Induction of Topoisomerase I Cleavage Complexes by the Vinyl Chloride Adduct 1, N 6 -Ethenoadenine

We used purified mammalian topoisomerases I (top1) and oligonucleotides to study top1-mediated cleavage and religation in the presence of a potent carcinogenic adduct, 1, N 6 -ethenoadenosine ( e A) incorporated immediately downstream of a unique top1 cleavage site. We found tha e A markedly enhanced top1 cleavage complexes when it was incorporated at the 1 1 position of the top1 cleavage. This enhancement was due to a reduction of the religation step of the top1 reaction. In addition, e A reduced the top1-mediated cleavage and decreased binding of the enzyme to DNA. We also studied the effects of the e A adduct on top1 trapping by camptothecin (CPT), a well known top1 inhibitor. CPT was inactive when e A was present at the 1 1 position. Alkylation of the top1 cleavage complex by 7-chloro-methyl-10,11-methylenedioxycamptothecin(7-ClMe-MDO-CPT)wasalsoblockedbythe e A adduct. Altogether, these results demonstrate that the e A carcinogenic adduct can efficiently trap human top1 and mimic CPT effects. Normal hydrogen bonding of the base pairs immediately downstream from the top1 cleavage site is probably essential for efficient DNA religation and binding of camptothecins in the top1 cleavage complex.


Induction of Topoisomerase I Cleavage Complexes by the Vinyl Chloride Adduct 1, N 6 -Ethenoadenine
Induction of Topoisomerase I Cleavage Complexes by the Vinyl Chloride Adduct 1,N 6 -Ethenoadenine* (Received for publication, May 29, 1998, and in revised form, July 22, 1998)  We used purified mammalian topoisomerases I (top1) and oligonucleotides to study top1-mediated cleavage and religation in the presence of a potent carcinogenic adduct, 1,N 6 -ethenoadenosine (⑀A) incorporated immediately downstream of a unique top1 cleavage site. We found tha ⑀A markedly enhanced top1 cleavage complexes when it was incorporated at the ؉1 position of the top1 cleavage. This enhancement was due to a reduction of the religation step of the top1 reaction. In addition, ⑀A reduced the top1-mediated cleavage and decreased binding of the enzyme to DNA. We also studied the effects of the ⑀A adduct on top1 trapping by camptothecin (CPT), a well known top1 inhibitor. CPT was inactive when ⑀A was present at the ؉1 position. Alkylation of the top1 cleavage complex by 7-chloromethyl-10,11-methylenedioxycamptothecin(7-ClMe-MDO-CPT) was also blocked by the ⑀A adduct. Altogether, these results demonstrate that the ⑀A carcinogenic adduct can efficiently trap human top1 and mimic CPT effects. Normal hydrogen bonding of the base pairs immediately downstream from the top1 cleavage site is probably essential for efficient DNA religation and binding of camptothecins in the top1 cleavage complex.
Vinyl chloride is a potent carcinogen. Its metabolites are reactive species that can induce cyclic adducts in DNA. Four adducts have been characterized: the 1,N 2 -ethenodeoxyguanosine (⑀G), 1 the N 2 ,3-ethenodeoxyguanosine, the 3,N 4ethenodeoxycytidine (⑀C), and the 1,N 6 -ethenodeoxyadenosine (⑀A) (1). These adducts can also be generated by other exogenous sources such as vinyl carbamate, ethyl carbamate, and other P450-mediated oxidized two-carbon compounds (2,3). More recently, ⑀A and ⑀C have been detected in normal hepatocytes, suggesting their endogenous formation (4), probably through multiple reactive intermediates such as lipid peroxidation products (5). In vivo cytotoxicity and carcinogenic effects of etheno adducts have been linked to their mutagenic properties (6 -9). For instance, ⑀G causes G/A transitions and ⑀dA causes A/C transversions in vivo (7,9). When introduced into DNA, ⑀A and ⑀C are recognized and removed by human cell extracts at similar rates (10). Two separate DNA glycosylase activities have been reported for ⑀A and ⑀C excision (11), but the corresponding proteins have not been identified yet.
