Chem. Res. Toxicol. 1996, 9, 165-171
Detection of PCB Adducts by the Technique†
165
32P-Postlabeling
Mitch R. McLean,‡ Larry W. Robertson,‡ and Ramesh C. Gupta*,‡,§ The Graduate Center for Toxicology and Department of Preventive Medicine and Environmental Health, University of Kentucky, Lexington, Kentucky 40536-0305 Received May 15, 1995X
The purpose of this study was to determine whether lower chlorinated biphenyls would be bioactivated to electrophilic metabolites by microsomes alone or in combination with peroxidase. Monochloro- and dichlorobiphenyls were incubated with liver microsomes of rats treated with phenobarbital and β-naphthoflavone, an NADPH-regenerating system, and deoxyguanosine 3′-monophosphate (dGp). The resultant adducts were analyzed by 32P-postlabeling either following microsomal incubation alone (“preoxidized”) or coupled with subsequent oxidation with horseradish peroxidase/H2O2 (“oxidized”). The incubation of 4-monochlorobiphenyl (4MCB) resulted in the formation of two minor adducts by microsomal activation alone. However, the oxidized sample showed two additional major adducts. Formation of the latter adducts was almost completely (>80%) inhibited when the oxidation reaction was performed in the presence of ascorbic acid. The other test mono- and dichlorobiphenyls also formed 1-3 major adducts. Compared with microsomal activation alone, these adducts were enhanced after the oxidation reaction or detected only in the oxidized samples. These data suggest that (1) some adducts of the lower chlorinated biphenyls are derived from arene oxides and (2) many adducts may be formed by metabolism of the parent compounds to catechol and p-hydroquinone species, which are oxidized to semiquinones and/or quinones. The involvement of quinones and/or semiquinones was supported by UV/vis spectroscopic measurements, which showed that metabolites of 4-MCB can be oxidized to products with spectra characteristic of quinones. These data raise the possibility that lower chlorinated biphenyls may be genotoxic and may explain the fact that commercial polychlorinated biphenyl mixtures are complete rodent carcinogens.
Introduction (PCBs)1
Polychlorinated biphenyls are resistant to chemical and thermal decomposition (1), which has contributed to both their commercial utility and their persistence in the global ecosystem (2). Although the higher chlorinated congeners (4-10 chlorines) exhibit the greatest degree of resistance, many of the lower congeneric PCBs (1-3 chlorines) are present in environmental samples (3) and biota (2, 4), including humans (5-7). The metabolic transformation of PCBs is inversely related to the degree of chlorination (8). The higher chlorinated compounds are relatively resistant to metabolism, while the lower chlorinated compounds are, in general, hydroxylated in reactions catalyzed by cytochromes P450 1A and 2B (5, 9-11). The major products are the monohydroxy metabolites formed by the isomerization of an arene oxide intermediate or via direct † Portions of this work were presented at the 13th annual meeting of the Society of Environmental Toxicology and Chemistry, Cincinnati, OH. * Address correspondence to this author at the Graduate Center for Toxicology, 354 Health Sciences Research Building, University of Kentucky Medical Center, Lexington, KY 40536-0305. Telephone: (606) 257-2397. Fax: (606) 323-1059. E-mail: RCGUPT00@ UKCC.UKY.EDU. ‡ The Graduate Center for Toxicology. § Department of Preventive Medicine and Environmental Health. X Abstract published in Advance ACS Abstracts, December 1, 1995. 1 Abbreviations: PCB, polychlorinated biphenyl; MCB, monochlorobiphenyl; DCB, dichlorobiphenyl; OPP, o-phenylphenol; PHQ, 2-phenyl1,4-hydroquinone; PBQ, 2-phenyl-1,4-benzoquinone; PB, phenobarbital; β-NF, β-naphthoflavone; DMSO, dimethyl sulfoxide; HRP, horseradish peroxidase; dGp, deoxyguanosine 3′-monophosphate; NADP+, (oxidized) nicotinamide adenine dinucleotide phosphate; PEI, poly(ethylene imine); QH2, hydroquinone; Q, quinone; QH•-, semiquinone; PHS, prostaglandin H synthase.
