Identification of a Mutant Human Topoisomerase I with Intact Catalytic Activity and Resistance to 9-Nitro-camptothecin*

Human U-937 myeloid leukemia cells were selected for resistance to increasing concentrations of the camptothecin derivative, 9-nitro-20(S)camptothecin (9-NC). The isolated single cell clone, designated U-937/CR, was approximately 20-fold resistant to 9-NC. Analysis of topoisomerase I (topo I) gene expression in U-937/CR cells demonstrated similar mRNA levels as compared with U-937 cells. Immunoblotting with an anti-top0 I serum revealed reactive proteins at 100, 75, and 67 kDa which were expressed at the same level in the parental and 9-NC-resistant clones. These cell lines also demon- strated similar levels of top0 I catalytic activity as determined by assaying nuclear extracts for relaxation of supercoiled plasmid DNA. In contrast, catalytic assays performed in the presence of 9-NC demonstrated that top0 I activity from U-937/CR cells was approximately 10-fold more resistant than that from U-937 cells. Nucleotide sequencing of top0 I cDNAs revealed the sub- stitution of phenylalanine ( E C ) at residue 361 in U-937 cells with serine (TCC) in the 9-NC-resistant clone. Ex- pression and partiii purification of the mutant top0 I protein in Escherichia coli products reactions.

other domains of top0 I are required for this event (7). Additional functions of top0 I may involve a direct role in gene transcription. Studies have shown that top0 I is required for transcription of supercoiled rRNA gene in vitro (8) and that microinjection of anti-top0 I antibodies into cells is associated with inhibition of transcription (9).
Top0 I has been identified as a cellular target for the plant alkaloid camptothecin (CPT) (3,lO-12). The available evidence indicates that CPT binds to the top0 I.DNA complex subsequent to the DNA cleavage step. While labeled CPT interacts reversibly with this complex, there is no detectable binding of this agent to isolated top0 I or purified DNA (13,14). Other studies have demonstrated that while CPT has little effect on top0 I-mediated DNA cleavage, this agent inhibits the religation step (15, 16). Moreover, CPT has been found to preferentially stabilize top0 I-mediated cleavage of T-G linkages following formation of the covalent bond between tyrosine and the 3"phosphate group of thymidine (17). These findings have supported the availability of a CPT-binding site upon formation of the top0 I.DNA complex. Although the nature of the CPT-binding site remains unknown, interaction of CPT and the top0 I.DNA complex appears to be necessary for the cytotoxic effects of this agent. For example, while CPT inhibits top0 I activity (lo), the concentration necessary for such inhibition is considerably higher than that required for induction of cell lethality. Indeed, the available evidence supports a model in which CPTinduced stabilization of the top0 I.DNA complex is associated with conversion of the single strand nicks to irreversible double strand breaks (18,19).
Other studies have demonstrated that yeast cells devoid of top0 I are resistant to the lethal effects of CPT (20,21). Conversely, cells that overexpress top0 I have been found to be hypersensitive to this agent (21,22). These findings are in concert with the demonstration that mammalian cell lines selected for resistance to CPT exhibit decreased levels of top0 I expression. In resistant murine P388 leukemia cells, CPT treatment is associated with decreased DNA single strand breaks compared to wild-type cells and decreased expression of topo I at the mRNA and protein levels (23). Similar findings have been obtained in hamster (24) and human (25,26) cell lines. Decreases in top0 I expression have been associated with rearrangement and hypermethylation of the top0 I gene (27). Other studies demonstrating that activity of purified top0 I from CPT-resistant cells is unimpaired in the presence of this agent have supported alterations in the enzyme that confer a resistant phenotype (25,28,29). In this context, several studies have identified mutations in the top0 I gene in association with the development of CPT resistance. A point mutation resulting in replacement of threonine with alanine at residue 729 has been found in a CPT-resistant human lung cancer line (30), while changes at residues 533 and 583 have been identified in resistant human leukemia cells (31). Thus, decreased expres- Recent work has resulted in the synthesis of certain CPT derivatives that are cytotoxic to tumor cells. For example, 9-nitro-2O(S)camptothecin (9-NC). inhibits the growth of human carcinoma and melanoma cells in vitro and induces regression of these tumors in immunodeficient mice (32-35). Other studies have demonstrated that exposure of human U-937 myeloid leukemia cells to this agent is associated with induction of early response gene expression and internucleosomal DNA fragmentation (36). The present work describes the isolation of a U-937 cell clone selected for resistance to 9-NC. The results demonstrate that these resistant cells exhibit similar levels of top0 I catalytic activity compared to that in parental cells and that the top0 I gene is mutated at a potential CPT-binding site in the enzyme.