Topoisomerase I (top1) is important for DNA replication and transcription, because it can remove the torsional constraint along the DNA (12)(13)(14). Its role in DNA repair is less well established (15,16). Mammalian top1 cleaves one strand of the DNA and remains covalently attached to the 3Ј-end of the transient break. This catalytic intermediate is referred to as the top1 cleavage complex. This reaction is reversible, and in normal conditions, only a small fraction of the cleaved DNA is detectable. Top1 inhibitors (poisons) such as camptothecin (CPT) increase (trap) top1 cleavage complexes by inhibiting the religation step (17,18).
Both topoisomerase I and II can be trapped by DNA modifications such as mismatches, abasic sites, nicks, and gaps, when such lesions are in the vicinity of the enzyme cleavage site (19 -22). In the present study, we used oligonucleotides containing an ⑀A at the ϩ1 position relative to the cleavage site and analyzed the effects on the top1-mediated DNA cleavage and religation and on top1 binding to DNA. We show for the first time that the carcinogenic adduct, ⑀A, can trap top1 and mimic CPT effects. We also demonstrate that the ⑀A adduct decreases top1 binding to DNA and the "on-rate" DNA cleavage. Finally, CPT activity was strongly inhibited when ⑀A was in the immediate vicinity of a top1 cleavage site.
Oligonucleotides Labeling and Annealing Procedures-3Ј labeling was performed using terminal deoxynucleotidyl transferase (Stratagene, La Jolla, CA) with ␣-32 P-cordycepin as described previously (23). Labeling mixtures were subsequently centrifuged through a G-25 Sephadex column to remove excess of unincorporated nucleotide. Radiolabeled single-stranded DNA oligonucleotides were annealed to the same concentration of unlabeled complementary strand in 1 ϫ annealing buffer (10 mM Tris-HCl, pH 7.8, 100 mM NaCl, 1 mM EDTA). Annealing mixtures were heated to 95°C for 5 min and slowly chilled overnight to room temperature.
Top1 Purification-Overexpression of the human recombinant top1 was performed in Sf9 insect cells using a baculovirus virus construct containing the top1 full-length cDNA (24). Nuclear extracts were from infected Sf9 cells 4 days after infection as described previously (20) and kept at Ϫ80°C before use. Human top1 was purified from nuclear extract using nickel-nitrilotriacetic acid-agarose beads (Qiagen, Santa Clarita, CA). 50 l of beads were placed into microcentrifuge tubes and washed three times with 1 ml of wash buffer (50 mM Hepes, pH 7, 0.5 mM dithiothreitol, 10 mM MgCl 2 , 3 mM MnCl 2 , 50 mM KCl). 500 l of nuclear extract were incubated with beads for 30 min at 4°C and * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
washed two times with 300 l of the wash buffer. Insect cells top1 was eluted with wash buffer containing 40 mM imidazole. Human recombinant top1 was eluted with wash buffer containing 300 mM imidazole and dialyzed 3 h with 30 mM potassium phosphate buffer (pH 7), 0.1 mM EDTA, 5 mM dithiothreitol. Bovine serum albumin (0.2 mg/ml) and glycerol (30% v/v) were added to the purified top1, and aliquots were stored at Ϫ20°C.
