Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

DNA topoisomerase I (Top1p) catalyzes the relaxation of supercoiled DNA through the transient cleavage of one strand of a DNA duplex and is fundamental to processes such as replication, recombination, and transcription (Hsiang et al., 1985; Chen and Liu, 1994; Wang, 1996; Nitiss, 1998). Eukaryotic Top1p is the cellular target of the antitumor drug camptothecin (Cpt), which reversibly stabilizes the cleavable complex, an intermediate in the enzyme's catalytic cycle (Hsiang et al., 1985; Chen and Liu, 1994; Pommier et al., 1998). Genetic and biochemical studies have led to the identification of amino-acid substitutions in Top1p that render the enzyme resistant to Cpt. Mutations that yield catalytically active, yet drug-resistant enzymes are clustered in highly conserved regions, which are themselves widely separated along the linear amino-acid sequence of the protein. One region surrounds the active site tyrosine, where several such mutations have been identified (Knab et al., 1993, 1995; Fujimori et al., 1995; Wang, 1997). Another is located around residue 363 (Gly363) of the human enzyme, where two mutations have been independently mapped (Benedetti et al., 1993; Rubin et al., 1994). Collectively, these data led to the working hypothesis that the Cpt-binding site is formed by a cleft that contains DNA and residues responsible for Cpt resistance.

Recently, the three-dimensional structures of a 26-kDa fragment of yeast DNA Top1p (yTop1p) (Lue et al., 1995) and of reconstituted and N-terminal truncated versions of human DNA Top1p (hTop1p) in complexes with a 22-base pair DNA molecule (Redinbo et al., 1998; Stewart et al., 1998) have been solved. The enzyme is organized in multiple domains that "clamp" around the DNA molecule. DNA relaxation has been proposed to proceed via "controlled rotation" in which the covalently bound enzyme holds one end of the DNA duplex, and allows the end downstream of the cleavage site to rotate around the remaining phosphodiester bond (Redinbo et al., 1998). The residues mutated in the Cpt-resistant mutants are positioned in regions of the enzyme engaged in DNA-protein interactions and are clustered along one face of the DNA molecule. In hTop1p, Gly363 lies within a loop structure that constitutes one of two "lips" that interact with the DNA and each other in the closed DNA bound protein clamp (Redinbo et al., 1998). The corresponding structure of the 26-kDa yTop1p fragment, called the Cpt loop, also contains this residue (Gly295) (Lue et al., 1995). Lue et al. (1995) suggested that the Cpt loop was opposite the active site tyrosine domain of the protein and formed a catalytic cleft. Indeed, the close proximity of these domains is borne out in the cocrystals of the reconstituted central and C-terminal fragments of hTop1p with DNA (Redinbo et al., 1998).

We describe a novel genetic screen to identify additional mutations that suppress the Cpt sensitivity of hTop1p without abolishing catalytic activity. Among the mutants isolated were additional substitutions of Gly363. Substitution of Ser or Val for Gly363, in mutants htop1G363S and htop1G363V, respectively, abolished the Cpt sensitivity of these mutant proteins with only marginal effects on enzyme activity. This was in contrast to the more than 100-fold reduction in specific activity seen in the corresponding yeast top1 mutant Top1G295Vp (Hann et al., 1998b). To address this apparent discrepancy in the activities of identical amino-acid substituents in hTop1p and yTop1p, despite the conservation of amino-acid sequence and structure in this domain, we further investigated the effects of these Gly363 substitutions in suppressing the lethal phenotype of a Top1 mutant that mimics Cpt. Herein, we show that substitution of Thr718 to Ala in hTop1p mirrors the alterations in catalytic activity observed with the same substitution in yTop1T722Ap, both in vitro and in vivo (Megonigal et al., 1997; Hann et al., 1998a,b) Moreover, we show that the Cpt-resistant mutations at Gly363 have widely varied affects in suppressing the lethal phenotype and catalytic activity of the hTop1T718A mutant enzyme. Furthermore, molecular models were created for the mutants, based on the three-dimensional structure of the human enzyme. Along with biochemical analyses, these data suggest that distinct interactions between the two conserved lip domains mediate the Cpt sensitivity and catalytic activity of wild-type DNA Top1p. The implications of this intramolecular suppression are discussed in the context of current structural models for eukaryotic DNA Top1p.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials, Yeast Strains, and Plasmids. Cpt (Sigma Chemical Co., St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO) at 4 mg/ml. Saccharomyces cerevisiae strains EKY2 (MATa, ura3-52, his3200, leu21, trp163, top1::HIS3) and JN2-134 (MATa, rad52::LEU2, trp1, ade2-1, his7, ura3-52, ise1, top-1, leu2) were described previously (Bjornsti et al., 1989; Kauh and Bjornsti, 1995). Strain JCW28 (Mata, ura3-52, his3200, leu21, trp163, top2-4, top1) was kindly provided by Dr. J. C. Wang, Harvard University. YCpGAL1-hTOP1 and ptac-hTOP1 have been described (Bjornsti and Wang, 1987; Bjornsti et al., 1989). YEptopA-pGPD was reported previously (Giaever and Wang, 1988). pBlueAK3-1 DNA contains a high-affinity DNA Top1p cleavage site (Knab et al., 1995).

