Alterations in the Catalytic Activity of Yeast DNA Topoisomerase I Result in Cell Cycle Arrest and Cell Death*

Eukaryotic DNA topoisomerase I catalyzes the relax- ation of supercoiled DNA through a concerted mechanism of DNA strand breakage and religation. The cyto- toxic activity of camptothecin results from the reversible stabilization of a covalent enzyme-DNA inter- mediate. Mutations in two conserved regions of yeast DNA topoisomerase I induced a similar mechanism of cell killing, albeit through different effects on enzyme catalysis. In Top1T722Ap, substituting Ala for Thr 722 reduced enzyme specific activity by 3-fold, yet enhanced the stability of the covalent enzyme-DNA complex. In contrast, Top1R517Gp was 1,000-fold less active and camptothecin resistant. Nevertheless, salt-stable DNA-enzyme intermediates were detected. Mutation of the active-site tyrosine abrogated mutant enzyme activity and cytotoxicity, while sublethal levels of top1T722A ex- pression increased rDNA recombination. In checkpoint proficient cells, p GAL1 -induced top1 expression coin- cided with the accumulation of a terminal G 2 -arrested phenotype. Although the acquisition of this phenotype did not require Rad9p, Top1R517Gp- and Top1T722Ap-induced lethality was enhanced in rad9 (cid:68) strains. Thus, despite mechanistic differences between Top1R517Gp and Top1T722Ap, the DNA lesions resulting from the enhanced stability of the covalent enzyme-DNA interme- diates were sufficient to cause cell cycle arrest and cell death. DNA topoisomerases catalyze changes in the linkage of DNA strands through a concerted mechanism of DNA strand breakage and rejoining (reviewed in Refs. 1–5). In eukaryotes, DNA topoisomerase I transiently

DNA topoisomerases catalyze changes in the linkage of DNA strands through a concerted mechanism of DNA strand breakage and rejoining (reviewed in Refs. [1][2][3][4][5]. In eukaryotes, DNA topoisomerase I transiently cleaves a single strand of duplex DNA to yield a covalent enzyme-DNA intermediate in which the active-site tyrosine is attached to the 3Ј-phosphate of the broken DNA strand. The transient formation of this proteinlinked gate in the DNA allows for the relaxation of supercoiled DNA, while conserving the energy of the broken phosphodiester backbone bond. The gene encoding DNA topoisomerase I, TOP1, has been cloned from a variety of sources, including the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe, Drosophila melanogaster, mouse and human, and exhibits a considerable degree of conservation (reviewed in Refs. 1, 4, 6, and 7). Genetic and biochemical studies indicate a role for this enzyme in DNA recombination, transcription, and rep-lication. In addition, DNA topoisomerase I constitutes the cellular target of the antitumor drug, camptothecin (reviewed in Refs. 4 -6 and 8).
Camptothecin reversibly stabilizes the covalent enzyme-DNA intermediate by interfering with the rejoining of the cleaved DNA strand (9 -11). Numerous studies suggest that during S phase, collision of the replication forks with the drugstabilized ternary complexes produces double-stranded DNA breaks, an inhibition of DNA synthesis, cell cycle arrest in G 2 and cell death (12)(13)(14). This mechanism of camptothecin-induced cell killing appears to be conserved in S. cerevisiae, as drug treatment of yeast cells expressing yeast or human TOP1 produces similar effects on cell cycle progression and cell viability (6,15,16).
Since the TOP1 gene is nonessential in yeast (17,18), this genetically tractable system has been exploited to address the mechanism of camptothecin-mediated cell killing and the effects of mutation on DNA topoisomerase I function and drug sensitivity (6, 15, 19 -23). For example, yeast strains deleted for the TOP1 gene (top1⌬) are resistant to camptothecin, while the expression of yeast or human DNA topoisomerase I from plasmid-borne sequences restores top1⌬ cell sensitivity to the drug (16). Camptothecin treatment of TOP1 ϩ yeast cells induces the expression of the DNA damage-responsive genes DIN3 and RNR3 (23,24), while deletion of the RAD52 gene, which is required for recombinational repair of doublestranded DNA breaks, enhances cell sensitivity to camptothecin (16,20,23). The camptothecin-induced toxicity of cells expressing wild-type DNA topoisomerase I has recently been shown to be mediated by the pleiotropic drug resistance (PDR) pathway (15,24). Specific amino acid substitutions in DNA topoisomerase I have been defined that reduce enzyme sensitivity to camptothecin, with little effect on catalytic activity (19,21,22). Substitution of Lys for Arg 420 in Top1-103p appeared to mimic the action of camptothecin, resulting in DNA damage and RAD9 dependent cell cycle arrest (25).