MATERIALS AND METHODS
Drugs-CPT and 9-NC were prepared and purified as described (37). Both drugs were suspended in polyethylene glycol (PEG 400; Aldrich), divided into small aliquots, and stored at -70 "C. Etoposide was obtained from Bristol-Myers (Evansville, IN).
Cytotoxicity Assays-Drug cytotoxicity assays were performed as described using the tetrazolium-based compound 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (38). Briefly, 100 pl of diluted drug or drug diluent alone was added to logarithmically growing cells (2 x 104/100 pl) in a 96-well plate. After incubation for 48 h, 50 pl of 3 mdml MTl' diluted in PBS were added and the cells incubated for an additional 3 h. Eighty pl of 23% SDS in 50% dimethylformamide were then added and the plates incubated overnight at 37 "C to allow solubilization of the formazan crystals. The absorbance values for each well were determined in an enzyme-linked immunosorbent assay reader (model MA310 Automated EIA Plate reader, Whittaker M. A. Bioproducts, Inc., Walkersville, MD) at a wavelength of 550 nm. ICso values were obtained by a linear regression analysis of percent absorbance versus log drug concentration.
RNA Isolation and Northern Blot Hybridization-Total cellular RNA was isolated by a modification of the guanidine isothiocyanate technique (36). RNA (20 pg/lane) was separated by electrophoresis in 1% agarose, 2.2 M formaldehyde gels, transferred to nitrocellulose filters, and hybridized to the following 32P-labeled DNA probes: (1) the 3.4-kb BarnHI-EcoRI insert of the plasmid ptac-hTOPl (39) and (2) the 2.0-kb PstI fragment of the chicken p-actin gene from the PA1 plasmid (40). Hybridizations were performed as described (36). The filters were washed and exposed to Kodak X-Omat XAR film using an intensifying screen. Autoradiograms were scanned and analyzed using an LKB (Bromma, Sweden) Ultroscan XL laser densitometer and the Gelscan XL software package.
Nuclear Extracts-Nuclear extracts were prepared as described (41).
Briefly, lo8 cells were collected by centrifugation and washed with icecold PBS and then ice-cold NB (2 m~ KH,P04, 5 m~ MgC12, 150 m~ NaCl, 1 m~ EGTA, 0.1 m~ dithiothreitol, 0.1 m~ PMSF, pH 6.5). All subsequent steps were performed at 4 "C. Washed cells were resuspended in 1 ml of NB and 9 ml of NB containing 0.35% Triton X-100. The cell suspension was gently rotated for 10 min and centrifuged at 1,000 x g for 10 min. The resulting nuclear pellet was washed with Triton-free NB, and resuspended in NB containing 0.35 M NaCl. The suspension was incubated with gentle rotation for 30 min, and the extract centrifuged at 17,000 x g for 10 min. The supernatant was collected and protein concentration determined by a modification of the Bradford method (42) (Bio-Rad Protein Assay, Bio-Rad). After adding glycerol to a final concentration of 40%, the extract was divided into aliquots and stored at -20 "C for up to 2 weeks. Zmrnunoblotting-Nuclear extracts were subjected to electrophoresis in 7.5% SDS-polyacrylamide gels and then transferred to nitrocellulose filters. The blots were incubated in 5% dry milk in PBST (PBS with 0.1% 'heen 20) for 1 h at 20 "C, followed by washing with PBST and incubation with a human top0 I antibody or a rabbit top0 I1 antibody ('Ibpogen, Inc., Columbus, OH) for I h at 20 "C. The blots were washed with PBST and incubated with an anti-human or anti-rabbit IgG peroxidase conjugate (Amersham Corp.) for 1 h at 20 "C. After additional washing, the blots were developed using an enhanced chemiluminescence technique (ECL detection system, Amersham Corp.) and Kodak X-Omat XAR film. Photographic negatives were scanned with a laser densitometer as described above.