Cell Viability Assays Top1 yeast strains were transformed with the YCpGAL1-hTOP1 constructs by LiOAc treatment and selected on synthetic complete (SC)-uracil plus 2% dextrose. Transformants were grown to an A₅₉₅ = 0.5 and 5-µl aliquots of serial 10-fold dilutions were spotted onto SC-uracil plates plus 2% dextrose or 2% galactose. Alternatively, cultures were diluted 1:100 into SC-uracil media containing 2% raffinose and, at an A₅₉₅ = 0.3, induced with a final 2% galactose. At various time points, aliquots were serially diluted and plated onto SC-uracil, 2%dextrose. The number of colonies was counted following incubation at 30°C.

DNA Top1p Activity In Vivo. Yeast strain JCW28 (top1, top2-4ts) was cotransformed with plasmid YEptopA-pGPD, which constitutively expresses the bacterial topA gene (Giaever and Wang, 1988), and either YCpGAL1, YCpGAL1-hTOP1, or YCpGAL1-htop1T718A. Transformants were grown at 25°C, a permissive temperature for the top2-4ts mutant, in SC-uracil-leucine plus dextrose. At A₅₉₅ = 1.0, Top1p expression was induced with 2% galactose. After 5 h, half of the culture was shifted to 37°C for 3 h to inactivate DNA Top II. The cells were harvested and disrupted as described (Worland and Wang, 1989; Gartenberg and Wang, 1992) and the linking number distributions of purified plasmid DNAs were determined by two-dimensional gel electrophoresis (Benedetti et al., 1993; Megonigal et al., 1997). DNA was transferred to a nylon membrane (Amersham Corp.) and probed for 2-µm plasmid sequences with a ³²P-labeled fragment prepared by random priming (Pharmacia, Piscataway, NJ). Phosphoimager analysis was used to establish the ratio between positive supercoils and relaxed DNA in each sample.

Purification of Mutant hTop1p. Partial purification of DNA Top1ps from galactose-induced EKY2 cells was achieved by phosphocellulose column chromatography as described (Benedetti et al., 1993; Knab et al., 1995). Coomassie-stained SDS-polyacrylimide gel electrophoresis gels indicate the purity of these enzymes to be ~50 to 60%. Alternatively, a one-step purification was achieved with the FLAG system (Kodak) as follows: sequences encoding the epitope DYKDDDY were inserted at the 5' end of hTOP1 by polymerase chain reaction. The tagged heTOP1 gene was sublconed into YCpGAL1 to produce YCpGAL1-heTOP1. Western blot and immunfluorescence analyses verified that full-length heTop1p was recognized by the epitope-specific monoclonal antibody M2 (data not shown). To purify heTop1p, strain EKY2 was transformed with YCPGAL1-heTOP1 or YCpGAL1-hetop1T718A, grown in SC-uracil plus dextrose and diluted 1:100 in SC-uracil plus raffinose. At an A₅₉₅ = 1.0, the cells were induced with 2% galactose for 6 h. Cells were harvested by centrifugation, washed with TEEG (TRIS, EDTA, EGTA, glycerol) buffer (50 mM Tris, pH 7.4, 1 mM EDTA, 1 mM ethylene glycol bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 10% (v/v) glycerol, 100 µg/ml phenylmethylsulfonyl fluoride, 800 µg/ml NaHSO₃, 20 µg/ml benzamidine, 2 µg/ml pepstatin A, 2 µg/ml leupeptin) and resuspended in TEEG buffer at 2 ml buffer/g cells (wet weight). Then 0.5 volumes of 0.5-mm glass beads were added, and the cells were disrupted with ten 30-s cycles of vortexing alternating with 30 s on ice. The lysate was clarified by centrifugation and incubated for 1 h at 4°C with the M2 antibody covalently coupled to an insoluble resin (Kodak). The resin was washed three times for 10 min with 10 mM Tris, pH 7.4, 150 mM NaCl, 0.05% Tween 20, and used directly in activity or cleavage assays.