In this study, oligonucleotide-directed mutagenesis of conserved residues in yeast Top1p, centering on residue Arg 517 or the active-site tyrosine, Tyr 727 , identified top1 mutants that are lethal when expressed from the galactose-inducible GAL1 promoter. Expression of mutants top1T722A and top1R517G (in which Thr 722 was mutated to Ala and Arg 517 was changed to Gly, respectively) coincided with a rapid drop in cell viability and the accumulation of a terminal, G 2 -arrested phenotype, independent of a functional RAD9 cell cycle checkpoint. Biochemical and genetic studies suggest that top1T722A-and top1R517G-induced cell killing was achieved by distinct mechanisms, as the single amino acid substitutions had different effects on catalytic activity and enzyme sensitivity to camptothecin. Yet, in both cases, cell lethality could be attributed to a stabilization of the covalent reaction intermediate. Apparently, eukaryotic cells are able to repair the low levels of DNA dam-age that occur as a consequence of wild-type DNA topoisomerase I function; however, the extent of DNA topoisomerase I-induced DNA damage may be exacerbated by mutation or by drug action, producing elevated rates of recombination, cell cycle arrest, and cell death.

EXPERIMENTAL PROCEDURES
Materials, Yeast Strains and Plasmids-Camptothecin (Sigma) was dissolved in Me 2 SO at a final 4 mg/ml and aliquots stored at Ϫ20°C.
The single copy vector, YCpGAL1-TOP1, contains yeast TOP1 under the control of the galactose-inducible pGAL1 promoter (21). Plasmid BM125 (herein called YCpGAL1) contains the pGAL1 promoter in a URA3, ARS/CEN vector (30). Mutant alleles of yeast TOP1 were generated by oligonucleotide-directed mutagenesis of TOP1 sequences cloned into the M13mp19 vector, using a kit supplied by Amersham. Substitution of Phe for the active-site tyrosine, Tyr 727 , in top1Y727F, has been described (31). The single mutant top1R517G (Arg 517 mutated to Gly) and the double mutant top1T722A,Y727F (Thr 722 changed to Ala, Tyr 727 changed to Phe) were generated using the oligonucleotides 5Ј-AAGTTTTCGGTACATATA-3Ј and 5Ј-CAGGTTTCACTGGGCGCT-TCCAAAATCAATTTTATAGACCCTAGAC-3Ј, respectively. A degenerate oligonucleotide spanning the region around the active-site tyrosine, 5Ј-GTTTCACTGGGCACTTCCAAAATCAATTATATAG-3Ј, was synthesized by spiking each nucleotide precursor with 1.7% of the other three nucleotides (32). This ratio maximized the yield of random single nucleotide substitutions following oligonucleotide-directed mutagenesis. The entire pool of mutated sequences in the M13 RF DNA was excised and used to replace the wild-type TOP1 sequences in plasmid ptacTOP1 (33). Of the individual top1 mutant plasmids identified, ptac-top1T722A, encoded a substitution of Ala for Thr 722 . To express the top1 mutants from the pGAL1 promoter, DNA fragments bearing the mutations were excised from ptacTOP1 and exchanged for the wild-type sequences in YCpGAL1-TOP1 to yield plasmids YCpGAL1-top1Y727F, YCpGAL1-top1T722A, YCpGAL1-top1T722A, Y727F, and YCpGAL1-top1R517G. The plasmid YCpGAL1-top1R517G, Y727F, was generated by replacing the corresponding wild-type sequences in YCpGAL-top1Y727F with a DNA fragment containing the Arg 517 to Gly mutation.
For constitutive expression, a 3302-base pair DNA fragment containing the TOP1 coding region and 5Ј-flanking promoter sequences (Sc-TOP1) was excised from plasmid pIG (17), then ligated into the SmaI/ NheI sites of YEp24 to yield YEpScTOP1. Plasmid YCpScTOP1 was constructed by ligating the ScTOP1 sequences from YEpScTOP1 into the ApaI/BamHI DNA sites of the ARS/CEN vector pRS416 (34). The wild-type sequences in YCpScTOP1 were replaced with the top1T722A allele in plasmid YCpSctop1T722A.