lbpo Z Enzyme Assays of Nuclear Extracts-'hpo I enzyme activity was measured by a DNA relaxation assay (43) using supercoiled piasmid DNA derived from pGEM-3Z (Promega, Madison, WI). Reaction mixtures (20 pl of final volume) contained 0.5 pg of supercoiled DNA, 10 m~ Tris-HC1, pH 7.5,50 m~ KCl, 5 m~ MgC12, 0.1 m~ EDTA, 15 pdml bovine serum albumin, 0.5 m~ dithiothreitol, and 2 pl of nuclear extract diluted in NB. In experiments using 9-NC, the mixtures also contained 2 pl of drug diluted in water immediately prior to use. Reactions were performed at 37 "C for 30 min and terminated by the addition of 2.2 pl of a 10 x loading buffer (5% SDS, 0.3% bromphenol blue, 16% Ficoll400, and 10 m~ NaH2P04). Samples were loaded onto a 0.8% agarose gel and subjected to electrophoresis for 6-8 h at 3 V/cm. The electrophoresis buffer consisted of 30 n m NaH2P04, 36 m~ Tris-HC1, and 1 m~ EDTA, pH 7.8. The gels also contained 2 d m 1 chloroquine to separate nicked, relaxed, and supercoiled DNA (17,44). After staining with ethidium bromide, gels were photographed with Kodak Ektapan film and the negatives scanned as described above. One unit of top0 I catalytic activity was defined as the amount of enzyme necessary to relax 50% of 0.5 pg of supercoiled DNA in 30 min at 37 "C. Top0 Z cDNA Sequencing-RNA was extracted from cells as described (36). For reverse transcription, 1 pg of total RNA was used in a 20-pl reaction mixture containing 1 m~ of each dNTP, 1 uniffpl RNase inhibitor, 2.5 unitdpl Moloney murine leukemia virus reverse transcriptase, and 2.5 p~ of random hexamers (GeneAmp RNA PCR Kit, Perkin Elmer Cetus, Norwalk, CT). The mixture was successively incubated at 42 " c for 15 min, at 99 "C for 5 min, and at 5 "C for 5 min. Oligonucleotide primers of 25-35 base pairs were added at a final concentration of 0.15 p i and were designed to amplify regions of 500-1000 base pairs of a human top0 I cDNA according to the GenBank sequence (Table I).
Oligonucleotides were synthesized on an Applied Biosystems DNA synthesizer model 394 using 8-cyanoethyl phosphoramidite chemistry and were purified by ethanol precipitation before use. After addition of 2.5 units of Tuq DNA polymerase, thermal cycling was performed at 95/60/ 72 "C using 30-s intervals for a total of 35 cycles. The products were analyzed by electrophoresis in a 1.5% agarose gel in TAE (0.04 M Tris acetate, 1 m ethylenediaminetetracetate), and the amplified fragments were removed and purified from the gel using an affinity matrix (Glassmilk; Bio 101, Inc., La Jolla, CA). Approximately 1 ng of the purified fragment was sequenced by the dideoxynucleotide method using thermocycling and modified Tup DNA polymerase. Briefly, 10 pmol of one of the flanking primers or an internal primer were end-labeled with [Y-~~PIATP (New England Nuclear) and T4 polynucleotide kinase (Promega, Madison, WI). The primers were used in 6 pl of reaction mixtures containing Sequencing Grade Taq DNA polymerase and dNTP/ddNTP mixtures (fmol DNA Sequencing System, Promega). Thermal cycling was performed at 95/70 "C using 30-s intervals for 30 cycles. The products were analyzed in a 6% polyacrylamide gel containing 8.3 M urea and TBE (0.1 M Tris-borate, 2 m EDTA, pH 8.3). When electrophoresis was terminated, the gel was dried on filter paper and exposed for 1648 h to Kodak film using an intensifylng screen.