0893-228x/96/2709-0165$12.00/0
insertion of hydroxy groups (12, 13). Although the arene oxide may be hydrated to a dihydrodiol by epoxide hydrolase, data suggest that 4-monochlorobiphenyl (4MCB) and several dichlorobiphenyls (DCB) undergo two successive hydroxylation steps to yield dihydroxy metabolites (10, 14-16). The binding of radiolabeled PCBs to microsomal proteins (17-23), RNA (17, 18), and DNA (17-19) frequently has been attributed to arene oxide intermediates. In contrast, Hesse et al. (16) presented evidence that arene oxide metabolites of 2,2′-dichlorobiphenyl were minor contributors to macromolecular binding. Their results led them to propose a role for semiquinones or quinones as the ultimate reactive agents. While this proposal apparently has received little attention, substantial literature has appeared that supports the involvement of quinoid metabolites in the genotoxicities of benzene (24) and o-phenylphenol (OPP) (25). Since PCBs, benzene, and OPP have similar structural features, the work with the latter two compounds strengthens the assertion by Hesse et al. (16), while also implicating quinoid metabolites in the genotoxicity of the lower chlorinated PCBs. Benzene has been epidemiologically linked to human leukemias (26). The toxicity of benzene is most likely due to the oxidation of its dihydroxy metabolites, i.e., catechol and p-hydroquinone, to semiquinones and quinones (27, 28). In rats, benzene adducts were not detected in the DNA isolated from liver, kidney, bone marrow, or mammary gland, but three minor adducts were detected in Zymbal gland DNA (29), which is the primary target tissue for benzene in rats (30). In mice, radioactively labeled benzene was covalently bound to © 1996 American Chemical Society
166 Chem. Res. Toxicol., Vol. 9, No. 1, 1996
macromolecules in liver, bone marrow, kidney, lung, spleen, blood, and muscle (31). The proportion of radioactivity bound to DNA was lowest in the liver (0.5%) and highest in the bone marrow (17%) (32). By using in vitro methods, benzoquinone, the two-electron oxidation product of p-hydroquinone, bound to deoxyguanosine and calf thymus DNA (33). p-Hydroquinone also formed DNA adducts in HL-60 cells, mouse bone marrow macrophages, and human bone marrow, as determined with the 32P-postlabeling technique (34). OPP (or 2-hydroxybiphenyl), a fungicide commonly used on citrus fruits, is metabolized to 2-phenyl-1,4hydroquinone (PHQ) (35). Kolachana et al. (36) demonstrated that prostaglandin H synthase (PHS), myeloperoxidase, and horseradish peroxidase could each oxidize PHQ to 2-phenyl-1,4-benzoquinone (PBQ) in the presence of the appropriate electron-accepting cosubstrate. Although intravesicularly injected PBQ induces bladder epithelial hyperplasia (37, 38), only PHQ resulted in DNA strand breaks during in vitro experiments (39). DNA cleavage by PHQ was preventable with superoxide dismutase, catalase, and oxygen-centered radical scavengers, suggesting that PHQ undergoes autoxidation in an aqueous medium leading to the production of superoxide and hydroxyl radicals. In addition, Grether et al. (40) have shown that PHQ binds to DNA in vitro. Evidently, the genotoxicities of OPP and benzene can be attributed to the hydroquinone metabolites of each of these compounds. Using rat liver microsomes, in the accompanying paper (41) we described the production of three dihydroxy metabolites of 4-MCB, i.e., 4-chloro-2′,3′/-3′,4′-/-2′,5′-dihydroxybiphenyl. We propose that these catechol and hydroquinone metabolites will either autoxidize or undergo enzymatic oxidation to produce electrophilic semiquinone and/or quinone species that bind to DNA bases. The purpose of this study was to detect adducts with lower chlorinated PCBs. Metabolism of 3and 4-MCBs and 2,2′-, 2,6-, and 3,4-DCBs led to products that react with 2′-deoxyguanosine 3′-monophosphate (dGp), both with and without an enzymatic oxidation system (horseradish peroxidase/H2O2), thereby suggesting that both arene oxides and quinoid metabolites may contribute to the genotoxicity of the lower chlorinated PCBs.
Materials and Methods 4-MCB was from EGA-Chemie (Steinheim/Albuch, Germany) and recrystallized from methanol to 99+% purity. 3-Chlorobiphenyl (99+%) and 2,2′- and 3,4-dichlorobiphenyl (each 99+%) were obtained from Riedel-De Haen AC (Hannover, Germany) and were used without further purification. 2,6-DCB was synthesized (42) and characterized as described (43). NADP+ (sodium salt), glucose 6-phosphate, glucose-6-phosphate dehydrogenase (EC 1.1.1.49, baker’s yeast), horseradish peroxidase (EC 1.11.1.7, type VI), catalase (EC 1.11.1.6, bovine liver), and inorganic reagents were purchased from Sigma Chemical Co. (St. Louis, MO). Sources of materials used in the 32P-postlabeling analysis have been described elsewhere (44, 45). Caution: PCBs should be handled as hazardous compounds in accordance with NIH guidelines (46). Microsomal Preparation. Male Sprague-Dawley rats were purchased from Harlan Sprague Dawley (Indianapolis, IN), provided with Rodent Chow (Purina, St. Louis, MO) and tap water ad libitum, and maintained on a 12 h light-dark cycle in a controlled environment at a temperature of 22 °C. After a 7 day quarantine period, the rats were injected ip on three consecutive days with 400 µmol of phenobarbital (PB, saline) per kilogram body weight and 100 µmol of β-naphthoflavone (β-