Cell Viability Assays-The indicated top1⌬ yeast strains were transformed with the YCpGAL1-TOP1 constructs and selected on synthetic complete medium lacking uracil and supplemented with 2% dextrose (SC-uracil, dextrose). To assess cell viability, exponentially growing cultures of individual transformants were serially 10-fold diluted and 5-l aliquots spotted onto SC-uracil plates supplemented with 2% dextrose or galactose. Alternatively, the cultures were diluted 1:100 into SC-uracil media containing 2% raffinose and, at an OD 595 of 0.3, induced with a final 2% galactose. At various time points, aliquots were serially diluted and plated onto SC-uracil, dextrose. Cell viability was assessed following incubation at 30°C.
DNA Topoisomerase I Activity in Vivo-Yeast strain JCW28 (top1⌬,top2ts) was co-transformed with plasmid YEptopA-PGPD, which constitutively expresses the bacterial topA gene (35), and one of the following plasmids; YCpGAL1, YCpGAL1-TOP1, YCpGAL1-top1R517G, YCpGAL1-top1T722A, or YCpGAL1-top1T722A,Y727F. Individual transformations were grown at the permissive temperature (25°C) in selective dextrose media, diluted into raffinose containing media and induced with 2% galactose at an OD 595 of 1.0. After 5 h, half the culture volume was shifted to the nonpermissive temperature (35°C) for an additional 3 h. Intracellular plasmid DNA topology was preserved by the addition of an equal volume of 20 mM Tris-HCl (pH 8.0), 95% ethanol, 3% toluene, 10 mM EDTA, prechilled to Ϫ20°C (36). The cells were converted to spheroplasts and the plasmid DNAs were purified, resolved in two-dimensional agarose gels (21), and blotted onto nylon membrane. The membranes were probed for 2-m plasmid sequences with a 32 P-labeled DNA fragment prepared using a random-primed DNA labeling kit from U. S. Biochemical Corp. Plasmid DNA topoisomer distributions were visualized by autoradiography.
Top1p Purification and in Vitro DNA Cleavage Assays-Wild-type and mutant Top1 proteins were purified to homogeneity from galactoseinduced cultures of JCW1 cells transformed with YCpGAL-TOP1, YCp-GAL1-top1T722A, or YCpGAL1-top1R517G, as described (21). The specific activity of the purified proteins was determined in plasmid DNA relaxation assays, while camptothecin sensitivity was assayed in DNA cleavage assays (21,22). For DNA cleavage, a single 3Ј-end labeled, 944-base pair DNA fragment was purified from plasmid pBlue-site, which is pBluescript II (Stratagene) containing a high affinity DNA topoisomerase I cleavage site (37) ligated into the EcoRI site. Approximately 10 ng (5100 cpm) of DNA was incubated with purified preparations of Top1p (37 ng), Top1T722Ap (25 ng), or Top1R517Gp (48 ng) in 50-l reaction volumes containing 50 mM Tris (pH 7.5), 50 mM KCl, 10 mM MgCl 2 , and 4% Me 2 SO. Where indicated, camptothecin was added to a final concentration of 100 M. After 30 min at 30°C, the reactions were terminated by the addition of 1% SDS and heating to 75°C for 10 min, treated with 0.4 mg/ml proteinase K, and the cleaved DNA fragments were resolved in a 7 M urea, 8% polyacrylamide gel and visualized by autoradiography.
To assess the relative stability of the cleavable complexes formed in the presence or absence of camptothecin, the DNA was incubated with Top1p (25 ng), Top1T722Ap (17 ng), or Top1R517Gp (32 ng) in 250-l reaction volumes. As above, a final 100 M camptothecin was added as indicated. After 10 min at 30°C, a 50-l aliquot was removed and treated as a no salt control. Additional KCl was added to a final 500 mM in the remaining volume. At the times indicated, 50-l aliquots were treated with 1% SDS and heating to 75°C and the extent of DNA cleavage determined as above.
Fluorescence Microscopy-Yeast cells were fixed with formaldehyde, stained with DAPI, 1 and mounted on glass slides with media containing calcofluor, as described (38). The cells were viewed on a Nikon Optiphot photomicroscope with an epi-fluorescence attachment EF-D mercury set with UV filter block.
For rapid staining of DNA in RAD9 and rad9⌬ strains, the cells were pelleted by centrifugation, washed with dH 2 O, and fixed in 70% ethanol. After 1 h, the cells were resuspended in 0.2 M Tris-HCl (pH 7.5) containing 3-5 l of DAPI (1 mg/ml) and incubated for 3 min at room temperature. Following three successive washes in 0.2 M Tris-HCl (pH 7.5), the DAPI-stained cells were mounted and viewed as described above. Yeast cells, fixed in 70% ethanol, were prepared for flow cytometry as described (15).