DNA Top1p Activity In Vitro. DNA Top1p activity was assayed in DNA relaxation reactions (Bjornsti and Wang, 1987; Bjornsti et al., 1989). Briefly, DNA Top1p preparations were incubated in 30-µl reaction volumes containing 0.5 µg of negatively supercoiled plasmid pHC624 DNA and reaction buffer [20 mM TRIS (pH 7.5), 0.1 mM Na₂EDTA, 10 mM MgCl₂, 50 µg/ml acetylated BSA, and 100 mM KCl]. After 1 h at 37°C, the reaction products were analyzed in 1% agarose gels. One unit of activity was the amount of enzyme needed to completely relax 0.5 µg of negatively supercoiled plasmid under standard reaction conditions. To assess the effects of Cpt on enzyme activity, 50 µM Cpt was added to the reactions and the extent of plasmid DNA relaxation at 0.5, 1, 5, 15, and 30 min was determined with a Bio-Rad GS 700 imaging densitometer with Multi Analyst software.

DNA Cleavage Assays. As described (Benedetti et al., 1993; Knab et al., 1995; Megonigal et al., 1997), cleavage of DNA by DNA Top1p was assayed by incubating enzyme preparations or immunotag-purified enzymes with a double-stranded DNA fragment, radiolabeled at one unique 3' end, in the presence or absence of camptothecin. A 900-base pair DNA fragment containing a high affinity DNA Top1p cleavage site (Bonven et al., 1985) was single ³²P-end labeled and purified from plasmid pBlueAK3-1 as described (Knab et al., 1995). Enzyme preparations were incubated in 50-µl reactions with DNA, reaction buffer and, where indicated, 100 µM camptothecin. DMSO was added to the no-drug controls. Following incubation at 37°C for 1, 5, or 30 min, reactions were terminated by the addition of 1% SDS, heated to 75°C for 15 min, and treated with 0.2 µg/ml proteinase K. The cleaved DNA fragments were resolved in 8 M urea/8% polyacrylamide gels and visualized by autoradiography (Benedetti et al., 1993; Knab et al., 1995)

Fluorescence Microscopy EKY2 cells, transformed with YCpGAL1-htop1T718A and induced with galactose for 6 h, were fixed by the addition of a final 3.7% formaldehyde. After 30 min, cells were pelleted and resuspended in 100 mM KPO₄, pH 7.0, and 3.7% formaldehyde. After 30 min, the fixed cells were washed three times with 100 mM KPO₄, sonicated, and stored at 4°C. To visualize nuclei, cells were resuspended in mounting medium (10 mM NaPO₄ pH 7, 150 mM NaCl, 90% glycerol, 1 µg/ml p-phenylenediamine) plus 1.0 µg/ml 4,6-diamidino-2-phenylindole and viewed in a Leitz Axioplan microscope equipped with a UV filter set. Large-budded cells (defined as cells in which the bud was at least one-half as large as the mother cell) were scored for the presence of a single nucleus, two nuclei, or one nucleus in the bud neck. One hundred cells were scored for each of five experiments, by an observer blinded to the identity of the samples.