Top1 Reactions-DNA substrates (approximately 50 fmol/reaction) were incubated with 5 units of calf thymus or with human recombinant top1 usually for 30 min at 25°C with or without drug in standard reaction buffer (10 mM Tris-HCl, pH 7.5, 50 mM KCl, 5 mM MgCl 2 , 0.1 mM EDTA, 15 g/ml bovine serum albumin). Reactions were stopped by adding either sodium dodecyl sulfate (SDS) (final concentration 0.5%) or NaCl (0.5 M for 30 min at 25°C followed by addition of 0.5% SDS for standard reactions or 0.35 M NaCl at 4°C for kinetics of reversal. Gel Electrophoresis and Analysis of Cleavage Products-3.3 volumes of Maxam Gilbert loading buffer (98% formamide, 0.01 M EDTA, 1 mg/ml xylene cyanol, and 1 mg/ml bromphenol blue) were added to the reaction mixtures before loading. 16% denaturing polyacrylamide gels (7 M urea) were run at 40 V/cm at 50°C for 2-3 h and dried on 3MM Whatman paper sheets. Imaging and quantitations were performed using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Top1 Is Reversibly Trapped by ⑀A at the ϩ1 Position of a Top1
Cleavage Site-In this study we have investigated the effect of ⑀A on top1 activity in vitro. As shown in Fig. 1A, the ⑀A bulky adduct differs from adenine by the presence of an additional ring resulting from a cyclization reaction of reactive species such as the vinyl chloride metabolite, chloroacetaldehyde (26). We used purified oligonucleotides derived from the Tetrahymena ribosomal DNA sequence (27) and analyzed the effects of ⑀A incorporation at the ϩ1 position of the scissile strand (X in Fig. 1B) on top1-mediated DNA cleavage and religation using purified recombinant human enzyme. As shown in Fig. 1C, incorporation of ⑀A led to approximately 6-fold increase of top1-mediated cleavage complexes (compare lanes 2). Top1 cleavage/religation equilibrium can be shifted toward religation by increasing the salt concentration in the reaction (28). In the case of the ⑀A oligonucleotide, 0.5 M NaCl was able to reverse the top1-mediated cleavage complexes (Fig. 1C, left  panel, lane 3).
Previous studies indicated that CPT interacted with the ϩ1 base immediately 3Ј to the top1 cleavage site (29) and increased the top1 cleavage complex (compare lanes 2 and 4 in the control panel) by inhibiting the religation step of the reaction (12,17,30). In the case of the ⑀A oligonucleotide, CPT did not enhance top1-mediated DNA cleavage further (Fig. 1C, compare lanes 2 and 4 on the right panel), and the reaction was still reversible in the presence of 0.5 M NaCl (lane 5, right panel).
These results demonstrate that the carcinogenic adduct ⑀A incorporated at the ϩ1 position of a top1 cleavage site reversibly traps top1.
Kinetics of Top1 Trapping by ⑀A-We next compared the kinetics of top1 cleavage complex formation between the con- trol oligonucleotide (A:T) and the ⑀A-containing oligonucleotide (⑀A:T) to investigate whether reversible trapping could be further enhanced by increasing the reaction time. As shown in Fig.  2A, cleavage complexes were formed within the first minute of incubation for both the normal and ⑀A oligonucleotides, and no further increase was detected. Instead, a slight decrease of the cleavage complexes was detected with longer incubation, probably resulting from a decrease of the enzyme activity. In accordance with the results presented in Fig. 1C, the plateau value was 5-6-fold higher when ⑀A was present at the ϩ1 position (Fig. 2B). These results demonstrate that the reversible trapping of top1 by ⑀A probably corresponds to an equilibrium that is reached quickly once top1 is added to the reaction.
⑀A at the ϩ1 Position Inhibits the Formation of Top1-mediated DNA Cleavage-Because top1-mediated DNA cleavage/ religation is an equilibrium process, reversible trapping of the enzyme by ⑀A could possibly result from either an increase of the cleavage (on-rate) or/and a decrease of the religation ("offrate"). In the following experiments, we separately investigated the effects of ⑀A on each step of the top1 reaction. We first studied the DNA cleavage formation in "suicide" oligonucleotides containing either A or ⑀A at the ϩ1 position relative to the top1 cleavage site (see Fig. 3A). Under these conditions, top1mediated cleavage was examined under single turnover conditions (25,30). Once top1 cleaves the DNA, it remains bound to the 3Ј-end of the scissile strand in a covalent complex (suicide complex), and the 3Ј-labeled 6-mer distal product is released (Fig. 3B, arrowhead). In our reactions, the 5Ј-end of the nonscissile strand was phosphorylated prior to annealing to avoid formation of recombination products (20,23). The kinetics experiment presented in Fig. 3B show a marked difference in the on-rate reaction between the two suicide substrates. Cleavage was slower in the oligonucleotide containing an ⑀A at the ϩ1 position than in the control (Fig. 3C). The apparent cleavage rate constant (k cl ) was calculated. Cleavage was normalized to the end point value (lanes 8 in Fig. 3B) of 72 and 78% for the A:T (open squares) and ⑀A:T (closed squares) oligonucleotide, respectively, and k cl was determined by fitting the data to the equation (100 Ϫ %Cl norm ) ϭ 100 e Ϫkcl⅐t (31). The k cl for the ⑀A-containing substrate was approximately 3-fold less than the control (0.047 s Ϫ1 versus 0.15 s Ϫ1 ). These results indicate that the top1-mediated DNA cleavage rate (on-rate step) is reduced by the presence of ⑀A at the ϩ1 position of a top1 cleavage site.