Mitotic Recombination at the rDNA Locus-CY185 cells, transformed with plasmids YCpScTOP1, YCpSctop1T722A, or YCp50, were grown in SC-uracil, -leucine, -adenine, dextrose media, and plated onto sectoring medium (39) containing: yeast nitrogen base, ammonium sulfate, 20 mg/liter each of histidine and tryptophan, 100 mg/liter of lysine; 6 mg/liter of adenine, and 2% glucose. The resultant colonies were scored for the presence of red sectors within a white colony. Completely red colonies resulted from the loss of the ADE2 marker from the rDNA locus prior to plating, whereas colonies with individual red sectors represented recombination events that occurred after plating. Representative colonies were photographed using Kodak Ektachrome 160T slide film at 8 ϫ magnification.

pGAL1-promoted Expression of top1T722A or top1R517G Is
Lethal in Yeast-Mutation of conserved amino acid residues in eukaryotic DNA topoisomerase I has previously been shown to affect enzyme activity and sensitivity to camptothecin (6,19,21,22). Substitution of Phe or Ser for the active-site tyrosine in yeast or human DNA topoisomerase I (Tyr 727 or Tyr 723 , respec-tively) abolishes catalytic activity (31,40,41). In the yeast mutant top1vac, changing the conserved Ile 725 -Asn 726 residues to Arg-Ala altered the sensitivity of the enzyme to a variety of DNA topoisomerase poisons, without any apparent effect on specific activity (21,22).
To expand these studies on DNA topoisomerase I function, oligonucleotide-directed mutagenesis of the conserved residues preceding the active-site tyrosine (Tyr 727 ) or of the conserved residues surrounding Arg 517 was undertaken as described under "Experimental Procedures." Two mutant alleles, top1T722A, which encodes a Thr 722 to Ala substitution, and top1R517G, in which Arg 517 is changed to Gly, were cloned under the GAL1 promoter in a single copy ARS/CEN plasmid and transformed into top1⌬ yeast strains.
Expression from the galactose-inducible GAL1 promoter produces a significant increment in TOP1 expression, relative to that obtained with the endogenous yeast TOP1 promoter (21,22). This increased level of wild-type DNA topoisomerase I is easily tolerated in top1⌬ strains (Fig. 1). However, GAL1 promoted expression of top1T722A or top1R517G produced a substantial drop in yeast cell viability that was evident 2 h following galactose induction (Fig. 1). By 8 h, cell viability had dropped by more than 30-fold in top1T722A expressing cells and 10-fold in top1R517G expressing cells. This time course closely paralleled the accumulation of Top1 protein in these cells, as assessed in immunoblots of crude cell extracts (data not shown).
Top1T722Ap and Top1R517Gp Exhibit Enhanced Stability of the Covalent Enzyme-DNA Intermediate-To examine the mechanism of top1-mediated cytotoxicity, the mutant proteins Top1T722Ap and Top1R517Gp were purified to homogeneity as described (21). When incubated in plasmid DNA relaxation assays (21,22), the specific activity of Top1T722Ap was reduced about 3-fold relative to wild-type Top1p (data not shown). However, as shown in Fig. 2, mutation of Thr 722 to Ala increased the stability of the covalent intermediate formed between the mutant enzyme and DNA. In these assays, purified Top1p was incubated with a single, 3Ј-end labeled DNA fragment in the presence or absence of camptothecin. After 30 min at 30°C, the covalent complexes were irreversibly trapped by the addition of 1% SDS and heating to 75°C. The extent of covalent-intermediate formation could then be assessed by re-solving the reaction products in a DNA sequencing gel and quantitating the relative amounts of cleaved and intact DNA substrates (21,22). In the absence of camptothecin, Top1T722Ap produced an average 9.4-fold increase in the intensity of cleaved DNA bands, relative to wild-type Top1p. Although the pattern of cleavage sites differed, this increment in Top1T722A-induced DNA cleavage was comparable to the 7.6-fold increase in DNA cleavage induced by camptothecin plus wild-type Top1p. Upon addition of camptothecin to the Top1T722A-containing reactions, the distribution of cleavage sites reverted to that of wild-type Top1p plus drug. In reactions corrected for protein concentration, overall band intensity was about 1.8-fold higher, suggesting that the Top1T722Ap was actually hypersensitive to camptothecin. Similar results were obtained with shorter incubation times.
