and causes drug-independent cytotoxicity in yeast and mammalian cells (28 -31). Although the step(s) in the catalytic cycle altered by the Thr to Ala replacement have not been described, the lethal phenotype does not appear to be a consequence of altered DNA binding (29).
Mutations involving the amino acid residues surrounding the active site tyrosine in eukaryotic DNA topoisomerase I have also been shown to affect enzyme sensitivity to camptothecin. Changing the conserved T-S-K-I(L)-N-Y* sequences in the yeast and human topvac mutants to the T-S-K-R-A-Y* found in vaccinia Top1p abrogated the camptothecin sensitivity of the cellular enzymes (26,27,30). Yet, these alterations in yeast Top1vac activity produced sufficient DNA damage to induce rad52⌬ cell lethality. Replacement of the Asn residue with Ser in the human enzyme (32) or with Leu in the yeast top1N726L mutant (26) also rendered the enzymes resistant to camptothecin, although the Leu mutation was accompanied by a reduction in catalytic activity and rad52⌬-dependent lethality (26). These data support a critical role for this Asn in mediating the camptothecin sensitivity of eukaryotic DNA topoisomerase I.
To more directly assess the function of Asn-726 in enzyme catalysis and drug sensitivity, we analyzed the effects of several substituents on yeast Top1 function. Our studies indicated that a basic residue, such as Asn or His, was essential for the Cpt sensitivity of the enzyme. The replacement of this residue with Ser (as previously shown for human Top1p (32)) or Asp abolished the ability of Cpt to stabilize the covalent complex. However, the Ser mutant was distinguished from all other Asn mutants studied by the lack of detectable alterations in catalytic activity. Unlike the previously reported yeast top1vac and top1N726L mutants, placement of Ser immediately N-terminal to the active site tyrosine was not accompanied by an increase in DNA damage and rad52⌬ cell death in the absence of camptothecin. What contribution the side chain -OH may play in enzyme catalysis has yet to be defined.
Two models were recently proposed to describe camptothecin bound to the covalent intermediate formed by human Top1 and DNA (6,40). Both models are consistent with the hypothesis that Cpt is stacked against the base 3Ј to the DNA cleavage site (the ϩ1 position) (41). However, the orientation of the drug is essentially flipped in these models, such that the proposed stacking interactions of Asn with camptothecin is a key feature of the Redinbo et al. (6,7) model, whereas a hydrogen bond between the oxygen of the Asn side chain and the 20-OH of Cpt is a prominent feature in the model by Fan et al. (40) model (3). Clearly, the data presented here regarding the alterations in drug sensitivity because of mutation of the Asn residue will allow for a refinement of these models.
Enzyme-catalyzed DNA Cleavage and Religation-The Asn-726 substituents had even more dramatic effects on the DNA cleavage and religation reactions catalyzed by Top1p. The His and Asp mutants were lethal even in the absence of Cpt. The results of DNA cleavage assays using end-labeled DNA fragments, nicked DNA molecules, or suicide substrates indicate that replacing Asn-726 with His (N726H) enhances the rate of DNA scission catalyzed by the enzyme (Fig. 10, step II). The net effect is the relative accumulation of covalent intermediates, FIG. 9. DNA topoisomerase I mutants exhibit different salt optima in plasmid DNA relaxation assays. Equal concentrations of Top1p, Top1N726Dp, and Top1N726Hp were incubated with supercoiled plasmid pHC624 DNA at 30°C for 30 min followed by termination with SDS. For each enzyme the concentration of KCl in the reaction mix varied from 50 -200 mM as indicated. Reaction products were electrophoresed and visualized as described in Fig. 5. Sc, negatively supercoiled DNA substrate; R, relaxed DNA topoisomers. albeit via a mechanism distinct from that ascribed to camptothecin. Nevertheless, this would also increase the probability of replication-induced DNA lesions as described for ternary camptothecin-enzyme-DNA complexes (12,14).
The increase in salt concentration necessary for optimal relaxation of supercoiled plasmid DNA (Fig. 9) is consistent with a higher affinity of the Top1N726H protein for DNA. However, the mechanistic basis for elevated rates of DNA scission remain unclear. This may reflect an increase in nucleophilicity of the active site tyrosine induced by the close proximity of the His residue. In this case, His may act as a base to accept the proton from the attacking hydroxyl of the tyrosine. Indeed, available crystallographic data have failed to identify an amino acid residue in close enough proximity to act as a general base in the cleavage reaction. With wild-type Top1p, it has been proposed that a water molecule may act in this capacity. In Top1N726Hp, the presence of His may usurp this function, facilitating DNA scission. Although, the clarification of this alteration in enzyme function awaits further structural studies, this view is consistent with the enhanced cleavage by Top1N726Hp of DNA containing 7,8-dihydroxy-8-oxoguanine, a common modification resulting from oxidative damage (42).
In the case of the Asp for Asn mutation, the reduction in specific activity was consistent with diminished binding of the enzyme for DNA. Results with DNA substrates bearing nicks in the nonscissile strand suggest that, once the covalent complex is formed, this diminished binding specifically affects the stability of the intermediate by altering the noncovalent binding of the enzyme to DNA downstream of the cleavage site. As suggested in Fig. 10 (step V), the tracking of large complexes along the DNA, such as replication forks, would selectively destabilize the Top1-DNA intermediates when approached from the 3Ј-side of the DNA scission. In contrast, any diminution in DNA binding on the 5Ј-side would be obviated by the phosphotyrosyl linkage. A similar polarity in camptothecin-induced DNA damage was observed in an in vitro SV40 replication system (13) and supported by recent genetic studies of yeast mutants defective in DNA replication (15). An alternate explanation is that Top1N726Dp binding to duplex DNA induces structural alterations that diminish the efficiency of DNA cleavage. However, to account for the enhanced production of irreversible covalent complexes formed with DNA substrates containing a nick at ϩ2/ϩ3 on the noncleaved strand, such alterations must also diminish protein binding to DNA in the covalent intermediate. These possibilities may be distinguished in DNA binding assays.
Taken together, these results support a model of novel mechanisms of DNA topoisomerase I poisoning. In contrast to camptothecin, which reversibly inhibits the religation of DNA (Fig.  10, step III), His enhances the rate of DNA scission (step II) and the Asp alters enzyme binding to DNA with dire consequences downstream of the cleavage site (step V). Although the net effect, in each case, is presumed to be the lethal accumulation of covalent complexes, the specific DNA lesions produced and the cellular responses to such damage have yet to be defined. Indeed, preliminary studies with rad9⌬ strains suggest widely different cellular responses. 2 Further genetic analyses are currently underway to address these questions. Nevertheless, the increased sensitivity of the Top1N726H mutant protein to camptothecin suggests that these distinct mechanisms of stabilizing the covalent complex are not mutually exclusive. Thus, the development of new classes of DNA topoisomerase I-targeted drugs that affect distinct steps in the catalytic cycle may prove efficacious in combination with camptothecin analogs.