Molecular Simulations. Models of enzyme mutants were constructed based on the coordinates of hTop1p (70 kDa) in noncovalent complex with a 22-base pair DNA duplex (Redinbo et al., 1998; Stewart et al., 1998). Hydrogens were added and the whole structure energy-minimized. Appropriate amino-acid substitutions were made and each structure was energy-minimized with only residues within a 20Å-radius from each mutation free to move. The programs Insight II, Biopolymer, and Discover (Molecular Simulations, Inc., Sunnyvale, CA) were used for viewing and creating the models. Minimizations were performed with the AMBER all-atom forcefield (Weiner, 1981; Weiner et al., 1984), a distance-dependent dielectric constant (4xR) and the conjugate gradients algorithm to final maximum derivative of 0.01 kcal/A.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Yeast Genetic Screen for Catalytically Active, Cpt-Resistant hTop1 Mutants. Previously, we examined DNA Top1 mutants in yeast (for review, see Bjornsti et al., 1994; Benedetti et al., 1998; Reid et al., 1998) and showed that cytotoxicity requires the presence of an active enzyme because top1 yeast lacking DNA Top1 are Cpt resistant. A large percentage of Cpt-resistant top1 yeast cells tranformed with a mutagenized pool of hTOP1 will express enzymes with reduced catalytic activity. To avoid brute-force screening of the subset of catalytically active, Cpt-resistant htop1 mutants, a genetic screen was devised that took advantage of the fact that yeast strains deficient in DNA repair (due to deletion of the RAD52 gene) tolerate only limited amounts of hTOP1. JN2-134 (rad52) cells tolerate pGAL1-promoted hTOP1 expression from a single copy vector, but exhibit decreased viability when the enzyme is expressed from pGAL1 on a multicopy vector (Bjornsti et al., 1989). Yet, leaky expression from this same multicopy YEpGAL1-hTOP1 vector is sufficient to render JN2-134 cells sensitive to 5 to 10 µg/ml Cpt on dextrose plates. Thus, catalytically active, Cpt-resistant htop1 mutants may be selected by replica plating Cpt-resistant colonies onto galactose plates and identifying the inviable cells (Fig. 1).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Isolation of Cpt resistant htop1 mutants. With the catalytically active Cptr mutant htop1G363C as an example, a yeast screen was developed to isolate catalytically active, Cpt-resistant hTop1p mutants. Exponentially growing cells in dextrose, transformed with YEpGAL1-hTOP1 or YEpGAL1-htop1G363C, were serially 10-fold diluted starting from A595 = 0.5 and 5 µl was spotted on selective media under repressed (Dex) or induced (Gal) conditions, in the presence or absence of 5 µg/ml Cpt. The figure shows that it is possible to score for resistance (YEpGAL1-htop1G363C in dextrose with Cpt) or sensitivity (YEpGAL1-hTOP1, under the same conditions).

hTop1G363Sp and hTop1G363Vp Are Cpt Resistant. htop1G363S and htop1G363V were cloned into the single copy vector YCpGAL1-hTOP1 to assess Cpt sensitivity in repair proficient EKY2 cells. As reported for htop1G363C, where Gly363 is mutated to Cys (Benedetti et al., 1993), cells expressing htop1G363S and htop1G363V were resistant to high doses of Cpt (5 µg/ml) (Fig. 2A).

View larger version (57K): [in this window] [in a new window]   Fig. 2.   htop1G363S and htop1G363V are resistant to Cpt. A, 5 µl of 10-fold serial dilutions starting from A595 0.5 of EKY2 (top1) cells transformed with YCpGAL1-hTOP1, YcpGAL1-htop1G363S, or YcpGAL1G363V were spotted onto SC-uracil plates containing 2% dextrose or 2% galactose in the presence or absence of 5 µM Cpt, and incubated at 30°C. B, to assess enzyme activity in DNA relaxation assays, partially purified hTop1p and Cpt-resistant mutant proteins were corrected for total protein concentration. Equal Top1p concentrations were confirmed in immunoblots. One microliter of serial 10-fold dilutions of hTop1G363Vp (lanes 1-4), hTop1G363Cp (lanes 5-8), hTop1G363Sp (lanes 9-12), and hTop1p (lanes13-16) were incubated with 0.5 µg of negatively supercoiled plasmid DNA (pHC624) as described in Experimental Procedures. C, no enzyme added.

hTop1T718A Is Lethal When Overexpressed in Yeast. The htop1T718A mutant, generated by site-directed mutagenesis, was cloned into plasmid YCpGAL1-htop1T718A and transformed into EKY2 (top1) cells. When cultured in galactose, cells expressing htop1T718A exhibited a dramatic reduction in cell viability (Fig. 3A). The cytotoxic activity of hTop1T718Ap required catalytic activity, as mutation of the active site Tyr723 to Phe abrogated htop1T718A-induced lethality. Moreover, low constitutive levels of htop1T718A expression from the yeast TOP1 promoter in YCpSc-htop1T718A was easily tolerated in EKY2 cells. In contrast, although JN2-134 cells transformed with YCpSc-hTOP1 were viable, no transformants were obtained with YCpSc-htop1T718A. Low levels of hTop1T718Ap were cytotoxic in rad52, repair deficient strains (data not shown). Thus, hTop1T718Ap-induced lethality was dependent upon cellular levels of mutant protein and the resultant DNA damage.