In contrast, the catalytic activity of Top1R517Gp was only detectable in purified preparations and was reduced ϳ1,000fold relative to wild-type DNA topoisomerase I (data not shown). An increase in Top1R517Gp-mediated cleavage was also observed in the absence of camptothecin, albeit at a small number of sites, such as the high affinity Top1p cleavage site (22,37) shown in Fig. 2. Top1R517Gp was resistant to camptothecin, as the pattern of DNA cleavage was unaffected by addition of the drug (Fig. 2). However, the Top1R517Gp-DNA complexes, once formed, appeared to be quite stable. This was evidenced by the increased half-life of the Top1R517Gp-DNA covalent complexes in high salt, relative to those formed either by Top1T722Ap or Top1p in the presence or absence of camptothecin (Fig. 3). In all cases, the observed differences in en- zyme activities were not due to any detectable differences in DNA binding, as assayed in DNA band shift assays (data not shown). Taken together, these data suggest a common mechanism of top1-induced cell killing, resulting from alterations in the stability of the covalent intermediate formed between the mutant enzyme and DNA.
The Active-site Tyrosine Is Required for the Cytotoxic Action and Catalytic Activity of the Mutant Top1 Proteins-As the covalent intermediate is formed by a linkage between the active-site tyrosine (Tyr 727 ) and the 3Ј-phosphate of the cleaved DNA, such a mechanism of top1-induced lethality could be directly addressed by mutating Tyr 727 to Phe. As seen in Fig. 4, galactose-induced expression of top1T722A produced a greater than 3-log drop in cell viability compared with the top1⌬ controls, or cells overexpressing TOP1 or top1Y727F. These cytotoxic effects were abolished in cells expressing the top1T722A,Y727F double mutant. The active-site tyrosine was also required for top1R517G-induced lethality ( Fig. 4 and data not shown). Top1Y727Fp is catalytically inactive (22,31), yet pGAL1 promoted expression of top1Y727F and top1T722A,Y727F produced a slight inhibitory effect on cell growth (Fig. 4). Although under investigation, these results cannot be attributed to the absence of DNA topoisomerase I activity, as the same effect was not seen in the top1⌬ controls. The introduction of the Y727F mutation had no discernible effect on Top1p stability, as judged in immunoblots of crude cell extracts (data not shown).
The lack of Top1T722A,Y727Fp activity was confirmed by assaying DNA topoisomerase I activity in vivo. In these experiments, JCW28 cells (top1⌬,top2ts) were co-transformed with plasmid YEptopA-PGPD, which constitutively expresses bacterial DNA topoisomerase I, and a second vector expressing the indicated top1 allele from the pGAL1 promoter. Transformed cell cultures were induced for 5 h with galactose at the permis- sive temperature (25°C), then shifted to the nonpermissive temperature (35°C) for an additional 3 h to inactivate DNA topoisomerase II. Changes in 2-m plasmid DNA topology were then assessed following plasmid purification and two-dimensional gel electrophoresis (21).
As described by Giaever and Wang (35), 2-m plasmid DNA transcription produces localized domains of DNA supercoiling. In the absence of endogenous DNA topoisomerase I or II activity at 35°C, the bacterial topA gene product preferentially relaxes negative supercoils, so that positively supercoiled plasmid DNA accumulates. However, if a catalytically active yeast DNA topoisomerase I is expressed from the pGAL1 promoter in these cells, the positively supercoiled DNA domains will also be relaxed (21). As shown in Fig. 5, in the absence of yeast DNA topoisomerase I (the pGAL1 control), the 2-m plasmid DNA is highly positively supercoiled. This corresponds to the plasmid DNA topoisomers present at the lower right hand tip of the arc, (labeled (ϩ)). Wild-type Top1p and TopT722Ap were both active inside yeast, as evidenced by a quantitative shift in the distribution of the plasmid DNA topoisomers to that of negatively supercoiled DNA (labeled (Ϫ)). In contrast, mutation of the active-site tyrosine in Top1T722A,Y727Fp abolished enzyme activity; the plasmid DNA topoisomers remained positively supercoiled. Surprisingly, a low level of Top1R517Gp activity was also detected, despite the more than 1,000-fold reduction in specific activity of the purified Top1R517Gp. This in vivo assay may be a more sensitive measure of DNA topoisomerase I catalytic activity.
top1T722A and top1R517G Expression Induces DNA Damage and a Terminal G 2 -arrested Phenotype-Treatment of yeast cells expressing TOP1 with camptothecin induces expression of the DNA damage responsive genes, DIN3 and RNR3 (23,24), leads to increased rates of rDNA recombination (23), and produces a terminal phenotype consistent with arrest in G 2 of the cell cycle (15). Given the biochemical similarities between mutant Top1p activity and camptothecin action, we examined the effects of mutant top1 expression on cell cycle progression and rDNA recombination.