Expression of Mutant U-937 Top0 Z in E. coli-A human topoisomerase I expression plasmid was constructed by cloning the BumHI-EcoRI fragment of the plasmid ptac-hTOP1 (kindly provided by Dr. J. C. Wang (39)) into the vector pGEX-lAT (Pharmacia LKB Biotechnology, Inc.). The resultant plasmid, pGEX-TOP1, contains a tuc promoter controlling the expression of a fusion protein consisting of glutathione Stransferase (GST) linked to the amino terminus of human top0 I. A plasmid expressing the mutant top0 I was constructed by replacing the NdeI-SphI fragment in pGEX-TOP1 with a fragment corresponding to this region in the resistant top0 I cDNA. The resultant plasmid, pGEX-MTOP1, was sequenced to confirm the presence of the mutation. DH5aF'IQ E. coli (Life Technologies, Inc.) infected with either pGEX-TOP1 or pGEX-MTOP1 were grown in 50 ml of 2 x YT media containing 2% (w/v) glucose and 75 pg/ml ampicillin to an OD, nm of approximately 0.5. ARer addition of isopropyl-P-D-thiogalactoside (IPTG) to 0.5 m, the bacteria were grown for 2 h and harvested by centrifugation. Lysates were prepared by resuspension of the pellets in 2.5 ml of icecold PBS containing 0.1 m PMSF and 10 pdml leupeptin, followed by sonication on ice. Triton X-100 was added to a concentration of 1% and the mixtures incubated on ice for 30 min. After centrifugation at 12,000 x g at 4 "C for 10 min, the supernatant was removed and incubated with 50 pl of a 50% (v/v) slurry of Glutathione Sepharose 4B (Pharmacia LKB Biotechnology, Inc.) for 2 h at 4 "C. The beads were washed three times with 250 pl of PBS containing 0.1 m PMSF and 10 pg/ml leupeptin. Bound proteins were eluted by incubation of the beads in 25 pl of elution buffer (10 rm glutathione, 50 m Tris-HC1, pH 8.0) for 10 min at room temperature. The elution step was repeated and the supernatants pooled. Glycerol was then added to a concentration of 30%, and the lysates were used immediately or stored at -20 "C. Top0 I catalytic activity was assayed as described above, with the exception that reactions were performed for 60 min in a buffer consisting of 10 n u Tris-HCl, pH 7.5, 10 rm EDTA, 100 m KCl, and 15 pdml bovine serum albumin.

RESULTS
Development of 9-NC-resistant Cells-U-937 cells were selected for resistance to 9-NC by continuous exposure to an initial concentration of 20 m drug for 2 weeks, and then growth for a similar period in media containing 40 m 9-NC. Cells were subsequently exposed to 60, 80, 100, 125, and 150 m 9-NC. Using this approach, a cell line, designated U-937/CR (CR, camptothecin resistant) was isolated as a single cell clone capable of sustained growth in media containing 150 m 9-NC. This resistant phenotype was stable aRer growth for over 2 months in media lacking 9-NC. The growth rate of U-937/CR cells was similar to that of the parental U-937 cells with a doubling time of approximately 24 h. Furthermore, U-937/CR cells remained tumorigenic when tested as xenografts in nude mice (data not shown, (46,47)).
Sensitivity of U-937 and U-937/CR cells to 9-NC was compared in MTT cytotoxicity assays. Parental u-937 cells were killed at a n IC50 of approximately 30 m 9-NC (Fig. 1). In  (Table 11). The finding that verapamil had no significant effect on the ICso for 9-NC in U-937/CR cells supported the lack of P-glycoprotein involvement in this resistant phenotype (Table  11). U-937/CR cells were also resistant to CPT, although the fold increase in IC5o compared with U-937 cells was less than that obtained with 9-NC (Table 11). Taken together, these findings indicated that U-937/CR cells are resistant to 9-NC, as well as CPT, and that the mechanism of resistance may involve alterations in top0 I expression.
Top0 Z Gene Expression-The basis for resistance of U-937/CR cells to 9-NC was studied by examining top0 I gene expression at the mRNA and protein levels. Northern blot analysis of total cellular RNA from U-937 cells revealed 4.0-kb transcripts that hybridized to the top0 I probe (Fig. 2). Similar findings were obtained with RNA from U-937/CR cells (Fig. 2). Moreover, intensity of the top0 I signals was comparable for the two lines (Fig. 2). The finding that the actin hybridization signals were similar for both lines confirmed equal loading of the lanes (Fig. 2).
Using polyclonal anti-topo I serum, three major immunoreactive proteins were detectable at approximately 100, 75, and 68 kDa in U-937 nuclear extracts (Fig. 3A). A similar pattern was obtained with extracts of U-937/CR cells, and the intensity of the signals was comparable for both lines (Fig. 3A). Analysis of duplicate preparations by Coomassie Blue staining indicated equal loading of protein (Fig. 3A). Since previous studies have demonstrated that CPT resistance is associated with increases in top0 I1 expression (481, we also reprobed the filters with a top0 I1 antiserum. Top0 I1 protein levels were increased 1.9-fold compared with the parental line (Fig. 3B). Taken together with the Northern analyses, these results demonstrate that top0 I expression in the resistant U-937/CR phenotype is indistinguishable from that in U-937 cells. Moreover, the 9-NC-resistant phenotype is associated with increased top0 I1 expression. Top0 Z Catalytic Activity-Catalytic activity of top0 I in U-937 and U-937/CR cells was determined by a plasmid relaxation assay performed in the absence of ATP (43). In this assay, varying amounts of nuclear protein were incubated with supercoiled plasmid DNA. In the absence of 9-NC, plasmid relaxation activity of nuclear extracts from U-937 cells was similar to that obtained with extracts from U-937/CR cells (Fig. a).