⑀A at the ϩ1 Position Inhibits Noncovalent Top1 Binding to DNA-To further determine whether presence of ⑀A altered the top1-DNA interaction, DNA binding assays were performed. For this purpose, we used the catalytically inactive yeast top1 in which the active tyrosine in position 727 has been mutated to a phenylalanine (Y727F). This enzyme does not perform cleavage, but is still able to bind to the DNA (32,33). We incubated 3Ј-labeled oligonucleotides containing either A (A:T) or ⑀A (⑀A:T) at the ϩ1 position in the presence of increasing amount of the yeast mutant top1 (Fig. 4). Y727F top1 was able to bind to both substrates, which resulted in the appearance of a slow migrating band corresponding to top1⅐DNA complexes in  Fig. 1B) were incubated with increasing amount of purified Y727F yeast mutant top1 for 5 min at 25°C. Lanes 1-4 correspond to 0, 0.75, 1.5, and 3.0 ng of the yeast mutant top1, respectively. Equal counts/min were loaded on a 6% native polyacrylamide gel in 0.25 ϫ TBE, and top1⅐DNA complexes were separated from free DNA at 100 V for 2 h. B, top1⅐DNA complexes were quantified from the dried gel shown in A using a PhosphorImager. For each concentration of enzyme, top1 binding to ⑀A:T was calculated relative to the A:T (control) oligonucleotide (binding of top1 to the control oligonucleotide was normalized to 1). native polyacrylamide gels (Fig. 4A). Top1 binding to the ⑀Acontaining substrate was expressed relative to the binding to the control oligonucleotide (which was set to 1), and the ratio was determined for each concentration of enzyme (Fig. 4B). Top1 binding to the oligonculeotide with ⑀A at the ϩ1 position was reduced by approximately 4 -5-fold compared with the control DNA. These results demonstrate that top1 binding is decreased by the presence of ⑀A at the ϩ1 position of the top1 site. They also suggest that reduction of the top1-mediated DNA cleavage on-rate (Fig. 3) is probably due to reduced binding of the enzyme to DNA.
⑀A Incorporation Mimics CPT Effects-Because the above data indicated that reduced top1 cleavage could not be responsible for top1 trapping by ⑀A, we studied the influence of ⑀A on the religation step (Fig. 5). For this purpose, we first incubated the normal and ⑀A oligonucleotides (see Fig. 1B) with purified top1 for 15 min at 25°C to reach cleavage equilibrium (see Fig.  2). Religation was then studied after addition of 0.35 M NaCl. These religation experiments were performed at 4°C to better follow religation kinetics (Fig. 5A). Under these conditions, religation is slower than at 37°C (28). Religation was markedly decreased with the ⑀A:T substrate as compared with the control oligonucleotide (Fig. 5A). As a comparison, we used the control oligonucleotide in the presence of 10 M CPT. CPT is known to increase top1 cleavage complexes by inhibiting religation (17,18). Cleavage products were quantitated and plotted as a function of time after salt addition (Fig. 5B). The kinetics of religation for the ⑀A-containing substrate and the control oligonucleotide in the presence of CPT were comparable (Fig. 5B,  compare open circles and closed box). These results demon-strate that reversible trapping of top1 by ⑀A is due to an inhibition of the religation step of the top1 reaction.
Lack of Effect of CPT in the Oligonucleotide with ⑀A at the ϩ1 Position-As shown in Fig. 1C, CPT did not increase further top1-mediated DNA cleavage when ⑀A was present at the ϩ1 position of the top1 site. We investigated this further by using different CPT concentrations (Fig. 6). Whereas CPT was active even at low concentration (0.1 M) in the control oligonucleotide (A:T), it did not increase top1-mediated cleavage even at high concentration in the ⑀A oligonucleotide (Fig. 6A, compare lanes  2 and 4 on the right panel). These results demonstrated that presence of ⑀A at the ϩ1 position relative to the top1 cleavage site prevents CPT activity.