Cells overexpressing wild-type TOP1 were distributed throughout the cell cycle, as evidenced by the presence of unbudded, small-budded, and large-budded cells (Fig. 6D). Staining of these cells with DAPI and calcofluor, to visualize the DNA and chitin-containing bud scars, respectively (38), indicated that the large-budded cells had completed nuclear segregation (Fig. 6E). In contrast, after 6 h induction with galactose, a greater proportion of large-budded cells were present in cells overexpressing top1R517G and top1T722A (Fig. 6, A-C and F). Furthermore, these cells typically contained a single mass of nuclear DNA at or extending across the neck that connects mother and bud (Fig. 6, B, C, and F). A similar terminal phenotype was previously reported for camptothecin-treated yeast cells overexpressing wild-type TOP1 (15) and is consistent with cell cycle arrest in G 2 (42).
Similar results were obtained by flow cytometry. In comparison with the normal cell cycle distribution obtained with yeast cells expressing TOP1, cells expressing top1T722A or top1R517G accumulated with a G 2 (2N) DNA content (Fig. 7). The greater than 2N DNA shoulder observed for top1T722A expressing cells was also reported for camptothecin-treated cells (15). This may represent increased propidium iodide binding to fragmented DNA or increased DNA content due to repair synthesis or increased mitochondrial DNA replication. Nevertheless, this shift is more pronounced in the cells expressing top1T722A than in cells expressing top1R517G, consistent with the increased lethality associated with top1T722A expression seen in Fig. 1. Cells overexpressing the catalytically inactive mutant, top1Y727F, exhibited a higher proportion of cells in G 1 relative to the TOP1 control. Although top1⌬ cells exhibit a slow growth phenotype at lower temperatures (17), these data might also reflect the reduced growth rate of cells overexpressing Top1Y727F protein observed in Fig. 4.
DNA topoisomerase I and II suppress mitotic recombination between the tandem repeats of the rDNA cluster (29,30). TOP1ϩ yeast cells treated with sublethal doses of camptothecin exhibit elevated rates of rDNA recombination, presumably due to the repair of drug-induced DNA damage (23). Hyper-rDNA recombination was also observed in yeast cells constitutively expressing low levels of the top1 mutant, top1-103 (25). Recombination at the rDNA locus was assessed by scoring for the loss of the ADE2 gene, inserted into the rDNA locus in the haploid yeast strain CY185 (top1⌬, ade2-1, rDNA::ADE2) (25). Constitutive expression of sublethal levels of top1T722A, from the TOP1 promoter, produces a substantial increase in rDNA recombination over the lower levels of red sectoring observed in the top1⌬ controls (compare Fig. 8, A and  C). Constitutive expression of TOP1, on the other hand, suppressed homologous rDNA recombination as the majority of the colonies were mostly white (Fig. 8B). We were unable to construct a similar expression vector with top1R517G. Since yeast TOP1 promoter function has been demonstrated in Escherichia coli (17) it is possible that top1R517G expression is especially cytotoxic in bacteria, resulting in plasmid rearrangements. While these data support the idea that top1 mutant expression is recombinogenic, this effect appears to be limited to the rDNA locus. Similar studies of mutant top1 expression in diploid strains did not produce a detectable increase in recombination between heteroalleles at a variety of genetic loci (data not shown).
Deletion of the RAD9 Checkpoint Enhances top1-induced Cell Killing-The RAD9 checkpoint is activated by DNA damage and induces a delay in cell cycle progression, to allow for the completion of DNA replication and repair before cells undergo mitosis (43)(44)(45). DNA lesions cause RAD9 cells to arrest in G 2 .
In contrast, rad9 mutant cells die more rapidly, as they proceed into mitosis with unrepaired DNA.
