Production of DNA Strand Breaks in Vitro and Reactive Oxygen Species in Vitro and in HL-60 Cells by PCB Metabolites

Anandi Srinivasan*, Hans-Joachim Lehmler*, Larry W. Robertson*,1 and Gabriele Ludewig{dagger}

* Graduate Center for Toxicology, 306 Health Sciences Research Building, University of Kentucky Medical Center, Lexington, Kentucky 40536–0305; and {dagger} Department of Nutrition and Food Science, 208 Funkhouser Building, University of Kentucky, Lexington, Kentucky 40506–0054

Received June 22, 2000; accepted October 30, 2000


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
PCBs are industrial chemicals that continue to contaminate our environment. They cause various toxic effects in animals and in exposed human populations. The mechanisms of toxicity, however, are not completely understood. PCBs are metabolized by cytochromes P450 to mono- and dihydroxylated compounds. Dihydroxy-PCBs can potentially be oxidized to the corresponding quinones. We hypothesized that reactive oxygen species (ROS) are produced by redox reactions of PCB metabolites. We tested several synthetic dihydroxy- and quinoid-PCBs with 1-3 chlorines for their potential to produce ROS in vitro and in HL-60 human leukemia cells, and DNA strand breaks in vitro. All dihydroxy-PCBs tested produced superoxide. The quinones generated superoxide only in the presence of GSH, probably during the autoxidation of the glutathione conjugates. We observed increased superoxide production with decreasing halogenation. Incubation of dihydroxy-PCBs or PCB quinones + GSH with plasmid DNA resulted in DNA strand break induction in the presence of Cu(II). Tests with various ROS scavengers indicated that hydroxyl radicals and singlet oxygen are likely involved in this strand break induction. Finally, dihydroxy- and quinoid PCBs also produced ROS in HL-60 cells in a dose- and time-dependent manner. We conclude that dihydroxylated PCBs, and PCB quinones after reaction with GSH, produce superoxide and other ROS both in vitro and in HL-60 cells, and oxidative DNA damage in the form of DNA strand breaks in vitro. The reactions seen in vitro and in cells may well be a predictor of the toxicity of PCBs in animals.

Key Words: PCBs; ROS production; DNA strand breaks; quinones; hydroquinones.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Polychlorinated biphenyls (PCBs) are industrial chemicals that were used as dielectrics in capacitors and as flame retardants in cooling fluids and lubricants. Unfortunately, an estimated one-third of the approximately 1.5 million tons of PCB produced worldwide have escaped into the environment (Tanabe, 1988Go), where they persist and accumulate in animal tissues. Humans are exposed to PCBs mostly through food intake, especially from fish and mothers' milk. These characteristics and their propensity to persist and to bioaccumulate raise concerns about the long-term health effects of these chemicals. Commercial PCB mixtures induce liver cancer in rodents (Mayes et al., 1998; reviewed in Silberhorn et al., 1990). In addition, about a dozen epidemiological studies have reported increased mortality due to liver, gall bladder, biliary tract, and digestive tract cancer, malignant melanoma, brain cancer, non-Hodgkin's lymphoma, and lung, and breast cancer in PCB-exposed populations (reviewed in Cogliano, 1998; Moysich et al., 1998). The role of PCBs in carcinogenesis, however, is in dispute. The effects of PCBs as promoters of hepatocarcinogenesis have been thoroughly studied and consistently reported (Safe, 1994Go; Silberhorn et al., 1990Go). The question remains whether PCBs may also act as initiators of carcinogenesis and what mechanism(s) and reactive species are involved.

Several mixtures of PCBs and a few synthetic congeners have been tested in short-term tests for genotoxicity (reviewed in Silberhorn et al., 1990). The results were mostly negative. This could be due in part to slow rate of metabolism of these compounds under the assay conditions. We and many others have shown that PCBs in general (reviewed in Sipes and Schnellmann, 1987), and lower halogenated PCBs in particular (McLean et al., 1996aGo), are metabolized by microsomal enzymes to mono- and dihydroxy-metabolites.2 Further metabolism of the dihydroxy-metabolites, in which the hydroxyl groups are ortho (catechols, Cat) or para to each other (hydroquinones, HQ), may result in the formation of semiquinones and quinones, themselves reactive electrophiles. We have also shown that the metabolic activation of PCBs leads to the formation of adducts with nucleotides and DNA in vitro (McLean et al., 1996bGo; Oakley et al., 1996aGo). Minor adducts may be derived from arene oxide intermediates, whereas several major adducts are more likely to be derived from (semi-) quinone(s). Chemically synthesized PCB quinones are potent electrophiles that react instantaneously with sulfhydryls (N-acetyl cysteine, GSH) and more slowly with N-nucleophiles (Gly, L-Arg, L-His and L-Lys), forming ring addition products (Amaro et al., 1996Go). Reaction of PCB-derived semiquinones or quinones with nucleotides and DNA has also been demonstrated, and several DNA adducts have been separated by thin layer chromatography and detected by 32P-postlabeling (Oakley et al., 1996aGo). By measuring NADPH oxidation, it was shown that synthesized PCB quinones redox-cycle in the presence of NADPH and rat liver microsomes (McLean et al., 2000Go).

The oxidation of dihydroxy-metabolites to quinones may generate superoxide anion radicals and other reactive oxygen species (ROS) via the reduction of molecular oxygen (Fridovich, 1999Go). ROS are known to cause DNA strand breaks that may play a role in mutagenesis and carcinogenesis (Flowers et al., 1997Go). Chronic exposure to PCBs was reported to result in chromosomal aberrations (Kalina et al., 1991Go), which are the result of chromosome breakage. Quinones of polycyclic aromatic hydrocarbons and dihydroxylated metabolites of benzene were shown to induce strand breaks in DNA in vitro in the presence of Cu(II), probably through redox reactions and ROS production (Flowers et al., 1997Go; Li and Trush, 1993Go, 1994Go). In vitro incubation of DNA with PCB hydroquinones in the presence of copper-dichloride (Cu II) increased the levels of 8-oxo-deoxyguanosine (8-oxodG) (Oakley et al., 1996bGo), probably because of ROS production during oxidation of dihydroxylated PCBs to quinones.