View larger version (46K): [in this window] [in a new window]   Fig. 3.   Overexpression of htop1-T718A is lethal in yeast. A, 5 µl of 10-fold serial dilutions starting from A595 0.5 of EKY2 (top1) cells transformed with YCpGAL1-hTOP1, YCpGAL1-htop1-T718A, or YCpGAL1-htop1T718A, Y723F were spotted onto SC-uracil plates containing 2% dextrose or 2% galactose, and incubated at 30°C. A comparison of human and yeast sequences around the active site tyrosine is shown. B, partially purified hTop1p (lanes 1-4) and hTop1 T718Ap (lanes 5-8) were serially 10-fold diluted in reaction buffer and 1 µl was assayed for their ability to relax 0.5 µg of supercoiled plasmid DNA as described in Experimental Procedures.

hTop1T718Ap Is Catalytically Active In Vitro and In Vivo. As shown in Fig. 3B, the specific activity of hTop1T718Ap was indistinguishable from that of wild-type hTop1p. The in vivo activity of hTop1p and hTop1T718Ap was assessed by examining the relaxation of the positive supercoils generated by transcription of the endogenous yeast 2-µm plasmid as described in Benedetti et al. (1993), Knab et al. (1993), and Megonigal et al. (1997). JCW28 (top1, top2-4ts) cells, deleted for TOP1 and bearing a temperature-sensitive mutation in DNA topoisomerase II, were cotransformed with YEptopA-pGPD and either YCpGAL1, YCpGAL1-hTOP1, or YCpGAL1-htop1T718A. YEptopA-pGPD constitutively expresses bacterial DNA Top1p. At 37°C, DNA topoisomerase II is inactivated and the negative supercoils that accumulate during transcription are selectively relaxed by bacterial DNA Top1p (Giaever and Wang, 1988). As seen in Fig. 4A, JCW28 cells transformed with YEptopA-pGPD and a control plasmid expressing no DNA Top1p accumulate positively supercoiled (PS) 2-µm plasmid DNA upon shift to the nonpermissive temperature (37°C). When JCW28 was transformed with YCpGAL1-htop1T718A or YCpGAL1-hTOP1 and shifted to nonpermissive temperature, the mutant or wild-type enzyme catalyzed the relaxation of the positive supercoils (Fig. 4, B and C). Thus, hTop1T718Ap is catalytically active in vivo.

View larger version (61K): [in this window] [in a new window]   Fig. 4.   hTop1T718A is active in vivo. Relaxation of positively supercoiled yeast 2-µm plasmid DNA by hTopo1p and htop1T718Ap were assayed in vivo. Strain JCW28 (top1, top2-4) constitutively expressing a plasmid-borne Escherichia coli topA gene encoding bacterial DNA topoisomerase I, was transformed with a control plasmid YCpGAL1 (A), YCpGAL1-htop1T718A (B), or YCpGAL1-hTOP1 (C). Cultures were incubated at 25°C, a permissive temperature for JCW28 strain, and induced with galactose. Cultures were then shifted to 37°C to inactivate yeast DNA topoisomerase II. The 2-µm plasmid DNA was recovered from each culture after the temperature shift and analyzed in southern blots of two-dimensional agarose gels electrophoresis. PS, positive supercoils.

Substitution of Ala for Thr718 in hTop1p Enhances Stability of Covalent Enzyme-DNA Intermediate. To assess the phenotypic consequences of htop1T718A expression, the stability of the covalent complex formed by the mutant enzyme was assessed. Equal concentrations of heTop1 and heTop1T718A proteins, epitope tagged at the N terminus, were bound to beads via the epitope specific M2 monoclonal antibody as described in Hann et al. (1998). The inclusion of the epitope had no effect on mutant or wild-type enzyme activity, either in solution or in the bead bound form (data not shown). The tethered proteins were incubated with a 3' end-labeled DNA fragment, in the presence or absence of 100 µM Cpt. Following incubation at 37°C for 1, 5, or 30 min, the cleavable complexes were trapped with SDS/proteinase K treatment and the cleaved DNA fragments resolved in a denaturing polyacrylamide gel. When wild-type heTop1p is incubated without Cpt, very little cleavage of the labeled DNA strand was detected (Fig. 5). However, when heTop1p is exposed to Cpt, a dramatic increase in cleaved DNA fragments was observed. In contrast, heTop1T718Ap-DNA complexes were stabilized even in the absence of Cpt, as evidenced by the increased intensity of cleaved fragments relative to heTop1p (Fig. 5). Yet, the apparent differences in band intensities produced in the presence and absence of Cpt (compare heTop1T718Ap lanes with Cpt and DMSO) indicate that heTop1T718Ap remains sensitive to Cpt. Moreover, because the pattern of bands was indistinguishable between heTop1T718Ap and heTop1p in the presence of Cpt, the preferred sites of cleavage by heTop1T718Ap are altered by the drug (see arrow in Fig. 5, A and C). Band (B) corresponds to the preferred cleavage site derived from rDNA (Bonven et al., 1985).