Previous studies demonstrated that RAD9 cells overexpressing mutant top1-103 arrested in G 2 phase of the cell cycle and remained viable for at least 24 h (25). However, an isogenic rad9⌬ strain rapidly lost viability. In the present studies, pGAL1 expression of top1R517G or top1T722A, in a RAD9 background, produced an immediate decrease in cell viability (see Figs. 1 and 9). Yet, the acquisition of a terminal G 2arrested phenotype in these cells (Fig. 6) implied the function of cell cycle checkpoint(s) in response to mutant Top1p-induced DNA damage. Indeed, as shown in Fig. 9, the lethality associated with top1R517G and top1T722A expression was enhanced by deletion of the RAD9 gene. After 8 h induction, rad9⌬ cells expressing top1R517G or top1T722A experienced a 10 3 drop in the number of viable cells, compared with a 10 -30-fold decrease, respectively, in the viability of RAD9 cells. The differences in the cytotoxic action of Top1T722Ap and Top1R517Gp was also abolished in rad9⌬ strains, suggesting that the RAD9 checkpoint was more responsive to Top1R517Gp-induced DNA damage. This is borne out by the data in Table I. While deletion of RAD9 has little effect on the overall percentage of largebudded cells in cultures expressing top1T722A, top1R517G or TOP1, rad9⌬ cells expressing top1R517G exhibited a 9-fold increase in the percent of large-budded cells having completed nuclear segregation. In contrast, the percentage of completely segregated nuclei in the large-budded cells expressing Top1T722Ap, only increased from 12 to 30%. Although overexpression of the top1 mutants is lethal even in the presence of a functional G 2 checkpoint, deletion of RAD9 enhances the cytotoxic action of the mutant proteins, with more pronounced effects on the segregation of DNA damaged in the presence of Top1R517Gp. activity of Top1T722Ap was ϳ3-fold less than wild-type Top1p and the mutant enzyme exhibited increased sensitivity to camptothecin-induced DNA cleavage. On average, Top1T722Ap also produced a 9.4-fold increase in DNA cleavage at a large number of sites. Taken together, these data indicate that substitution of Ala for Thr 722 increases the stability of the cleavable complex without affecting sequence specificity.
Top1R517Gp activity, on the other hand, was down about 1,000-fold and no increment in DNA cleavage was detected in the presence of camptothecin. Although a significant reduction in the number of DNA cleavage sites was observed with Top1R517Gp, relative to the wild-type enzyme, the complexes once formed were extremely salt stable. In contrast to wild-type Top1p and Top1T722Ap, these results suggest the formation of persistent Top1R517Gp-DNA covalent intermediates. Despite these mechanistic differences, the mutant proteins both exhibited alterations in the stability of the covalent intermediate formed between the active-site tyrosine in the enzyme and the 3Ј-phosphoryl end of the cleaved DNA.
A number of residues surrounding the active-site tyrosine have been shown to be important for enzyme activity and camptothecin sensitivity (21,22,46,47). 2 Indeed substitution of Arg-Ala for Asn 725 -Ile 726 in the yeast mutant top1vac and Leu for Asn 726 in top1N726L, have previously been shown to adversely affect the viability of cells deficient in the repair of double strand DNA breaks, due to the deletion of RAD52 (21,22). Thus, substitutions of adjacent, conserved residues, such as Thr 722 , might also be expected to affect DNA topoisomerase I function. Indeed, mutation of Thr 718 , at the corresponding position in human DNA topoisomerase I, has similar cytotoxic effects when overexpressed in yeast top1⌬ strains. 3 Experiments are currently underway to determine if the cytotoxic action of top1T722A results from a decrease in the rate of DNA religation, as proposed for camptothecin (8,11), or from an increase in the rate of DNA cleavage. In either case, the fact that the active-site tyrosine is required both for Top1T722Ap catalytic activity and cytotoxicity underscores the importance of the formation of the covalent complex in top1T722A-induced cell killing.
In the case of top1R517G, however, the effects of the Arg 517 to Gly substitution on catalytic activity, camptothecin sensitivity, and cell viability were quite surprising. Arg 517 lies COOHterminal to the "camptothecin-binding loop" proposed by Lue et al. (48), based on their recent report of the structure of a more amino-terminal fragment of yeast DNA topoisomerase I and the mapping of several camptothecin-resistant mutations to this region of the enzyme (19,49). The data presented herein suggest a mechanism of cell killing by Top1R517Gp via the stabilization of the covalent enzyme intermediate. The possibility is currently being explored that this lethal phenotype results from alterations in enzyme structure, rather than a direct interaction of this region of the enzyme with the active site. For example, the changes in the pattern of DNA cleavage sites observed in vitro may reflect altered structural requirements for efficient binding of DNA substrates by the mutant enzyme (reviewed in Ref. 50). Changes in enzyme structure could also alter the binding site for camptothecin, as well as the rates of DNA cleavage and religation. Nevertheless, the low level of long-lived protein-linked DNA breaks observed in vitro would be sufficient to induce cell killing, as even a single, persistent double-stranded DNA break is cytotoxic in yeast (51,52).