We hypothesize that PCBs are involved in carcinogenicity by several mechanisms, one involving the production of ROS that may cause DNA damage. Although several dihydroxy-PCBs have been shown to produce 8-oxo-dG in vitro, no direct evidence of ROS production by these metabolites has been shown. Moreover, the effect of PCB quinones with respect to ROS production has not been studied, and it remains to be shown whether these PCB metabolites will result in oxidative stress in cells. In this study, we therefore analyzed the potential of various dihydroxy- and quinone-PCBs (mostly hydroquinones and p-quinones) to generate superoxide anion radicals and other ROS both in vitro and in cells in culture and their ability to induce DNA damage in the form of strand breaks in vitro.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Materials.
Cu(II), GSH, EDTA, bathocuproine, superoxide dismutase (SOD) (EC 1.15.1.1), catalase (EC 1.11.1.6), dimethyl sulfoxide (DMSO), horseradish peroxidase (HRP), and lactoperoxidase (LactoPx) (EC 1.11.1.7), Tris, nitroblue tetrazolium (NBT) tablets, ethidium bromide, hexadecyl trimethyl ammonium bromide (CTAB), tiron, and trypan blue were purchased from Sigma Chemical Co. (St. Louis, MO). Sodium benzoate, thiourea, and H2O2 (3%) were purchased from Aldrich Chemical Co. (Milwaukee, MI). All salts used for buffers were ACS certified and meet ACS specifications. RPMI 1640, fetal calf serum (FCS), penicillin, and streptomycin were obtained from Gibco BRL (Grand Island, NY). Agarose (Grand Island Biological, Grant Island, NY), sodium azide (Fisher Scientific, Pittsburgh, PA), plasmid bluescript sk (pBS sk; Stratagene, La Jolla, CA), and 2',7'-dichloroflourescein-diacetate (DCFH-DA; Molecular Probes, Eugene, OR) were obtained from the sources indicated.

Synthesis and characterization of dihydroxy-PCBs and PCB p-benzoquinones.
The structures and abbreviated nomenclature of the test compounds used in this study are shown in Figure 1Go. The p-benzoquinones of PCBs were synthesized using the Meerwein arylation as described (Brassard and Ecuyer, 1958Go) and shown in Scheme 1Go. The PCB hydroquinones were prepared from the corresponding p-benzoquinones by reduction with sodium dithionite (Oakley et al., 1996bGo). The detailed synthetic procedure and characterization of the p-benzoquinones and several PCB hydroquinones are given elsewhere (Bauer et al., 1995Go; McLean et al., 1996aGo; Oakley et al., 1996bGo). The PCB catechols were obtained from Suzuki coupling of a suitable dimethoxy benzene boronic acid with a bromochlorobenzene and subsequent deprotection of the dimethoxy-PCB with boron tribromide (BBr3) (Bauer et al., 1995Go). Three new compounds (3ClPh-HQ, 3,5ClPh-HQ, and 3,4,5ClPh-HQ) were characterized by 1H and 13C NMR, FT-IR, and GC/MS spectroscopy. The IR spectra were obtained using a Nicolet Magna-IR 560 Spectrometer E.S.P. The 1H and 13C NMR spectra were recorded on a Varian Gemini-200 spectrometer by using acetone-d6 or deuteriated chloroform (CDCl3, Cambridge Isotope Laboratories Inc., Andover, MA) as solvents and internal standards. GC/MS spectra were recorded in the mass spectrometer facility of the University of Kentucky on a Finnigan INCOS 50 and Kratos Concept 1H by using a fused silica capillary column (DB-5MS 15 m x 0.25 mm and OV-1, 25 m, J&W Scientific, Folsom, CA). For GC/MS all dihydroxybiphenyls were silylated with bis(trimethylsilyl)trifluoroacetamide (BSTFA)/pyridine (1:1). All dihydroxybiphenyls showed a purity of > 99 % by GS/MS based on relative peak area after several recrystallizations from chloroform/petroleum ether. The p-benzoquinones were analyzed with a Hewlett Packard 5890A gas chromatograph equipped with an HP-1 (methyl silicone gum) column (Hewlett Packard, Avondale, PA) and showed a purity of > 99 % based on relative peak area. Melting points were determined on an MEL-TEMP apparatus and are uncorrected. All solvents were obtained from commercial sources and used without further purification. The dihydroxybiphenyls were stored at 4°C under an argon atmosphere to minimize oxidation to the corresponding p-benzoquinone. To avoid depletion of any single compound, 2ClPh-HQ was used as standard compound in the NBT assays and 4ClPh-HQ in the strand break assays. For the assays stock solutions of PCB metabolites were prepared in DMSO. Caution: Synthetic PCB metabolites should be considered potentially toxic and hazardous and should therefore be handled in an appropriate manner.



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 1. Structures of the synthesized PCB metabolites studied in this paper.

 


View larger version (7K):
[in this window]
[in a new window]
 
SCHEME 1. Synthesis of p-quinone and p-dihydroxy derivatives of chlorinated biphenyls.

 
2-(3'-Chlorophenyl)-1,4-hydroquinone (3ClPh-HQ)1: mp = 75–77°C (white solid from chloroform/petroleum ether at 4°C). 1H-NMR (acetone-d6, 200 MHz) {delta} 6.25–6.61 (3H, m, H-3/4/6), 6.97 ("t", J = 8.0 Hz, "t", J = 1.6 Hz, 4'-H), 7.06 ("t", J = 8.0 Hz, 5'-H), 7.17 ("t", J = 8.0 Hz, "t", J = 1.8 Hz, 6'-H), 7.29 ("t", J = 1.6 Hz, 2'-H), 8.01 (2H, br s, OH). 13C-NMR (acetone-d6, 50 MHz) {delta} 116.69, 117.51, 118.01, 127.42, 128.25, 128.58, 130.09, 130.56, 134.17, 142.05, 148.03, 151.74. IR [cm–1]: 3273 ({nu}(OH)), 1596, 1565, 1510, 1472, 1454, 1414, 1362, 1248, 1204, 1184, 1121, 807, 784, 754, 691. MS m/z (relative intensity, %): 45 (22), 73 (100), 93 (10), 314 (29), 364/366 (35/15, C16H25Si2O2Cl+•).

2-(3',5'-Dichlorophenyl)-1,4-hydroquinone (3,5ClPh-HQ)1: mp = 118°C (white solid from chloroform/petroleum ether at 4°C). 1H-NMR (acetone-d6, 200 MHz) {delta} 6.75 (d, J = 8.7 Hz, d, 2.7 Hz, 4-H), 6.84 (d, J = 2.7 Hz, 6-H), 6.86 (d, J = 8.7 Hz, 3-H), 7.39 (t, J = 1.8 Hz, 4'-H), 7.58 (d, J = 1.8 Hz, 2'- and 6'-H), 8.04 (br s, 2- and 5-OH). 13C-NMR (acetone-d6, 50 MHz) {delta} 117.07, 117.32, 118.09, 127.07, 130.13, 130.80, 131.01, 132.07, 132.16, 140.51, 148.03, 151.80. IR [cm–1]: 3211 ({nu}(OH)), 1588, 1559, 1452, 1409, 1384, 1242, 1192, 1126, 857, 811, 790, 782, 758, 684. MS m/z (relative intensity, %): 45 (21), 73 (100), 93 (10), 275 (20), 348 (10), 398/400 (32/23, C16H24Si2O2Cl2+•).