View larger version (90K): [in this window] [in a new window]   Fig. 5.   hTop1T718Ap increases the stability of the covalent enzyme-DNA intermediate in the absence of Cpt. In vitro DNA cleavage activity of bead-bound epitope-tagged ehTop1p and ehTop1T718Ap was assayed. A 3' 32P end-labeled DNA fragment was incubated with immunoprecipitated ehTop1p or ehTop1T718Ap for the times indicated, in the presence of 4% DMSO or 100 µM Cpt. Control, no enzyme added (C). Total protein (110 ng), with comparable specific activity, was used in each assay. Arrows A and C indicate cleavage sites mostly evident in the hTop1T718Ap in the absence of Cpt and arrow B indicates the preferred cleavage site derived from Tetrahymena rDNA.

top1

View larger version (66K): [in this window] [in a new window]   Fig. 6.   Nuclear segregation is defective in strains overexpressing htop1T718A. EKY2 (top1) cells were transformed with YCpGAL1-hTOP1 or YcpGAL1-htop1-T718A, and hTop1p expression was galactose-induced for 6 h before cells were processed for microscopy. The table shows the percentage of large-budded cells displaying a single or double nucleus. Scoring was done by an observer blinded to the identity of the samples.

Gly363 Substitutions Are Distinguished in Their Ability to Suppress Cytotoxic Activity of hTop1T718A. As detailed in Hann et al. (1998), random mutagenesis identified a Gly295 to Val substitution as an intragenic suppressor of ytop1T722A-induced lethality. This resulted from a dramatic decrease in DNA binding and, therefore, the catalytic activity of the double mutant. In contrast, our results demonstrate that the same substitution (Gly363 to Val) in hTop1p had only a minimal effect on enzyme activity, yet suppressed the Cpt sensitivity of the human enzyme. To investigate whether the interactions between the "lip" domain and the active-site tyrosine residues differ in the yeast and human enzymes, the ability of all three Gly363 substituents to suppress htop1T718A-induced lethality was assessed.

View larger version (58K): [in this window] [in a new window]   Fig. 7.   Mutations at position 363 can suppress the lethality of htop1T718A. A, viability of EKY2 (top1) cells transformed with the indicated YCpGAL1-hTOP1 construct. Five microliters of serial 10-fold dilutions of cultures starting from A595 = 0.5, were spotted onto SC-uracil plates containing 2% dextrose or 2% galactose and 5 µg/ml Cpt. B, 1 µl of serial 5-fold dilutions of partially purified wild-type or mutant hTop1p were assayed for their ability to relax 0.5-µg supercoiled plasmid DNA. Lanes 1-5, hTop1; lanes 6-10, hTop1G363C, T718Ap; lanes 11-15, hTop1G363V, T718Ap; lanes 16-20, hTop1G363S, T718Ap. Control (C), no protein added. Equivalent protein concentrations were verified by western blotting with an antibody specific for hTop1p. C, 1 µl of partially purified wild-type hTop1 (1-5) hTop1G363C, T718Ap (6-10) hTop1G363V, T718Ap (11-15), or hTop1G363S, T718Ap (16-20) were assayed for their ability to relax 0.5-µg supercoiled plasmid DNA in the presence of the KCl concentrations.