Camptothecin reversibly stabilizes the covalent intermediate formed between DNA topoisomerase I and DNA, resulting in DNA damage, inhibition of DNA synthesis, cell cycle arrest in G 2 , and cell death (4,6,8). Lethal doses of the drug produce a G 2 -arrested terminal phenotype, with a single DNA mass in the mother cell or caught in the neck between mother and bud (15). Similar effects were seen following the induction of top1T722A and top1R517G expression; an immediate decrease in cell viability was accompanied by an increase in G 2 -arrested cells. Flow cytometric analysis established that the majority of these cells had a G 2 (2N) content. Moreover, as with camptothecin-treated cells (15) Aliquots of galactose-induced cultures of RAD9 and rad9⌬ cells, transformed with YCpGAL1-TOP1, YCpGAL1-top1R517G, YCpGAL1-top1T722A, or YCpGAL1, were taken at the indicated times, serially diluted, and plated on selective media containing dextrose. Shown are the number of viable cells relative to t ϭ 0. Cellular and nuclear morphologies were also examined (see Table I).

TABLE I RAD9 checkpoint function affects the cell cycle progression of cells overexpressing top1R517G
As in Fig. 9, expression of the indicated TOP1 allele was induced for 8 h in cultures of RAD9 and rad9⌬ cells. Cellular morphologies were distinguished as follows: large-budded cells had daughter buds greater than half the diameter of the mother cell. These were further subdivided into cells with an undivided nucleus (a single nuclear mass positioned either at or in the neck region between mother and bud) and those with completely segregated nuclei. An average of 250 cells were examined for morphology and ϳ100 large-budded cells were evaluated for differences in nuclear segregation. staining of fragmented chromosomal DNA or an increase in mitochondrial DNA synthesis (53).
Mutant top1 expression also produced an increase in rDNA recombination. Recombination between heteroalleles in diploid strains appeared to be unaffected. The ability of these mutant proteins to enhance the rate of illegitimate recombination has not yet been addressed. However, since DNA topoisomerase I has been implicated as playing a role in illegitimate recombination (54), it is tempting to speculate that a low level of DNA damage accompanying wild-type DNA topoisomerase I function may be enhanced by drug treatment or top1 mutation, resulting in increased genomic DNA fragmentation and rearrangement. This idea is supported by the fact that extremely high levels of wild-type DNA topoisomerase I are lethal in a rad52⌬ background, presumably due to an increase in double-stranded DNA breaks. Moreover, such a mechanism of genomic rearrangements has been invoked for DNA topoisomerase II in epipodophyllotoxin-associated translocations (55). Similar effects may also result from the chemotherapeutic application of camptothecin derivatives.
The RAD9 gene product is required to arrest cell division in response to DNA damage, such as that induced by X-irradiation or replication defective cdc mutants (43)(44)(45). Recently, Lydall and Weinert (43) proposed that the coordinated action of Rad9p with the Rad17, Rad24, and Mec3 checkpoint proteins functions in the repair of damaged DNA. Mec1p and Rad53p, on the other hand, are required for cell cycle arrest at G 2 /M and in S phase in response to inhibition of DNA replication (56).
Expression of the DNA topoisomerase I mutant top1-103 caused a RAD9-dependent cell cycle arrest in G 2 (25) and the cells remained viable for over 24 h. In comparison, rad9⌬ cells did not arrest in G 2 and rapidly lost viability in response to top1-103 expression. The cytotoxic activity of Top1T722Ap and Top1R517Gp, on the other hand, was evident even in checkpoint proficient, repair competent cells. Although rad9⌬ strains experienced enhanced top1-mediated killing, the cells still accumulated with a terminal large-budded phenotype. This increment in rad9⌬ cell killing was more pronounced in top1R517G expressing cells, and was accompanied by an increase in the percentage of large-budded cells having completed nuclear segregation. Indeed, in the absence of Rad9p, the DNA damage may be amplified due to unregulated DNA degradation (43). Nevertheless, these results suggest that while the RAD9 checkpoint, and presumably other DNA damage-responsive checkpoints, respond to top1-induced DNA damage, overexpression of Top1T722Ap and Top1R517Gp exceeds the cells' ability to repair these lethal lesions.