2-(3',4',5'-Trichlorophenyl)-1,4-hydroquinone (3,4,5ClPh-HQ)1: mp = 196–197°C (white solid from chloroform/petroleum ether at 4°C). 1H-NMR (acetone-d6, 200 MHz) {delta} 6.34–6.61 (3H, m, H-3/4/6), 7.43 (2H, s, H-2'/6'), 7.78 (2H, br s, OH). 13C-NMR (acetone-d6, 50 MHz) {delta} 117.19, 117.68, 118.22, 125.83, 130.58 (2C), 133.99, 140.69, 148.07, 151.86. IR [cm–1]: 3248 ({nu}(OH)), 1582, 1540, 1508, 1470, 1429, 1383, 1320, 1305, 1258, 1209, 1192, 1154, 1131, 868, 813, 802, 776, 730, 694, 634, 617. MS m/z (relative intensity, %): 45 (22), 73 (100), 93 (12), 382/384 (17/22), 442/434 (30/33, C16H23Si2O2Cl3+•).

Measurement of superoxide production in vitro (NBT assay).
The NBT assay used is a modified version of the assay as described by Epe et al. (1986). Two milliliters of solutions containing 2 mM phosphate buffer (pH 7.0), 5 mM CTAB, and 100 µM NBT alone or with a) Cu(II) (100 µM); b) SOD (600U/ml); c) HRP (2.5U/ml) or LactoPx (0.1U/ml) and H2O2 (10 µM); or d) GSH (50, 100, 200 µM) were prepared and added to both the reference and sample quartz cuvettes. The reaction was started by adding PCB metabolite to a final concentration of 100 µM to the sample cuvette and DMSO to the reference cuvette. The rate of NBT reduction was monitored at 540 nm (37°C) for 10 min with a Shimadzu MPS 2000 UV/VIS spectrophotometer. Statistical analysis was performed using a SAS program (PROC mixed for repeated measures).

Measurement of strand break induction in plasmid DNA.
Supercoiled plasmid DNA (pBS sk, about 3kb size; 1 µg/15 µl) was exposed to the test compound and modifiers for 1 h at 37°C in 2 mM phosphate buffer (total sample volume = 15 µl). Immediately after exposure, the DNA was loaded onto 0.8% agarose gels in 1x Tris acetate/EDTA (TAE) buffer (40 mM Tris acetate, 1 mM EDTA, pH 7.0) and gel electrophoresis was carried out for 2 h at 60 V in a horizontal gel electrophoresis apparatus (Owl Scientific, MA). The gels were stained with ethidium bromide (1 µg/ml water) and the DNA visualized with a UV transilluminator. Pictures were taken with a Sony video graphics printer video camera attached to a Biophotonica gel print 2000i transilluminator and used for documentation. The gel pictures were scanned and total amount of DNA in each lane as well as the relative amounts of plasmid DNA with no strand breaks (supercoiled), single strand breaks (open circle), and double strand breaks (linear form) were determined. The amount of DNA in each form was expressed as percentage of the total DNA in each lane (% of intensity of optical density, IOD).

In order to establish standard assay conditions for our experiments, we performed first strand break experiments with 4ClPh-HQ + Cu(II) (100 µM each) using different buffers and pHs. We observed the same extent of strand break induction in 2 mM phosphate and 4 mM Tris buffer, but reduced strand breaks when 2 mM Na-citrate or a Tris-EDTA buffer was used. Because citrate and EDTA chelate Cu(II) and thus inhibit copper mediated dismutation of ROS, and because Tris is a reactive oxygen scavenger (Tsou et al., 1996Go), we decided to use 2 mM phosphate buffer for all experiments. We also found that at acidic pHs, less damage was observed than at alkaline or neutral pHs. Therefore, all experiments described here were performed at pH 7.0.

Cell culture and viability test.
The human promyelocytic leukemia cell line HL-60 was obtained from Dr. M. Doukas, University of Kentucky, Lexington, KY. Cells (passages 70–90) were cultured in RPMI 1640 medium supplemented with 5% FCS, penicillin (0.1 U/ml) and streptomycin (0.1 µg/ml) and grown in a humidified atmosphere at 5% CO2 and 37°C.

For viability testing, 1 x 106 HL-60 cells in 1 ml PBS (pH 7.4) were exposed to the test compound or solvent alone. After 1 and 3 h at 37°C, aliquots of cells were removed and treated with an equal volume of 10% trypan blue solution. The number of viable cells in each sample was determined by counting unstained cells using a hemocytometer and a light microscope. All determinations were done in triplicate. Data are presented as percent of solvent-treated control. Statistical significance was determined using the ANOVA and post hoc Tukey's test.

Measurement of ROS in HL-60 cells using DCFH-DA.
3 x 105 HL-60 cells in 3 ml PBS (pH 7.4) were treated with various concentrations of 4ClPh-pQ, 4ClPh-HQ, or solvent alone (control) in the presence of 10 µM DCFH-DA. After 1 h or 3 h of incubation at 37°C, fluorescence of the cell suspension was determined using a RF-5301 PC Shimadzu spectrofluorophotometer with an excitation wavelength at 485 nm and emission wavelength at 530 nm. All experiments were carried out in triplicate, and the means and standard deviations (SD) were calculated per data point. Net ROS production was calculated by subtracting the fluorescence of the compounds in PBS without cells from the corresponding incubation with cells in suspension. Statistical significance was determined using the ANOVA and post hoc Tukey's test.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
NBT Reduction
NBT is stable in aqueous solution, but is reduced by superoxide to a blue formazan, the formation of which can be monitored spectrally at 540 nm (Epe et al., 1986Go). CTAB greatly facilitates NBT reduction by various compounds (Takabe et al., 1978Go), and in prior tests we made the same observation with PCB metabolites. Therefore, 5 mM CTAB was added to each NBT experiment. Under these conditions, the addition of 100 µM of 2ClPh-HQ caused a statistically significant increase (p < 0.0001) in absorbance at 540 nm in the absence of any added oxidant (Fig. 2AGo). The addition of SOD to the solution containing 2ClPh-HQ completely inhibited NBT reduction compared to 2ClPh-HQ alone (p < 0.001; Fig. 2AGo). The presence of 100 µM Cu(II) also inhibited formazan formation (p < 0.001). We compared the effects of 6 PCB hydroquinones with 1 to 3 chlorines in the second ring in the NBT assay (Fig. 2BGo). There was a clear structure-activity relationship for NBT reduction. The order of decreasing activity for the various PCB hydroquinones is 2ClPh-HQ ~ 3ClPh-HQ ~ 4ClPh-HQ > 3,5ClPh-HQ ~ 3,4ClPh-HQ > 3,4,5ClPh-HQ (Fig. 2BGo). The monochlorinated PCB metabolites were significantly different from higher chlorinated PCB metabolites after 1 min of incubation (p < 0.05); the dichlorinated compounds became significantly different from the trichlorinated metabolite at later time points (p < 0.05). Obviously the position of chlorines did not have a major effect on superoxide production, but an increase in the number of chlorines dramatically decreased NBT reduction.