top1, rad52

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Time course of Cpt inhibition of hTop1p, hTop1T718Ap, hTop1G363Vp, and hTop1G363V, T718Ap relaxation activity. A, 1 µl of partially purified hTop1p or hTop1T718Ap was used in each assay in the presence or absence of 50 µM Cpt. At the following times, the reactions were terminated with 0.5% SDS: 0.5 (lanes 1, 6, 11, 16), 1 (lanes 2, 7, 12, 17), 5 (lanes 3, 8, 13, 18), 15 (lanes 4, 9, 14, 19), or 30 min (lanes 5, 10, 15, 20). Reaction products were resolved in agarose gels and visualized by staining with ethidium bromide. The percentage of relaxed DNA, quantitated with the Bio-Rad imaging densitometer and Multi Analyst software, is presented in graphic form. , hTOP1p (50 µM CPT); , hTOP1p (no drug); , htop1T718Ap (50 µM CPT); , htop1T718Ap (no drug). B, as described in (A), 1 µl of partially purified hTop1G363Vp or hTop1G363V, T718Ap was assayed in the presence or absence of 50 µM Cpt. Reactions were terminated with 0.5% SDS at 0.5 (lanes 1, 6, 11, 16) 1 (lanes 2, 7, 12, 17), 5 (lanes 3, 8, 13, 18), 15 (lanes 4, 9, 14, 19), or 30 min (lanes 5, 10, 15, 20). The percentage of relaxed DNA is presented in graphic form. , hTop1G363Vp (50 µM CPT); , hTop1G363Vp (no drug); , hTop1G363Vp (50 µM CPT); , hTop1G363Vp (no drug).

Molecular Modeling Analysis of Mutations. Molecular simulations were based on the coordinates of the three-dimensional structure of hTop1p (70 kDa) in noncovalent complex with a 22-base pair DNA duplex (Stewart et al., 1998). The complete structure was energy-minimized, after addition of the hydrogens, with the conjugate gradients algorithm and AMBER forcefield. Individual substitutions, T718A, G363C, G363S, and G363V, and double combinations were introduced and the structures were again minimized for all residues within a 20-Å radius from each mutation.

View larger version (27K): [in this window] [in a new window]   Fig. 9.   A, minimized structures in the region of the Thr718 and the active site tyrosine. Superimposition of wild-type hTop1p (red) and hTop1T718Ap (Ala substituted for Thr-718) (white) structures after minimization. The yellow DNA ribbon indicates the nonscissile strand. B, minimized structures in the region of the Cpt-loop. Superimposition of the wild-type hTop1p (red) and Val substituted for Gly-363 (green).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Genetic, biochemical, and crystallographic studies contribute to our current understanding of eukaryotic DNA Top1p structure and mechanism (Lue et al., 1995; Wang, 1996; Berger, 1998; Pommier et al., 1998; Redinbo et al., 1998; Stewart et al., 1998). Recent crystallographic data (Redinbo et al., 1998; Stewart et al., 1998) reveal that the enzyme forms a protein clamp around the DNA with basic residues in contact with the phosphate backbone and the active site tyrosine poised to attack the phosphodiester linkage. Stewart et al. (1998) proposed that the enzyme catalyzes the relaxation of supercoiled DNA via a mechanism of "controlled rotation" in which the protein clamp binds the DNA helix to create the single-stranded break. The DNA molecule 3' to the cleavage site rotates about the phosphodiester bond opposite the nick, with the positively charged surfaces of the "linker region" (a coiled-coil structure that connects the central globular core of the enzyme with the active site-tyrosine domain) and a set of cone helices extending from the top of the clamp, acting as a brake to slow the rotation of the negatively charged DNA. Specific contacts with the DNA involve several Top1p domains. The residues of the Cpt loop/lip region of core subdomain I also are involved in interdomainal interactions with the lip region of core subdomain III and in directly binding the DNA helix to form the closed clamp structure.

Among the Cpt-resistant mutants isolated in this screen, substitution of Ala653 with Pro is interesting given its location in the coiled-coil region of the linker domain. This mutant has reduced catalytic activity (data not shown) and is particularly sensitive to proteolysis. Substitution of Pro for Ala could disrupt the alpha helix of the coiled-coil, thereby affecting the interaction of the enzyme with DNA. Although the linker domain is dispensable for relaxation activity (Stewart et al., 1997), the presence of a modified structure could alter the controlled rotation of DNA and influence the relaxation activity of the enzyme.