Quantitation of the supercoiled bands by densitometry confirmed similar levels of activity in U-937 and U-937/CR cells (Fig. a). These   Expressed as mean 2 standard error.
The CR values for 9-NC with or without verapamil are not statistically different ( p = 0.277, paired two-tailed t test). phenotype is unrelated to changes in top0 I activity.
Other assays of top0 I catalytic activity were performed in the presence of 9-NC. Under these conditions, the addition of 50 p~ 9-NC resulted in nearly complete inhibition of the plasmid relaxation activity in U-937 cell extracts (Fig. 5 4 ) . In contrast, the activity of U-937/CR cell extracts was inhibited only in part in the presence of 500 p~ 9-NC (Fig. 5 A ) . Densitometric quantification of the supercoiled bands indicated that the concentration of 9-NC required for 50% inhibition of top0 I activity was 10-fold greater for U-937/CR cells compared with U-937 cells (Fig. 5B). These results suggested that the resistant U-937KR phenotype is related to alterations in the ability of top0 I to interact with 9-NC and not to a decrease in top0 I catalytic activity.
Detection of a Mutated Top0 Z Gene-The finding that top0 I gene expression and catalytic activity are unchanged in U-937/CR cells prompted further studies on potential alterations in the top0 I gene. The nucleotide sequence of top0 I cDNAs from both cell lines was determined using reverse transcription followed by PCR. Overlapping partial top0 I cDNAs of 500-1000 base pairs were isolated which together encompassed the top0 I coding region (primer positions and sequences are listed in Table I). Direct sequencing of these products by the dideoxynucleotide method demonstrated that the top0 I sequence from both U-937 and U-937/CR cells contained two alterations compared to the GenBank sequence (49) utilized products from independent PCR reactions.

Expression of the Mutant Top0 Z Gene in E. coli-In order to
determine whether the alteration in the top0 I gene from U-937KR cells confers CPT resistance, we expressed the mutant cDNA in E. coli and assayed for sensitivity to 9-NC. In order to eliminate potential confounding effects of E. coli topoisomerases, we expressed human top0 I as a fusion protein linked to GST and partially purified the enzyme from bacterial lysates using glutathione Sepharose beads. lbpo I catalytic activity in the purified bacterial lysates was also assayed in the absence of Mg2+ (1). Under these conditions, there was no detectable catalytic activity in purified lysates from bacteria expressing GST alone (data not shown). In contrast, purified lysates from bacteria expressing a GST-topo I fusion protein exhibited top0 I catalytic activity that was inhibited by 50 p~ 9-NC (Fig. 7). Moreover, purified lysates from bacteria expressing a fusion protein containing the top0 I mutation at position 361 exhibited catalytic activity that was unaffected by 50 or 100 p~ of 9-NC (Fig. 7). These findings indicated that substitution of phenylalanine with serine at position 361 of human top0 I is sufficient to confer resistance to the inhibitory effects of 9-NC on catalytic activity.