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 2. Reduction of NBT by PCB-generated superoxide, monitored by measuring absorbance at 540 nm for 10 min. (A) Reduction of NBT by 2ClPh-HQ (100 µM) alone (black circles) or in the presence of lactoperoxidase + H2O2 (white circles), SOD (black triangles), or Cu(II) (white triangles). (B) Reduction of NBT by selected PCB hydroquinones (100 µM). (C) Reduction of NBT by 100 µM 2ClPh-pQ in the presence of various amounts of GSH and SOD or Cu(II).

 
The PCB quinone metabolites by themselves caused no NBT reduction (data not shown). As shown with 2ClPh-pQ, as an example, adding increasing amounts of GSH (ratios of GSH:quinone of 0.5:1, 1:1, and 2:1) significantly enhanced reduction of NBT, especially at the two higher GSH concentrations (p < 0.01) (Fig. 2CGo). The addition of SOD to a quinone:GSH (1:1) reaction completely inhibited NBT reduction, proving that superoxide is involved (p < 0.01). The addition of Cu(II) also completely inhibited formazan formation (p < 0.01). The amount of superoxide generated by PCB-quinone:GSH (1:1) was much more than with the corresponding PCB-hydroquinone. However, in the presence of GSH, 2ClPh-HQ produced a reduction of NBT similar to that by 2ClPh-pQ + GSH (data not shown).

Strand Break Induction
Strand break induction in DNA can be monitored using supercoiled plasmid DNA. When treated with an agent that induces DNA strand breaks, supercoiled DNA is converted to the open-circle form by a single strand break and to linear DNA by a double strand break. Extensive double strand breaks lead to DNA degradation. These effects can be monitored by gel electrophoresis. The migration pattern of a 3-kb plasmid under our gel electrophoresis conditions is: supercoiled > linear > open circle (Fig. 3AGo).



View larger version (42K):
[in this window]
[in a new window]
 
FIG. 3. Strand break induction measured by conformation analysis of supercoiled plasmid DNA (pBS sk) after incubation with 4ClPh-HQ/Cu(II) in PBS at 37°C for 1 h and gel electrophoresis. (A) Effect of increasing concentrations of 4ClPh-HQ in the presence of 100 µM Cu(II). (B) Effect of different concentrations of Cu(II) on strand break induction by 4ClPh-HQ (100 µM). The percentage of each form of the plasmid was determined by densitometry (white: supercoiled; striped: open circle; black: linear plasmid DNA).

 
To explore the optimal assay conditions, 4ClPh-HQ was used as test compound. Incubation of 4ClPh-HQ with plasmid DNA alone resulted in no strand break induction (Fig 3BGo, lane 1). However, adding increasing concentrations of 4ClPh-HQ in the presence of 100 µM Cu(II) resulted in increased strand scission (Fig. 3AGo). The highest concentration of 4ClPh-HQ tested, 2.5 mM, led to degradation of DNA due to extensive double strand breaks. Cu(II) alone did not induce DNA strand breaks. Fig 3BGo shows the results of an incubation of plasmid DNA with 100 µM 4ClPh-HQ and increasing concentrations of Cu(II). At about equimolar ratios of Cu(II) to 4ClPh-HQ, single strand breaks were seen as indicated by the disappearance of supercoiled plasmid and increase in open-circle form plasmid. Doubling the concentration of Cu(II) resulted in some linear DNA. Increasing Cu(II) up to 5 mM had no further effect. Beyond 5 mM the DNA was completely degraded. The extent of strand breaks induced by 100 µM 4ClPh-HQ and 100 µM Cu(II) increased with time of exposure (Fig. 4Go). After 10 min of incubation, a decrease in supercoiled and an increase in open-circle DNA were observed. Longer incubation times led to increasing formation of linear plasmid DNA. Exposure times > 1 h resulted in degraded DNA, which formed a DNA smear. Therefore, 100 µM Cu(II) and a 1-h incubation time were used in all following experiments.



View larger version (69K):
[in this window]
[in a new window]
 
FIG. 4. Time course of DNA strand break induction by 4ClPh-HQ plus Cu(II) (100 µM each) in supercoiled plasmid BS sk DNA in PBS at 37°C.

 
Various dihydroxy-PCB metabolites, hydroquinones and catechols were analyzed for comparison of strand break induction. Visual comparison of the bands in the gels gives the impression that all PCB hydroquinones tested induced strand breaks to an almost equal extent in the presence of Cu(II) (Fig. 5Go). We also analyzed the influence of the position of the dihydroxy groups on activity using 4-chlorobiphenyl metabolites. There was a small difference visible in the amount of strand break induction by 4ChPh metabolites; more linear form was seen with 4ClPh-HQ than with 4ClPh-3',4'-catechol or 4ClPh-2',3'-catechol (Fig. 5Go, lanes 4, 6, 7). The oxidized derivative of the hydroquinone, 4ClPh-pQ, which was included for comparison, induced very few strand breaks under these conditions (Fig. 5Go, lane 8). 4ClPh-pQ produced only a small reduction in supercoiled plasmid DNA in the presence of Cu(II) (Fig. 6Go). When GSH was added to the incubation mixture, however, an increase in strand breaks was observed. This effect increased with dose up to a 1:1 ratio of GSH:4ClPh-pQ.



View larger version (29K):
[in this window]
[in a new window]
 
FIG. 5. Strand break induction in supercoiled plasmid BS sk by dihydroxylated PCBs and 4ClPh-pQ in the presence of Cu(II). (1) Cu(II) alone, with (2) 2ClPh-HQ, (3) 3ClPh-HQ, (4) 4ClPh-HQ, (5) 3,5ClPh-HQ, (6) 4ClPh-2',3'-catechol, (7) 4ClPh-3',4'-catechol, (8) 4ClPh-pQ. All compounds and Cu(II) are present in 100 µM concentrations.