McLean et al. NF, corn oil) per kilogram body weight. The notation “PB/βNF” will be used to refer to the treatment of rats with both compounds. On the fourth day, each rat (200-220 g) was sacrificed by asphyxiation with carbon dioxide, and the liver was perfused in situ with ice-cold 0.25 M sucrose/0.1 mM EDTA (pH 7.3) and then collected. Microsomes were prepared as previously described (41). Protein was determined by the method of Lowry et al. (47) using bovine serum albumin as standard. The microsomal suspensions were frozen in aliquots at -80 °C until use. Microsomal Incubation. The incubation medium consisted of 50 mM sodium citrate buffer (pH 7.4), 2 mM MgCl2, 2 mg of microsomal protein/mL, and a NADPH-regenerating system of 5 mM glucose 6-phosphate, 0.5 mM NADP+, and 0.75 unit glucose-6-phosphate dehydrogenase/mL. After a 2 min preincubation, substrate, dissolved in dimethyl sulfoxide (DMSO), was added to a final concentration of 1 mM (0.5% DMSO). All incubations were at 37 °C in a shaking water bath. UV/vis Spectroscopy. Microsomal reaction mixtures (20 mL) were prepared and incubated at 37 °C for 1 h with and without 4-MCB. Each mixture was then extracted with ethyl acetate (2 × 10 mL) and dichloromethane (1 × 10 mL). After the solvent was reduced with a stream of N2, 200 µL of DMSO was added, and the residual ethyl acetate and dichloromethane were evaporated with gentle heating. Aliquots of the extracts in DMSO were then subjected to UV/vis spectroscopy as follows: A Shimadzu 2000 spectrophotometer was programmed to record the absorbance spectrum every min between 350 and 550 nm. The reference and sample cuvettes (1 mL) were maintained at 37 °C, and both contained (final concentrations) 50 mM sodium citrate buffer (pH 7.2) and 5% Triton X-100 for emulsifying the lipids. Aliquots (generally 10-20 µL) of the extracted material in DMSO were then added to the cuvettes; the reference received extract without 4-MCB metabolites. After H2O2 was added to 1 mM, a baseline absorption curve was determined and the oxidation was started by the addition of horseradish peroxidase (HRP) to the sample. Adduction of Deoxyguanosine 3′-Monophosphate (dGp) with 4-MCB and DCBs and Analysis by 32P-Postlabeling. The microsomal incubation (1 mL) was as described earlier, except that 40 µg of dGp was added. After 1 h, half of the incubation mixture was treated with 0.5 unit of HRP and 1 mM H2O2 for an additional 30 min. A 10 min treatment with catalase (10 units) was used to remove the remaining H2O2. Thus, one sample will be referred to as “preoxidized” and the other as “oxidized.” Unreacted metabolites and parent chlorinated biphenyls were extracted with ethyl acetate (2 × 1 vol, saturated with sodium citrate buffer, pH 7.4). For 32P-postlabeling analysis, adducts were enriched by extraction with water-saturated n-butanol and labeled with 32P using carrier-free [γ-32P]ATP (∼100 µCi, e2 µM, specific activity g 3000 Ci/mmol) and T4 polynucleotide kinase phosphorylation as described (48). Aliquots were spotted onto poly(ethylene imine)/cellulose sheets with a 5 cm wick (Whatman 17 Chr chromatography paper). The plate was first developed with 1.7 M sodium phosphate (pH 6) overnight (D1). The adducts were then separated by development in the same direction as D1 with 0.8 M lithium chloride/0.5 M Tris-HCl/7 M urea (pH 8) (D3). The chromatograms were then developed with 2-propanol/4 M ammonium hydroxide (1:1) (D4) perpendicular to D3. The developments in both D3 and D4 were 2 and 4 cm, respectively, onto a Whatman 3 mm paper wick. A final development with 1.7 M sodium phosphate (pH 6) in the D4 direction was given to further reduce background radioactivity. The adducts were visualized by intensifying screen-enhanced autoradiography at -80 °C.
Results UV/vis Spectroscopy of the HRP/H2O2-Oxidized MCB and DCB Metabolites. Absorption spectra of the HRP/H2O2-oxidized metabolites, produced by microsomemediated activation of 4-MCB, are shown in Figure 1;
PCB Adduct Formation
Chem. Res. Toxicol., Vol. 9, No. 1, 1996 167
Figure 3. 32P-Postlabeling maps demonstrating the effect of ascorbic acid on the formation of adducts between 4-monochlorobiphenyl metabolites and dGp. Experimental conditions were the same as described in legend to Figure 2, except that the preoxidation reaction was performed in the absence (B) and presence (C) of 5 mM ascorbic acid. (A) Control, DMSO only. The autoradiography exposure time was 44 h at -80 °C. Figure 1. Absorption spectra of oxidized metabolites of 4-monochlorobiphenyl (4-MCB). 4-MCB was incubated with β-NF-induced rat liver microsomes and a NADPH-regenerating system. An identical incubation without PCB was used as a control. After 1 h, the mixtures were extracted with ethyl acetate and methylene chloride and finally dissolved in DMSO. Aliquots of the extracted 4-MCB metabolites and the control were placed into the sample and reference cuvettes, respectively, containing 50 mM sodium citrate (pH 7.4), 5% Triton X-100, and 1 mM H2O2. After the baseline curve was recorded, 0.1 unit of horseradish peroxidase (HRP) was added to the sample, and the absorption was recorded every minute for 10 min. Additional HRP (0.4 unit) was then added, and the absorption was recorded every minute for an additional 10 min.