⑀A at the ϩ1 Position Prevents Alkylation of Guanine N3 by 7-ClMe-MDO-CPT-Because CPT has been reported to interact directly with the ϩ1 base from the top1 cleavage site (29), we tested whether the lack of effect of CPT on top1 cleavage was associated with alterations of the drug with the ⑀A. For this purpose, we used the alkylating CPT, 7-ClMe-MDO-CPT. This drug alkylates the N3 position of the ϩ1 guanine or adenine in the presence of top1 (29,34). As reported previously, in the control oligonucleotide, 7-ClMe-MDO-CPT stimulated top1 cleavage complexes (Fig. 7, compare lane b with other lanes in the control panel) and formed the alkylated product that can be detected as a retarded band in denaturing polyacrylamide gel (Fig. 7, arrow). By contrast, when ⑀A was incorporated at the ϩ1 position, no increase of DNA cleavage was detectable with 7-ClMe-MDO-CPT. Moreover, the alkylation product was not detectable regardless of the concentration of 7-ClMe-MDO-CPT used. These results suggested that the lack of CPT activity when ⑀A was present at the ϩ1 position relative to a top1 cleavage site might be related to the altered structure of the ϩ1 base pair. DISCUSSION In this study, we report for the first time that a potent carcinogenic adduct, ⑀A can trap eukaryotic top1. ⑀A incorporation results in altered base pairing with the thymine on the opposite DNA strand. Circular dichroism analysis demonstrates that the presence of the extra 4-membered ring suppresses the two hydrogen bonds (35). Our observation that ⑀A can reversibly trap top1 is consistent with previous observations indicating that base mismatches at the ϩ1 position can enhance top1 cleavage complexes (19,36).
Reduced religation of top1-cleavable complexes appears to be the mechanism by which ⑀A traps top1. Thus, carcinogenic adducts can mimic CPT effects. Once the DNA is cleaved by top1, perfect positioning of the ϩ1 base is probably required for optimal top1-mediated DNA religation. This positioning closely depends on the hydrogen bonding of the ϩ1 base pairs. True mismatches such as A:A base pairing or uracil misincorporation leading to wobble base pairing such as U:C or U:G also increase top1 cleavage complex formation, whereas U:A has no effect (19). Thus, any modification of the hydrogen bonding would potentially affect the religation (off-rate) of the top1 reaction.
We also found that binding and subsequent cleavage of the ⑀A-containing DNA was significantly reduced. The activity of other DNA processing enzymes such as NheI or DNA polymerases has also been shown to be reduced by the presence of ⑀A adducts (35,37). In our system, top1 was still able to cleave the DNA next to the ⑀A and cleavage was globally enhanced. This demonstrates that inhibition of religation by ⑀A is the predominant mechanism responsible for the increase of top1-mediated cleavage complexes.
CPT has been shown to trap top1 cleavage complexes probably by interacting with the ϩ1 base immediately downstream of the top1 cleavage site (29). Two models for CPT stacking to the ϩ1 base have been proposed recently (16,39). However, no crystal structure of the top1⅐DNA complex in the presence of CPT is available yet. In this study, we found that the effect of CPT was not detectable when ⑀A was incorporated at the ϩ1 position. Moreover, we could not detect alkylation of the DNA in the top1 cleavage complex by 7-ClMe-MDO-CPT when ⑀A was next to the top1 cleavage site. Thus, ⑀A at the ϩ1 position prevents CPT interaction. This could be due to altered base pairing at the ϩ1 position (19). Such base pairing might be important for optimum CPT interaction (stacking with the ϩ1 base pair) within the top1-DNA complex (16,39).
In conclusion, this study indicates that carcinogenic adducts generated from vinyl chloride and other industrial by-products can trap top1 and mimic CPT effects. This trapping might be part of the cell death pathways or repair mechanisms resulting from these lesions.