 


View larger version (50K):
[in this window]
[in a new window]
 
FIG. 6. Influence of GSH on strand break induction by 4ClPh-pQ + Cu(II) (100 µM each). (1) 4ClPh-pQ alone, with (2) Cu(II) and (3) 25 µM GSH, (4) 50 µM GSH, (5) 100 µM GSH, (6) 500 µM GSH.

 
Various free radical scavengers and transition metal chelators were tested in the strand break assay to analyze which reactive species were involved in strand break induction. For these studies 4ClPh-HQ + Cu(II) was used for strand break induction. Bathocuproine, a Cu(I)-specific inhibitor, completely prevented strand breaks (Fig. 7Go, Table 1Go). In the presence of the divalent cation chelator EDTA (100 µM), strand break induction was completely inhibited. GSH, a redox cofactor, was found to protect against 4ClPh-HQ + Cu(II)-induced strand breaks. SOD and Tiron, both superoxide scavengers, did not reduce the amount of strand breaks by 4ClPh-HQ + Cu(II) (Table 1Go). In the presence of 4ClPh-HQ + Cu(II), the enzyme catalase, which dismutates H2O2 to water, had a slightly protective effect. However, denatured (boiled) catalase was also slightly protective, which indicates that other mechanisms than enzyme activity (protein binding?) may be involved. Of three hydroxyl radical scavengers tested, sodium benzoate had only a slight protective effect on 4ClPh-HQ + Cu(II)–induced strand breaks at 50–75 mM. However, sodium benzoate induced DNA degradation at >= 100 mM concentrations. DMSO conferred some protection, but only at very high concentrations (Fig. 7Go). Thiourea protected completely. Of the singlet oxygen scavengers, both sodium azide and Tris prevented 4ClPh-HQ + Cu(II)–induced strand break induction.



View larger version (57K):
[in this window]
[in a new window]
 
FIG. 7. Effect of different modifiers on 4ClPh-HQ + Cu(II) induced strand breaks after 1-h exposure at 37°C. The percentage of each form of the plasmid was determined by densitometry and calculated as % of total DNA (white: supercoiled; striped: open circle; black: linear plasmid). Cuvettes contained 100 µM Cu(II) and 4ClPh-HQ each, in addition to the modifier, unless otherwise stated.

 

View this table:
[in this window]
[in a new window]
 
TABLE 1 Effect of Modifiers on Strand Break Induction by 4ClPh-HQ
 
ROS Generation and Cytotoxicity in HL-60 Cells
The probe DCFH-DA is converted by ROS in cells to the highly fluorescent 2',7'-dichlorofluorescein (DCF; LeBel et al., 1992). An increase in fluorescence due to intracellular DCF formation by ROS can easily be monitored with a fluorometer. Figure 8Go shows the significant increase in fluorescence intensity (DCF formation) over time with both PCB metabolites tested after 1 h of incubation. 4ClPh-pQ significantly increased DCF fluorescence at concentrations as low as 0.5 µM, whereas 4ClPh-HQ increased fluorescence significantly at concentrations of 2.5 µM or higher. After 3 h of incubation, both metabolites increased ROS production almost 5 times compared to the fluorescence at the same concentrations after 1 h. At the 3-h time point, fluorescence in 4ClPh-pQ–treated samples was significantly higher than DMSO controls at all concentrations tested. 4ClPh-HQ significantly increased fluorescence compared to solvent controls at concentrations > 0.5 µM.



View larger version (29K):
[in this window]
[in a new window]
 
FIG. 8. ROS generation determined by measuring increase in fluorescence due to DCFH-DA oxidation. Effect of time and concentration on 4ClPh-pQ and 4ClPh-HQ induced DCF fluorescence formation. HL-60 cells (3 x 105 cells in 3 ml PBS with 10 µM DCFH-DA) were exposed to the compounds for 1 or 3 h at different concentrations of 4ClPh-pQ and 4ClPh-HQ. All exposures were done in triplicate. Background fluorescence of compounds in PBS alone was subtracted. *Significantly different from solvent control (p < 0.05).

 
The viability of cells in PBS after exposure to 4ClPh-pQ and 4ClPh-HQ was tested with the trypan blue exclusion test. The 4ClPh-HQ was not toxic at the concentrations tested (Table 2Go). 4ClPh-pQ significantly reduced the number of viable cells compared to controls only after 3-h exposure in PBS at 5 µM concentration (p < 0.05).


View this table:
[in this window]
[in a new window]
 
TABLE 2 Viability of HL-60 Cells after Treatment with 4ClPh-HQ or 4ClPh-pQ
 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
We hypothesized that there are at least two mechanisms by which ROS may be produced nonenzymatically by these PCB metabolites: a) by autoxidation of dihydroxy-PCBs, and b) by reduction of quinones via Michael addition of GSH with autoxidation of the thioether. Both are depicted in Scheme 2Go and may result in superoxide production. We used the NBT reduction assay to test if superoxide is produced during autoxidation of the PCB metabolites. All PCB hydroquinones tested resulted in NBT reduction. Because SOD nearly completely inhibited NBT reduction (Fig. 2Go), superoxide is most likely involved. The addition of equimolar concentrations of Cu(II) to the NBT + PCB metabolite mixtures also inhibited NBT reduction. Cu(II) reacts with superoxide to yield Cu(I) and oxygen, thus eliminating superoxide from the reaction mixture before it can reduce NBT. This strongly suggests that superoxide was formed during PCB hydroquinone oxidation. The yield of nonenzymatic superoxide production in the NBT assay in the presence of CTAB was very high. Takabe and coworkers have shown that CTAB, a surface-active cation, significantly accelerates NBT reduction by polyphenols by a faster electron transfer from superoxide to NBT (Takabe et al., 1978Go). In the absence of CTAB, autoxidation was much slower, and longer incubation times and/or enzymatic oxidation were needed to obtain measurable changes.



View larger version (19K):
[in this window]
[in a new window]
 
SCHEME 2. Proposed mechanism for strand break induction and ROS production by the PCB metabolites. PCB metabolites are oxidized to (semi)quinones accompanied by the production of superoxide, which dismutates to other, DNA-damaging forms of reactive oxygen species.

 
The ROS production should result in oxidative DNA damage. Previously we had reported that PCB metabolites form adducts in vitro with nucleotides and DNA (McLean et al., 1996bGo; Oakley et al., 1996aGo). Here we tested for additional DNA damage due to oxidative stress in the form of DNA strand breaks in vitro. All dihydroxy-PCBs tested caused DNA strand breaks in the presence of Cu(II). Most likely, a ROS derived from superoxide is involved, although a semiquinone as strand-breaking agent cannot be excluded. There was a dose-dependent increase in damage while increasing either the concentrations of dihydroxy-PCBs or Cu(II). Quantitatively it can be estimated from our results that under these conditions about 0.34 pmoles supercoiled plasmid DNA (two-thirds of the 1-µg plasmid) was converted by 1.9 nmoles 4ClPh-HQ to open-circle DNA, or a strand break to PCB-HQ ratio of about 1:5.3 x 103/h. This seems very low. Strand break induction by 4ClPh-HQ was also time dependent, with increasing amounts of open-circle and linear DNA between 10–60 min, and deteriorated DNA after longer incubation times. This suggests that autoxidation is happening as a slow and continuing process in the absence of enzymes or CTAB. (CTAB could not be included in this assay, as it caused strand breaks by itself.)