Figure 2. 32P-Postlabeling maps of adducts of 4-monochlorobiphenyl and 3,4-dichlorobiphenyl with dGp: (A) control, DMSO only; (B) 4-monochlorobiphenyl; (C) 3,4-dichlorobiphenyl. 4-Chlorobiphenyl (1 mM) was incubated with microsomes from PB/βNF-treated rats, a NADPH-regenerating system, and 40 µg of dGp. After 60 min, half of the mixture was oxidized with horseradish peroxidase and H2O2 for an additional 30 min. Following a 10 min treatment with catalase, the mixture was extracted two times with ethyl acetate to remove unreacted 4-MCB and metabolites. Adducts were enriched with the butanol procedure and then labeled with 32P using carrier-free [γ-32P]ATP and T4 polynucleotide kinase as described (48). Chromatography was on PEI/cellulose sheets: D1, 1.7 M sodium phosphate (pH 6); D3, 0.8 M lithium chloride/0.5 M Tris-HCl/7 M urea (pH 8); D4, 2-propanol/4 M ammonium hydroxide (1:1). The autoradiography exposure time was 12 h at -80 °C.
the nonoxidized metabolites exhibited an absorption maximum at about 285 nm (not shown) and otherwise did not absorb in the visible region (350-700 nm). After
the baseline absorption was recorded (time zero), 0.1 unit of HRP was added to the sample, and the absorption spectrum was recorded every minute for 10 min. An absorption maximum was observed at 400 nm (Figure 1). Although HRP reportedly absorbs at about 402 nm (49), the concentrations of HRP used in these studies did not produce an observable absorption peak (not shown). However, by using a 20-fold excess of HRP a small peak was observed at about 417 nm (not shown), which corresponded to the expected oxidation of HRP to compound II (49). The rate of production of the oxidizable species seemed to increase after the addition of additional HRP (0.4 unit) (Figure 1, top 10 curves). In all experiments, the absorption increased over time until reaching a maximum, followed by a comparatively rapid drop (not shown). Neither the addition of H2O2 or HRP nor increases or decreases in the amount of extracted metabolite prevented the drop in absorption. In all cases, oxidation was dependent upon both HRP and H2O2 since use of either alone did not result in the appearance of absorption peaks from 350 to 700 nm. Binding of MCBs and DCBs to dGp. The conversion of MCBs and DCBs to electrophilic metabolites was measured by the addition of dGp to trap the electrophilic metabolites, followed by analysis of the resultant adducts by 32P-postlabeling. Butanol enrichment (48) instead of nuclease P1 (50) enrichment was used throughout this study, since the former method resulted in higher adduct recoveries, particularly for MCBs. One major adduct and one minor adduct were detected after the incubation of 4-MCB with microsomes prepared from rats treated with PB/β-NF (Figure 2B, adducts 1 and 2). When half of the mixture was oxidized, two additional adducts (Figure 2E, adducts 3 and 4) were found. Controls lacking 4-MCB showed one spot (Figure 2A,D, not numbered) as a contaminant. Inclusion of 5 mM ascorbic acid, a substrate for the HRP/H2O2 oxidation system, in the incubation mixture substantially (>80%) reduced adducts 3 and 4 (Figure 3), but had no effect on the formation of adducts 1 and 2 (Figure 3). Incubation of 3,4-DCB with dGp and liver microsomes from rats treated with PB/β-NF resulted in the formation of three major and several minor adducts. Most of the adducts were found under both preoxidation (Figure 2C) and oxidation conditions (Figure 2F), although they were much more prominent after the oxidation reaction.
168 Chem. Res. Toxicol., Vol. 9, No. 1, 1996
McLean et al. Table 1. Summary of Mono- and Dichlorobiphenyl Adducts Formed with Deoxyguanosine 3′-Monophosphate Following Microsome-Mediated Activation Alone (Preoxidized) or Followed by HRP/H2O2 Oxidation (Postoxidation) number of adducts detecteda parent biphenyl compoundb
Figure 4. 32P-Postlabeling maps showing the formation of adducts between dGp and the horseradish peroxidase/H2O2oxidizable metabolites of 3-monochlorobiphenyl (B), 2,2′-dichlorobiphenyl (C), and 2,6-dichlorobiphenyl (D). (A) Control, DMSO only. For experimental conditions, consult the legend to Figure 2. The autoradiography exposure time was 16 h at -80 °C.