DISCUSSION
Previous work has demonstrated that cellular resistance to CPT is associated with decreases in top0 I activity. Selection of P388 cells resistant to CPT has been associated with decreases (24-fold) in top0 I mRNA, immunoreactivity, and extractable enzymatic activity (23). This induction of CPT resistance has also been associated with rearrangements and increases in methylation of the top0 I gene (23,271. Other studies in human tumor cell lines have shown that resistance to CPT and its derivative CPT-11 is associated with a reduction, but not qualitative changes, in top0 I protein (26). Selection of human lung cancer cells for resistance to CPT-11 has been associated with a A. reduction in the amount of top0 I, as well as a decrease in sensitivity of the purified enzyme to this agent (25). Similar results have been obtained in V79 hamster lung cells (241, while another CPT-resistant line derived from hamster lung exhibited decreased catalytic activity and no change in immunoreactive top0 I protein (29). In contrast to these findings, the present results demonstrate that U-937 cells resistant to 9-NC exhibit little if any apparent decrease in top0 I catalytic activity. There was also no detectable change in top0 I mRNA and protein levels associated with acquisition of 9-NC resistance. These findings are thus in contrast to previous CPT-resistant isolates which have associated decreases in top0 I expression. Nonetheless, as demonstrated in other CPT-resistant lines (261, U-937/CR cells were found to have increased levels of top0 I1 protein and catalytic activity (data not shown). Moreover, the U-937/CR cells exhibited an increase in sensitivity to the top0 I1 inhibitor etoposide (IC50 = 0.14 2 0.018 p~; mean r S.E. of three experiments performed in duplicate) compared to parental U-937 cells (IC50 = 0.33 2 0.02 p~). These findings are consistent with the induction of top0 I1 expression in association with 9-NC resistance and support a defect in top0 I function not reflected in the in vitro assay of catalytic activity.

u-937 CR
While top0 I is quantitatively similar in U-937 and U-937/CR cells, the results demonstrate that this resistant phenotype is associated with expression of an altered enzyme. Top0 I activity in U-937/CR cells was approximately 10-fold less sensitive to 9-NC than that in parental cells. In the absence of quantitative
changes, this loss of 9-NC sensitivity probably accounts for the comparable increases in cellular resistance to this agent. Immunoblot analysis of U-937/CR extracts with an anti-top0 I serum revealed reactive proteins a t 100,75, and 67 kDa. While previous studies have demonstrated that the 100-and 67-kDa proteins are catalytically active (43, 49), less is known about the 75-kDa species (50). However, the finding that the pattern and intensity of the three bands was similar in both U-937 and Moreover, expression of a wild-type allele was undetectable in U-937/CR cells. As determined by Southern analyses, this finding was not related to genomic rearrangements (data not shown). Although we were able to detect differences in methylation of the top0 I gene in the resistant cells (data not shown), the basis for lack of expression of a wild-type allele is unclear. While these findings, taken together, supported the involvement of a mutation at codon 361 in cellular resistance to 9-NC, a more direct analysis was performed using partially purified human top0 I expressed in E. coli. The results of these experiments demonstrate that expression of top0 I containing the observed mutation at codon 361 is sufficient for conferring resistance to 9-NC. Comparison of amino acids 352 to 371 in human top0 I with corresponding sequences of this enzyme from other species has revealed a region that is highly conserved in eukaryotes (Table   111). Moreover, vaccinia virus top0 I, which is resistant to CPT, has a peptide sequence that differs in this region. Thus, one potential explanation for the present findings is that this highly conserved domain in eukaryotes is involved in the binding of 9-NC to the top0 I-DNA complex. A mutation at codon 361 could directly interfere with 9-NC binding. Alternatively, an alteration at this site could influence secondary structure and thereby binding of 9-NC. In this context, use of a computer algorithm to predict protein secondary structure has indicated that residues 364-370 of the wild-type protein are probably part of a loop structure (51). In contrast, a similar analysis of the mutant enzyme indicates that this loop region is lengthened to include residues 350-369. The present results would suggest that this alteration in secondary structure has little if any effect on catalytic activity and may contribute to CPT resistance. The presence of a mutation at phenylalanine 361 could also affect post-translational modification of the enzyme. Previous studies have demonstrated that top0 I is phosphorylated and activated by the serinekhreonine protein kinase C (52,53). Interestingly, substitution of phenylalanine with serine at residue 361 in the context of carboxyl-terminal arginine residues results in a potential protein kinase C phosphorylation site in the mutant enzyme (54). Since the phosphorylation state of top0 I has been hypothesized to be important in CPT-  which result in changes at each site from aspartic acid to glycine (31). Although glycine is found at position 583 in a CPTsensitive top0 I from human placenta (491, these regions represent potential sites for CPT interaction. Other studies with purified top0 I from a CPT-11-resistant human lung cancer cell line have demonstrated that a mutation (residue 729) near the active site of the enzyme is associated with decreased sensitivity to this agent in catalytic assays (30). Recent work has also implicated a mutation of top0 I Gly-503 to serine in the development of a resistant hamster cell line (55). Taken together with the results of the present studies, these findings suggest that multiple residues (361, 503, 533/583, 729) in top0 I are involved in the interaction of CPT with the top0 I.DNA complex and that CPT interacts with non-contiguous domains in the enzyme.