For risk assessment, it is important to have information about structure-activity relationships. We observed more open-circle form after treatment with 4ChPh-HQ than with the two catechols (Fig. 5Go). This may be explainable by differences in the reduction potentials. Indeed, Amaro and coworkers (Amaro et al., 1996Go) reported that in nonaqueous solution the reduction potentials for the pQ/SQ and the SQ/HQ pairs of 4ClPh- and 3,4ClPh-para-metabolites were about 100–130 mV lower than the ones for the corresponding 2,3-ortho metabolites, indicating a lower likelihood of oxidation for the catechols. PCB-HQs were also more active than 3',4'-catechols in inducing 8-oxodG formation in vitro (Oakley et al., 1996bGo), probably due to a common mechanisms for induction of strand breaks and 8-oxodG.

We did not see an influence of the number or position of chlorines on strand break induction in our gels, possibly due to the generally slow autoxidation rate. However, we observed a structure-activity relationship in the NBT reduction assay: mono-ClPh- > di-ClPh- > triClPh-HQ (Fig. 2Go). The reported reduction potentials for mono- and dichlorinated metabolites point to a small (about 20 mV), but consistent decrease in the redox potential with the higher chlorination (Amaro et al., 1996, measured in nonaqueous medium), but other factors such as solubility or protonation may also be involved in this effect.

The PCB-quinones were not (NBT) or hardly (strand breaks) active, but were activated by the addition of GSH (Figs. 2C and 6GoGo, respectively). Amaro and coworkers have shown that PCB-quinones react instantaneously with GSH by nucelophilic addition and have characterized the resulting hydroquinone thioethers (Amaro et al., 1996Go). As the addition of GSH to a quinone can actually lower the reduction potential, as was shown for menadione (reviewed in Monks and Lau, 1997), GSH adducts may repeatedly autoxidize with the formation of superoxide. In combination with the PCB quinones, GSH was clearly activating by promoting the generation of superoxide. The reduction of NBT by PCB hydroquinones was also increased, perhaps by the same quinone-recycling mechanism. Strand break induction by PCB-HQs + Cu(II), however, was strongly reduced in the presence of GSH, possibly by stabilizing copper in the cuprous (+1) form (Milne et al., 1993Go), thus preventing redox cycling of copper and OH formation from H2O2 by a Fenton reaction, and/or scavenging of OH (Hanna and Mason, 1992Go).

Copper, which was essential for strand break induction, is an essential micronutrient present in all mammalian tissues. It has a preferential accumulation in the nucleus (Agarwal et al., 1989Go), where it is involved in the formation of higher order chromatin structures due to its role in the attachment of DNA to nuclear scaffolding proteins (Lewis and Laemmli, 1982Go). Copper is highly active as a fenton reagent in converting H2O2 to the very reactive OH (Cu+ + H2O2 -> Cu2+ + OH + OH) (Goldstein et al., 1993Go). It may also result in singlet oxygen production (Yamamoto et al., 1993Go). Because copper is localized in such close proximity to DNA, it would produce locally high concentrations of these radicals; thus, its role in PCB toxicity is of particular interest. We have shown previously that the presence of copper significantly increased 8-oxodG formation in DNA by various dihydroxy-PCBs (Oakley et al., 1996bGo). It was reported that DNA-bound copper ions produce primarily base modifications (Drouin et al., 1996Go). In this paper we report that copper is also required for strand break induction by dihydroxy and quinone metabolites of PCBs, an effect that was also seen with other compounds like o-quinones of polycyclic aromatic hydrocarbons (PAHs; Flowers et al., 1997), hydroquinones (Carstens et al., 1990Go; Li and Trush, 1993Go), estrogen metabolites (Chen et al., 1998Go), and polyphenolic compounds (Li and Trush, 1994Go). EDTA, which chelates Cu(II) by forming a redox inert complex (Agarwal et al., 1989Go; Li and Trush, 1993Go), prevented 4ClPh-HQ + Cu(II)–induced strand breaks, as did bathocuproine, a Cu(I)-specific chelator. This suggests that it is not Cu(II)-mediated HQ oxidation, but Cu(I)-mediated formation of other ROS from superoxide or its dismutation product H2O2 that is required for strand break induction.

Some scientists reported that the reaction of cuprous ions with hydrogen peroxide generates a Cu(I)-OOH complex that may cause the DNA damage (Murata et al., 1999Go). Others suggest that this complex is only formed initially and then decomposes into OH, which is the major species derived from the reaction of Cu(I) with H2O2 (Eberhardt et al., 1989Go). To determine experimentally the exact nature of the DNA-breaking radical, we included various modifiers in the incubations during the strand break assay. The superoxide scavenger Tiron did not decrease stand break induction, possibly because the concentration tested was not high enough. SOD had no effect, perhaps because superoxide production and not nonenzymatic dismutation of superoxide is the rate-limiting step. Adding H2O2, the product of SOD metabolism of superoxide, to plasmid DNA + Cu(II) resulted in totally destroyed DNA (unpublished observation), indicating that H2O2 or a H2O2-derived oxygen radical is the ultimate clastogen. Two highly reactive oxygen species are OH and singlet oxygen. The OH scavenger Na-benzoate caused only slight protection of the DNA from 4ClPh-HQ + Cu(II)–induced strand breaks, and itself produced DNA degradation in this system. The OH scavengers GSH, DMSO, and thiourea decreased strand breaks in a dose-dependent manner—DMSO marginally, GSH and thiourea completely. Hence, OH seems to be a major player in strand break induction by the PCB metabolites. This is in agreement with results from strand break experiments with PAH-quinones, polyphenols, and estrogen metabolites (Carstens et al., 1990Go; Flowers et al., 1997Go; Li and Trush, 1993Go, 1994Go). We observed in addition, however, that the singlet oxygen scavengers sodium azide and Tris also reduced strand break generation by 4ClPh-HQ + Cu(II). This radical could be produced by spontaneous dismutation of superoxide (Corey et al., 1987Go), or be more likely derived from Cu/peroxide (Yamamoto et al., 1993Go). This suggests that these two oxygen species, OH and singlet oxygen, are involved in PCB strand break induction.