MCBs 234DCBs, same ring 2,32,42,52,63,43,5DCBs, both rings 2,2′2,3′2,4′3,3′3,4′4,4′-
pre
major post
minor pre post
total
-c 1
1 3 2
1
1 -
2 3 4
4d 3d
2 6 3
-
1 1 2 2 -
1 1 2 2 8 3
-
1 1 -
2 1 2
1 3 1 2
-
a
Major adducts were defined as a spot intensity representing >10% of the total adducts detected. Minor adducts had a spot intensity of 2-10%. b MCBs, monochlorobiphenyls; DCBs, dichlorobiphenyls. c -, adduct not detectable. d Adducts found under both preoxidized and postoxidized conditions, although much enhanced by HRP/H2O2 oxidation, suggesting autoxidation of the putative hydroquinone metabolites prior to enzyme-mediated oxidation.
The formation of PCB-dGp adducts in the presence of HRP/H2O2 was also demonstrated by incubating dGp with several other mono- or dichlorobiphenyls: 3-MCB, 2,2′-DCB, and 2,6-DCB. Each of these compounds resulted in three, one, and two major adducts, respectively (Figure 4B-D); the control showed no such spots (Figure 4A). No major spots were detected in the preoxidized incubations (not shown). 2-MCB and the other possible DCBs were also tested (Table 1). No adducts were detected with 2-MCB under preoxidation conditions, but two adducts (one major and one minor) were found when the mixture was oxidized with HRP/H2O2. 2,3-, 2,4-, and 2,5-DCB also formed adducts, but only after the oxidation reaction (Table 1). With the exception of 2,2′- and 3,3′-DCB, only minor adducts were formed with test compounds containing a chlorine substituent in each ring. 3,3′-, 3,4′-, and 4,4′DCB produced three, one, and two adducts, respectively; all adducts were detected only after oxidation with HRP/ H2O2 (Table 1). 2,3′- and 2,4′-DCB did not form adducts under either preoxidized or oxidized conditions (Table 1).
experiments the black precipitate eventually became apparent, which qualitatively seemed to coincide with the gradual drop in absorption following the peak. Since neither the concentration nor the identities of the oxidizable metabolites were known, quantitative studies were not possible. The mechanism of HRP-catalyzed oxidation, as well as those of peroxidases in general, has attracted considerable interest for over 40 years (49). Native HRP undergoes a two-electron oxidation with H2O2 serving as the electron acceptor. The products (eqs 1-3) of this reaction are water and compound I. Compound I is a strong oxidant that is reduced in two, one-electron steps with compound II (eq 2) as an intermediate (51, 52). By using a hydroquinone (QH2) as a model substrate, the reaction proceeds as follows:
Discussion
HRP + H2O2 f compound I + H2O
(1)
Formation of Quinones from Metabolites of 4-MCB and 3,4-DCB. Catechols and hydroquinones absorb in the UV region of the electromagnetic spectrum. Subsequent chemical or enzymatic oxidation to quinones produces an absorption peak in the visible region, which is readily apparent as the solution changes color. By using UV/vis spectroscopy, we were able to demonstrate that the microsome-mediated metabolites of 4-MCB were readily oxidized with HRP using H2O2 as a cosubstrate (Figure 1). Oxidation with more than 0.5 unit of HRP/ mL was very rapid, as judged by the almost immediate color change (to dark yellow, which was reversible with sodium dithionite) and the occurrence of a black precipitate, which was likely due to coupling products (see the following). However, by keeping the concentration of HRP low, the time-dependent oxidation of 4-MCB metabolites was observed. Nevertheless, in all of these
compound I + QH2 f compound II + QH•-
(2)
compound II + QH2 f HRP + QH•- + H2O (3) The quinone (Q) arises from the disproportionation of two semiquinones (QH•-):
QH•- + QH•- f QH2 + Q
(4)
The presence of a reactive semiquinone in peroxidasemediated oxidation greatly increases the probability for binding to macromolecular targets. In fact, observations of the black precipitate suggested free radical coupling of the semiquinones to yield polymeric pigments reminiscent of melanin (53). Therefore, binding to a nucleophile may occur by reaction with (1) a semiquinone, (2)
PCB Adduct Formation
a 1,4-Michael addition of the quinone, or (3) a Schiff base (1,2-addition) formed between an amino group and the quinone. Identification of dGp Adducts with MCBs and DCBs. 32P-Postlabeling was used to detect the adducts between dGp and metabolites of MCBs and DCBs. The 32P-postlabeling technique is both very sensitive, being able to detect one adduct in 1010 nucleotides, and does not require the use of a radiolabeled parent compound (44, 48). In addition, by using multidirectional TLC, individual adducts can be separated on the same plate by using two or more solvent systems. 