Cpt resistance was previously ascribed to mutations in the Cpt loop/lip region of core subdomain I (Benedetti et al., 1993; Rubin et al., 1994; Lue et al., 1995; Li et al., 1997). Herein, we report that substitution of Val or Ser for Gly363 had little effect on catalytic activity, yet decreased the Cpt sensitivity of the enzyme to levels similar to those observed with the Cpt-resistant htop1G363C mutant. Subtle differences highlight the strict structural context required for Cpt sensitivity and presumably drug binding. Our molecular models suggest that substitutions in this lip region could affect enzyme activity and sensitivity to Cpt by altering the conformation of the loop. Arg364, which makes direct contact with the minor groove of the DNA opposite the site cleaved by the active site tyrosine (Redinbo et al., 1998) is slightly distorted in the minimized structures, with the G363C mutation being most pronounced. This distortion may influence the hydrogen bonding with Cpt as proposed by Redinbo et al. (1998) or might affect the overall structure of the loop important for Cpt binding. In addition, as a single salt bridge exists between the two lip domains, any alteration in the correct positioning of these two regions might adversely affect the interdomainal interactions needed to effect clamp closure. The opening and closing of the enzyme clamp around the DNA also could influence the Cpt binding-equilibrium. If hTop1p mutations alter the efficiency of clamp closure, the equilibrium would shift toward the unbound state.

Fan et al. (1998) recently proposed an alternate model for Cpt binding, based on molecular-modeling studies of different Cpt analogs and DNA. Substitution of Ala for Thr718 produces different patterns of DNA cleavage in the presence or absence of Cpt. In the presence of Cpt, however, the pattern was indistinguishable between hTop1T718Ap and hTop1p. Similar results were obtained with yTop1T722Ap (Megonigal et al., 1997), suggesting that Cpt binding to the mutant enzyme-DNA complex restores the microenvironment around the active site. This is in agreement with the Fan et al. (1998) model.

Mutation of Gly363 to Val provided the strongest suppression of htop1T718A lethality. The molecular models suggest the introduction of the branched side chain of Val, as opposed to the smaller, linear side chains of Cys or Ser, produces a more severe alteration in the Cpt loop. This was reflected in a pronounced decrease in catalytic activity, but only in the context of the Thr718-to-Ala substitution. Thus, the flexibility imparted to the Cpt loop by Gly363 may be more critical in maintaining the altered catalytic activity of hTop1T718Ap than wild-type hTop1p. Indeed, substitution of Val, Ser, or Cys for Gly363 had comparable effects on wild-type enzyme activity and Cpt sensitivity: a modest decrease in activity at higher salt and resistance to Cpt.

The results of Gly363 substitutions in hTop1p differed from that obtained following the substitution of Val for the corresponding residue (Gly295) in yTop1p, where a 100-fold decrease in yTop1G295Vp was attributed to a reduction in DNA binding (Hann et al., 1998b). Yet, consistent with the results presented herein, this effect was exacerbated in the context of the lethal Thr to Ala substitution. Thus, interactions of the Cpt loop/lip region with residues near the active site tyrosine affect the catalytic activity of both yTop1p and hTop1p.

Alterations in the catalytic activity of hTop1p imparted by the Thr718 to Ala mutation mirror those reported for yTop1T722Ap (Megonigal et al., 1997). The DNA relaxation activity of hTop1T718Ap was essentially identical with that of the wild-type enzyme. Furthermore, hTop1T718Ap mimics the cytotoxic effects of Cpt. Wild-type and mutant enzyme sensitivity to Cpt are comparable with regards to the degree and site specificity of DNA cleavage. Yet, in the absence of Cpt, hTop1T718Ap formed elevated levels of cleavable complexes. Therefore, either treatment with Cpt or mutation of Thr718 to Ala enhances the stability of the covalent enzyme-DNA intermediate. It is possible that different steps in the catalytic cycle are affected, which would not be distinguished by the assays used. Indeed, differences in the DNA cleavage patterns created by hTop1T718Ap in the presence or absence of Cpt support the hypothesis that the Thr718-to-Ala mutation and Cpt treatment interfere with distinct aspects of the catalytic cycle.

Yeast and hTop1p share a high degree of similarity in terms of sequence, structure, and enzyme mechanism. Moreover, the human enzyme functions in yeast (Bjornsti et al., 1989) and the yeast enzyme functions in mammalian cells (Hann et al., 1998). Yet, despite the overwhelming similarities in enzyme structure and function, significant differences have been reported (Merino et al., 1993; Knab et al., 1995; Reid et al., 1997). For example, yTop1p could not substitute for hTop1p in suppressing the basal level of transcription in vitro. Because hTop1p catalytic activity was not required, this implied a critical difference in structure between enzymes. In these studies, mutation of most conserved residues elicited similar alterations in enzyme activity, yet significant differences were apparent. As numerous Cpt analogs and other Top1p inhibitors are being developed as chemotherapeutic agents, deciphering the structural features of Top1p critical for Cpt sensitivity is important.