The data as discussed above clearly show that PCB metabolites do produce ROS and consequently DNA damage under various conditions in vitro, but what are the effects of these metabolites in cells in culture? Our data provide for the first time direct evidence of intracellular production of ROS by PCB metabolites. The DCFH-DA dye has been used to detect ROS in many types of cells and tissues (LeBel et al., 1992Go). DCFH-DA is a nonionic and nonpolar substance that can cross the cell membrane and is hydrolyzed by cellular esterases to nonfluorescent DCFH. In the presence of ROS, DCFH is converted rapidly to highly fluorescent DCF (LeBel et al., 1992Go). It was reported that oxidation of DCFH occurs almost exclusively in the cytosol (LeBel et al., 1992Go). We saw some background production of DCF in PBS, much less, however, than in the presence of cells, indicating that the PCB metabolites do produce a strong net increase in intracellular ROS in HL-60 cells in a dose- and time-dependent manner (Fig. 8Go). To confirm that the fluorescence was intracellular and not due to leakage of DCFH or DCF from damaged cells, we examined the cells by confocal fluorescence microscopy. A clear increase in the number of fluorescing cells and intensity of intracellular fluorescence was visible in samples treated with PCB metabolites. In addition, our toxicity tests show that the HL-60 cells were mostly viable at the concentrations tested for ROS production, with the exception of the high concentration of the quinone. HL-60 cells contain considerable amounts of myeloperoxidase (Shen et al., 1996Go), which is efficient in PCB-HQ oxidation to the quinone. The PCB quinone was even more active than the PCB-HQ in producing ROS, probably by being activated by conjugation and by depletion of intracellular GSH.

Altogether, our studies demonstrate the production of ROS by PCB metabolites both in vitro and in cells in culture. The activated oxygen species generated may attack cellular DNA and other nucleophiles, resulting in oxidative damage. Indeed, DNA strand breaks were produced in vitro with these metabolites. These mechanisms may play an important role in the different toxic effects of PCBs.


    ACKNOWLEDGMENTS
 
The authors thank P. Brown and J. Shaw for synthesizing some of the compounds used, Dr. P. Espandiari, Graduate Center for Toxicology, for technical support, Dr. J. O'Reilly, Department of Chemistry, for helpful discussions, and Dr. Marta S. Mendiondo, Sanders-Brown Center on Aging, and Dr. Richard J. Kryscio, Biostatistics Consulting Unit, for the statistical analysis of the NBT data. This publication was made possible by grant DAMD 17-96-1-6262 from the Department of Defense, by grant P42 ES 07380 from the National Institute of Environmental Health Sciences (NIEHS)with funding provided by the U.S. Environmental Protection Agency (EPA), and by grant 85-001-13-IRG from the American Cancer Society (ACS). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the ACS, NIEHS, National Institutes of Health, or the U.S. EPA.


    NOTES
 
1 To whom correspondence should be addressed. Fax: (859) 323-1059. E-mail: lwrobe01{at}pop.uky.edu. Back

2 Dihydroxy is used as a general term for all dihydroxylated PCBs. Para metabolites are described by the term hydroquinone (HQ). For ortho metabolites, the term catechol (Cat) is used. Back

A portion of this work was presented at the PCB Workshop, Recent Advances in the Environmental Toxicology and Health Effects of PCBs, April 9–12, 2000, Lexington, KY.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Agarwal, K., Sharma, A., and Talukder, G. (1989). Effects of copper on mammalian cell components. Chem.-Biol. Interact. 69, 1–16.[ISI][Medline]

Amaro, A. R., Oakley, G. G., Bauer, U., Spielmann, H. P., and Robertson, L. W. (1996). Metabolic activation of PCBs to quinones: reactivity toward nitrogen and sulfur nucleophiles and influence of superoxide dismutase. Chem. Res. Toxicol. 9, 623–629.[ISI][Medline]

Bauer, U., Amaro, A. R., and Robertson, L. W. (1995). A new strategy for the synthesis of polychlorinated biphenyl metabolites. Chem. Res. Toxicol. 8, 92–95.[ISI][Medline]

Brassard, P., and Ecuyer, P. L. (1958). L' arylation des quinones par les sels de diazonium. Can. J. Chem. 36, 700–708.[ISI]

Carstens, C. P., Blum, J. K., and Witte, I. (1990). The role of hydroxyl radicals in tetrachlorohydroquinone induced DNA strand break formation in PM2 DNA and human fibroblasts. Chem.-Biol. Interactions 74, 305–314.[ISI][Medline]

Chen, Y., Shen, L., Zhang, F., Lau, S. S., van Breeman, R. B., Nikolic, D., and Bolton, J. L. (1998). The equine estrogen metabolite 4-hydroxyequilenin causes DNA single-strand breaks and oxidation of DNA bases in vitro. Chem. Res. Toxicol. 11, 1105–1111.[ISI][Medline]

Cogliano, V. J. (1998). Assessing the cancer risk from environmental PCBs. Environ. Health Perspect. 106, 317–23.[ISI][Medline]

Corey, E. J., Mehrotra, M. M., and Khan, A. U. (1987). Water induced dismutation of superoxide anion generates singlet molecular oxygen. Biochem. Biophys. Res. Commun. 145, 842–846.[ISI][Medline]

Drouin, R., Rodriguez, H., Gao, S. W., Gebreyes, Z., O'Connor, T. R., Holmquist, G. P., and Akman, S. A. (1996). Cupric ion/ascorbate/hydrogen peroxide-induced DNA damage: DNA-bound copper ion primarily induces base modifications. Free Radical Biol. Med. 21, 261–273[ISI][Medline]

Eberhardt, M. K., Ramirez, G., and Ayala, E. (1989). Does the reaction of Cu+ with H2O2 give OH radicals? J. Org. Chem. 54, 5922–5926.[ISI]

Epe, B., Schiffmann, D., and Metzler, M. (1986). Possible role of oxygen radicals in cell transformation by diethylstilbestrol and related compounds. Carcinogenesis 7, 1329–1334.[Abstract]

Flowers, L., Ohnishi, S. T., and Penning, T. M. (1997). DNA strand scission by polycyclic aromatic hydrocarbon o-quinones: Role of reactive oxygen species, Cu(II)/Cu(I) redox cycling, and o-semiquinone anion radicals. Biochemistry 36, 8640–8648.[ISI][Medline]

Fridovich, I. (1999). Fundamental aspects of reactive oxygen species, or what's the matter with oxygen? Ann. N.Y. Acad. Sci. 893, 13–8.[Abstract/Free Full Text]