4-MCB was bioactivated with a microsomal system to metabolites that bound to exogenously added dGp. 4-MCB gave two adducts with dGp that were formed during microsome-mediated metabolism. These adducts were interpreted to be due to arene oxides, which are intermediates formed during the metabolism of 4-MCB (17, 18). Oxidation of the mixture with HRP/H2O2 resulted in the formation of two additional adducts. It was previously reported that [3H]-4-MCB binds to microsomal protein, rRNA, and DNA (17). In addition, poly(guanylic acid) was the best nucleophilic trapping agent for reactive metabolites of [3H]-4-MCB, followed, in decreasing order, by poly(adenylic acid), denatured calf thymus DNA, and the poly(pyrimidine)s (18). In the accompanying paper (41), the metabolism of 4-MCB by liver microsomes from rats treated with both phenobarbital and 3-methylcholanthrene yielded three diol metabolites: 2′,3′-, 2′,5′-, and 3′,4′-diol. Each of these metabolites could presumably be oxidized to a semiquinone and/or quinone by the HRP/H2O2 oxidation system and, therefore, form adducts with dGp. Which metabolite(s) forms the adducts is not known, but as indicated in the Introduction, 4-chloro2′,5′-dihydroxybiphenyl is structurally similar to the proposed genotoxic metabolites of benzene (24) and OPP (25). Although the metabolism of 2- and 3-MCB has not been studied extensively, the MCBs appear to have similar pathways (9), i.e., the formation of monol and diol products. No adducts were detected when 2- and 3-MCB were incubated with hepatic microsomes and dGp alone. However, the adducts formed from the oxidizable metabolites of each compound exhibited different chromatographic characteristics. Adducts were also formed between dGp and several dichlorobiphenyls. Kennedy et al. (10) reported that DCBs are metabolized primarily to monols, with lesser amounts of dihydrodiols and, presumably, diol products. Under their experimental conditions the diols were found to be unstable, which suggested that they may undergo nonenzymatic oxidation. All six of the DCBs (2,3-, 2,4-, 2,5-, 2,6-, 3,4-, and 3,5-DCB) with both chlorine substituents in the same ring produced adducts with dGp under oxidation conditions. The largest number of adducts were produced from metabolites of 3,4- and 3,5-DCB; these were also the only DCBs that gave adducts under preoxidized conditions. Since the preoxidized and oxidized adducts of 3,4- (Figure 2C,F) and 3,5-DCB (not shown) were chromatographically similar, the preoxidized adducts may have been derived from the autoxidation of hydroquinone metabolites instead of arene oxides (see the following). With the exception of 2,2′-DCB, DCBs with chlorine substituents in each ring formed no adducts (2,3′ and 2,4′) or only minor adducts (3,3′, 3,4′, and 4,4′) when subjected to HRP/H2O2 oxidation. The reason for this is not known
Chem. Res. Toxicol., Vol. 9, No. 1, 1996 169
since Kennedy et al. (10) showed that these compounds are metabolized by rat hepatic microsomes to monol and, presumably, diol products. Nevertheless, 2,2′-DCB did form one major adduct when the reaction mixture was supplemented with HRP/H2O2. Hesse et al. (16) have shown that [14C]-2,2′-DCB binds to rat liver microsomal protein in a reaction that was inhibited by superoxide dismutase, ascorbic acid, catechol, epinephrine, reduced glutathione, and UDP-glucuronic acid. These results were strong evidence for the involvement of quinones and/or semiquinones in the protein binding (16) and support our hypothesis for the involvement of quinones and/or semiquinones in the binding of metabolites of lower chlorinated PCBs to DNA bases. A unique component of our experimental method was the separation of the adducts into “preoxidized” and “oxidized.” At this time we do not know the nature of the reactive metabolite species, i.e., whether semiquinone, quinone, or both; however, binding was HRPmediated since ascorbic acid greatly diminished the binding of oxidizable metabolites of 4-MCB, but not the putative arene oxide adducts. Ascorbic acid is a substrate for HRP-catalyzed oxidation (54) and therefore may competitively inhibit oxidation of the hydroquinone metabolites or may deplete H2O2 by promoting the HRP/ H2O2 catalytic cycle. Ascorbic acid may also reduce the quinone back to a hydroquinone (not shown), thereby preventing the formation of Michael addition products or a Schiff base. Although inhibition of adduct formation by ascorbic acid does not confirm the mechanism of binding, these data do show that oxidation products are involved in the binding to dGp. In fact, ascorbic acid may react with the electrophilic oxidation product(s) to form an inactive covalent adduct, as has been recently described for 6-(sulfoxymethyl)benzo[a]pyrene (55), which has been proposed as the ultimate carcinogenic agent of 6-(hydroxymethyl)benzo[a]pyrene. We used HRP as the oxidizing enzyme, but oxidases with an identical or similar mechanism are present in most mammalian tissues (56). For example, prostaglandin H synthase (PHS) bioactivates many aromatic amines and phenols via a one-electron oxidation (57). Therefore, it is conceivable that hydroquinone metabolites of PCBs are oxidized by PHS to generate semiquinones and perhaps