Goldstein, S., Meyerstein, D., and Czapski, G. (1993). The fenton reagents. Free Radical Biol. Med. 15, 435–445.[ISI][Medline]

Hanna, P. M., and Mason, R. P. (1992). Direct evidence for inhibition of free radical formation from Cu(I) and hydrogen peroxide by glutathione and other potential ligands using the EPR spin-trapping technique. Arch. Biochem. Biophys. 295, 205–213[ISI][Medline]

Kalina, I., Sram, R. J., Konecna, H., and Ondrussekova, A. (1991). Cytogenetic analysis of peripheral blood lymphocytes in workers occupationally exposed to polychlorinated biphenyls. Teratog. Carcinog. Mutagen. 11, 77–82.[ISI][Medline]

LeBel, C. P., Ischiropoulos, H., and Bondy, S. C. (1992). Evaluation of the probe 2',7'-dichlorofluorescin as an indicator of reactive oxygen species formation and oxidative stress. Chem. Res. Toxicol. 5, 227–231.[ISI][Medline]

Lewis, C. D., and Laemmli, U. K. (1982). Higher order metaphase chromosome structure: Evidence for metalloprotein interactions. Cell 29, 171–181.[ISI][Medline]

Li, Y., and Trush, M. A. (1993). DNA damage resulting from the oxidation of hydroquinone by copper: Role for a Cu(II)/Cu(I) redox cycle and reactive oxygen generation. Carcinogenesis 14, 1303–1311.[Abstract]

Li, Y., and Trush, M. A. (1994). Reactive oxygen-dependent DNA damage resulting from the oxidation of phenolic compounds by a copper-redox cycle mechanism. Cancer Res. (Suppl.) 54, 1895s–1898s.[Abstract]

Mayes, B. A., McConnell, E. E., Neal, B. N., Brunner, M. J., Hamilton, S. B., Sullivan, T. M., Peters, A. C., Ryan, M. J., Toft, J. D., Singer, A. W., Brown, J. F., Menton, R. G., and Moore, J. A. (1998). Comparative carcinogenicity in Sprague-Dawley rats of the polychlorinated biphenyl mixtures Arochlors 1016, 1242, 1254, and 1260. Toxicol. Sci. 41, 62–76.[Abstract]

McLean, M. R., Bauer, U., Amaro, A. R., and Robertson, L. W. (1996a). Identification of catechol and hydroquinone metabolites of 4-monochlorobiphenyl. Chem. Res. Toxicol. 9, 158–164.[ISI][Medline]

McLean, M. R., Robertson, L. W., and Gupta, R. C. (1996b). Detection of PCB adducts by the 32P-postlabeling technique. Chem. Res. Toxicol. 9, 165–171.[ISI][Medline]

McLean, M. R., Twaroski, T. and Robertson, L. W. (2000). Redox cycling of 2-(x'-mono, -di,-trichlorophenyl)-1,4-benzoquinones, oxidation products of polychlorinated biphenyls. Arch. Biochem. Biophys. 376, 449–455.[ISI][Medline]

Milne, L., Nicotera, P., Orrenius, S., and Burkitt, M. J. (1993). Effects of glutathione and chelating agents on copper-mediated DNA oxidation: Pro-oxidant and antioxidant properties of glutathione. Arch. Biochem. Biophys. 304, 102–109.[ISI][Medline]

Monks, T. J., and Lau, S. S. (1997). Biological reactivity of polyphenolic-glutathione conjugates. Chem. Res. Toxicol. 10, 1296–1313.[ISI][Medline]

Moysich, K. B., Ambrosone, C. B., Vena, J. E., Shields, P. G., Mendola, P., Kostyniak, P., Greizerstein, H., Graham, S., Marshall, J. R., Schisterman, E. F., and Freudenheim, J. L. (1998). Environmental organochlorine exposure and postmenopausal breast cancer risk. Cancer Epidemiol. Biomarkers Prev. 7, 181–188.[Abstract]

Murata, M., Moriya, K., Inoue, S., and Kawanishi, S. (1999). Oxidative damage to cellular and isolated DNA by metabolites of the fungicide ortho-phenylphenol. Carcinogenesis 20, 851–857.[Abstract/Free Full Text]

Oakley, G. G., Robertson, L. W., and Gupta, R. C. (1996a). Analysis of polychlorinated biphenyl-DNA adducts by 32P-postlabeling. Carcinogenesis 17, 109–114.[Abstract]

Oakley, G. G., Devanaboyina, U., Robertson, L. W., and Gupta, R. C. (1996b). Oxidative DNA damage induced by activation of polychlorinated biphenyls (PCBs): Implications for PCB-induced oxidative stress in breast cancer. Chem. Res. Toxicol. 9, 1285–1292.[ISI][Medline]

Safe, S. H. (1994). Polychlorinated biphenyls (PCBs): Environmental impact, biochemical and toxic responses and implications for risk assessment. CRC Crit. Rev. Toxicol. 24, 87–149.

Shen, Y., Shen, H-M., Shi, C-Y., and Ong, C-N. (1996). Benzene metabolites enhance reactive oxygen species generation in HL 60 human leukemia cells. Hum. Exp. Toxicol. 15, 422–427.[ISI][Medline]

Silberhorn, E. M., Glauert, H. P., and Robertson, L. W. (1990). Carcinogenicity of polyhalogenated biphenyls: PCBs and PBBs. Crit. Rev. Toxicol. 20, 440–496.[Medline]

Sipes, I. G., and Schnellmann, R. G. (1987). Biotransformation of PCBs: Metabolic pathways and mechanisms. In Polychlorinated Biphenyls (PCBs): Mammalian and Environmental Toxicology, pp. 97–110. Springer-Verlag, Berlin.

Takabe, T., Miyakawa, M., and Nikai, S. (1978). The micellar acceleration of the reduction of nitroblue tetrazolium by superoxide anion radicals during the autoxidation of polyphenols. Bull. Chem. Soc. Jpn. 51, 321–322.[ISI]

Tanabe, S. (1988). PCB problems in the future: Foresight from current knowledge. Environ. Poll. 50, 5–28.[ISI][Medline]

Tsou, T. C., Chen, C. L., Liu, T. Y., and Yang, J. L. (1996). Induction of 8-hydroxydeoxyguanosine in DNA by chromium(III) plus hydrogen peroxide and its prevention by scavengers. Carcinogenesis 17, 103–108.[Abstract]

Yamamoto, K., Inoue, S., and Kawanishi, S. (1993). Site-specific DNA damage and 8-hydroxydeoxyguanosine formation by hydroxylamine and 4-hydroxyaminoquinoline 1-oxide in the presence of Cu(II): Role of active oxygen species. Carcinogenesis 14, 1397–1401.[Abstract]