quinones. Since PHS is a constituent of microsomal membranes, it is possible that the formation of the “oxidizable-like” adducts during the preoxidation stage of the 3,4-DCB incubation may be mediated by this enzyme. Additional work is needed to support this possibility since the hydroquinones may also autoxidize to the semiquinone and/or quinone, which then bind to the target nucleotide. Genotoxicity and Carcinogenicity. Numerous studies have demonstrated that commercial PCB formulations, which are complex mixtures of individual isomers and congeners, are rodent carcinogens (reviewed in ref 58). Commercial PCBs induce preneoplastic lesions, neoplastic nodules, and hepatocellular carcinomas in animals. PCBs therefore may be considered complete carcinogens. PCB mixtures and individual isomers and congeners are efficacious tumor promoters in both rats and mice when given for extended periods of time after an initiating agent. Generally, the higher halogenated congeners, especially those that induce cytochrome P450, are efficacious promoters in two-stage hepatocarcinogenesis. The traditional view has been that PCB mixtures, as well as individual PCB congeners, are not
170 Chem. Res. Toxicol., Vol. 9, No. 1, 1996
mutagenic or genotoxic, although there are several findings to the contrary (58). Generally, Ames assays, as well as most short-term tests for genotoxicity, are not optimized for the relatively slow rate of metabolism of individual PCB congeners. In addition, few of the 209 possible PCBs have been tested. The present study provides, for the first time, unequivocal evidence that individual PCBs may be activated (via at least two mechanisms) to electrophiles that bind to DNA bases. We also found that dG in calf thymus DNA is the predominant target for these electrophiles.2 These data raise the possibility that the individual PCB isomers and congeners may be initiators and promoters of hepatocarcinogenesis and may explain the fact that commercial PCB formulations are complete animal carcinogens. Conclusions. We have provided evidence for the formation of adducts between dGp and monochlorobiphenyls and dichlorobiphenyls. Most of the adducts were attributable to the oxidation of secondary metabolism products and much less so to arene oxides. The putative diol metabolites may participate in binding to nucleic acids following oxidation to the semiquinone, quinone, or both.
Acknowledgment. This work was supported by Grant CA 57423 from the National Institutes of Health. The authors acknowledge many useful discussions with Drs. Achal Garg and Kevin Stansbury. Kevin Thomas is acknowledged for technical assistance.
References (1) Rappe, C., and Buser, H. R. (1980) Chemical properties and analytical methods. In Halogenated biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products (Kimbrough, R. D., Ed.) pp 41-76, Elsevier/North-Holland Biomedical Press, Amsterdam. (2) Risebrough, R. W., Rieche, P., Herman, S. G., Peakall, D. B., and Kirven, M. N. (1968) Polychlorinated biphenyls in the global ecosystem. Nature 220, 1098-1102. (3) Brown, J. F., Bedard, D. L., Brennan, M. J., Carnahan, J. C., Feng, H., and Wagner, R. E. (1987) Polychlorinated biphenyl dechlorination in aquatic sediments. Science 221, 709-712. (4) Buckley, E. H. (1982) Accumulation of airborne polychlorinated biphenyls in foliage. Science 216, 520-522. (5) Safe, S. (1984) Polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs): biochemistry, toxicology, and mechanism of action. CRC Crit. Rev. Toxicol. 13, 319-395. (6) Jacobson, J. L., Humphrey, H. E. B., Jacobson, S. W., Schantz, S. L., Mullin, M. D., and Welch, R. (1989) Determinants of polychlorinated biphenyls (PCBs), polybrominated biphenyls (PBBs), and dichlorodiphenyl trichloroethane (DTT) levels in the sera of young children. Am. J. Pub. Health 79, 1401-1404. (7) Falck, F., Ricci, A., Wolff, M. S., Godbold, J., and Deckers, P. (1992) Pesticides and polychlorinated biphenyl residues in human breast lipids and their relation to breast cancer. Arch. Environ. Health 47, 143-146. (8) Sundstro¨m, G., and Hutzinger, O. (1976) The metabolism of chlorobiphenylssa review. Chemosphere 5, 267-298. (9) Kennedy, M. W., Carpentier, N. K., Dymerski, P. P., Adams, S. M., and Kaminsky, L. S. (1980) Metabolism of monochlorobiphenyls by hepatic microsomal cytochrome P-450. Biochem. Pharmacol. 29, 727-736. (10) Kennedy, M. W., Carpentier, N. K., Dymerski, P. P., and Kaminsky, L. S. (1981) Metabolism of dichlorobiphenyls by hepatic microsomal cytochrome P-450. Biochem. Pharmacol. 30, 577-588. (11) Kaminsky, L. S., Kennedy, M. W., Adams, S. M., and Guengerich, F. P. (1981) Metabolism of dichlorobiphenyls by highly purified isozymes of rat liver cytochrome P-450. Biochemistry 20, 73797384. (12) Jerina, D. M., and Daly, J. W. (1974) Arene oxides: a new aspect of drug metabolism. Science 185, 573-582. 2 G. Oakley, L. W. Robertson, and R. C. Gupta, Carcinogenesis, in press, 1996.
McLean et al. (13) Tomaszewski, J. E., Jerina, D. M., and Daly, J. W. (1975) Deuterium isotope effects during formation of phenols by hepatic monoxygenases. Evidence for an alternative to the arene oxide pathway. Biochemistry 14, 2024-20331. (14) Safe, S., Hutzinger, O., and Jones, D. (1975) The mechanism of chlorobiphenyl metabolism. J. Agric. Food Chem. 23, 851-853. (15) Greb, W., Klein, W., Coulston, F., Golberg, L., and Korte, F. (1975) In vitro metabolism of polychlorinated biphenyls-14C. Bull. Environ. Contam. Toxicol. 13, 424-432. (16) Hesse, S., Mezger, M., and Wolff, T. (1978) Activation of [14C]chlorobiphenyls to protein-binding metabolites by rat liver microsomes. Chem.-Biol. Interact. 20, 355-365. (17) Wong, A., Basrur, P., and Safe, S. (1979) The metabolically mediated DNA damage and subsequent DNA repair by 4-chlorobiphenyl in Chinese hamster ovary cells. Res. Commun. Chem. Pathol. Pharmacol. 24, 543-550. (18) Wyndam, C., and Safe, S. (1978) In vitro metabolism of 4-chlorobiphenyl by control and induced rat liver microsomes. Biochemistry 17, 208-215. (19) Narbonne, J. F., and Daubeze, M. (1980) In vitro binding of hexachlorobiphenyl to DNA and proteins. Toxicology 16, 173175. (20) Shimada, T., and Sato, R. (1978) Covalent binding in vitro of polychlorinated biphenyls to microsomal macromolecules. Biochem. Pharmacol. 27, 585-593. (21) Shimada, T. (1976) Metabolic activation of [14C]polychlorinated biphenyl mixtures by rat liver microsomes. Bull. Environ. Contam. Toxicol. 16, 25-32. (22) Shimada, T., Imai, Y., and Sato, R. (1981) Covalent binding of polychlorinated biphenyls to proteins by reconstituted monooxygenase system containing cytochrome P-450. Chem.-Biol. Interact. 38, 29-44. (23) Shimada, T., and Sato, R. (1980) Covalent binding of polychlorinated biphenyls to rat liver microsomes in vitro: nature of reactive metabolites and target macromolecules. Toxicol. Appl. Pharmacol. 55, 490-500. (24) Snyder, R., Witz, G., and Goldstein, B. D. (1993) The toxicology of benzene. Environ. Health Perspect. 100, 293-306. (25) Hiraga, K., and Fujii, T. (1981) Induction of tumors of the urinary system in F344 rats by dietary administration of sodium ophenylphenate. Food Cosmet. Toxicol. 19, 303-310. (26) International Agency for Research on Cancer (1982) Some industrial chemicals and dyestuffs. In Evaluation of the Carcinogenic Risk of Chemicals to Humans, Vol. 29, pp 93-148, IARC, Lyon, France. (27) Smith, M. T., Yager, J. W., Steinmetz, K. L., and Eastmond, D. A. (1989) Peroxidase-dependent metabolism of benzene’s phenolic metabolites and its potential role in benzene toxicity and carcinogenicity. Environ. Health Perspect. 82, 23-29. (28) Sadler, A., Subrahmanyam, V. V., and Ross, D. (1988) Oxidation of catechol by horseradish peroxidase and human leukocyte peroxidase: reactions of o-benzoquinone and o-benzosemiquinone. Toxicol. Appl. Pharmacol. 93, 62-71. (29) Reddy, M. V., Blackburn, G. R., Schreiner, C. A., Mehlman, M. A., and Mackerer, C. R. (1989) 32P analysis of DNA adducts in tissues of benzene-treated rats. Environ. Health Perspect. 82, 253-257. (30) Low, L. K., Lambert, C. E., Meeks, J. R., Naro, P. A., and Mackerer, C. R. (1995) Tissue-specific metabolism of benzene in Zymbal gland and other solid tumor target tissues in rats. J. Am. Coll. Toxicol. 14, 40-60. (31) Snyder, R., Lee, E. W., and Kocsis, J. J. (1978) Binding of labeled benzene metabolites to mouse liver and bone marrow. Res. Commun. Chem. Pathol. Pharmacol. 20, 191-194. (32) Gill, D. P., and Ahmed, A. E. (1981) Covalent binding of 14Cbenzene to cellular organelles and bone marrow nucleic acid. Biochem. Pharmacol. 30, 1127-1131. (33) Jowa, L., Witz, G., Snyder, R., Winkle, S., and Kalf, G. F. (1990) Synthesis and characterization of deoxyguanosine-benzoquinone adducts. J. Appl. Toxicol. 10, 47-54. (34) Le´vay, G., Ross, D., and Bodell, W. J. (1993) Peroxidase activation of hydroquinone results in the formation of DNA adducts in HL60 cells, mouse bone marrow macrophages and human bone marrow. Carcinogenesis 14, 2329-2334. (35) Nakao, T., Ushiyama, K., Kabashima, J., Nagai, F., Hakagawa, A., Ohno, T., Ichikawa, H., Kobayashi, H., and Hiraga, K. (1983) The metabolic profile of sodium o-phenylphenate after subchronic oral administration to rats. Food Chem. Toxicol. 21, 325-329. (36) Kolachana, P., Subrahmanyam, V. V., Eastmond, D. A., and Smith M. T. (1991) Metabolism of phenylhydroquinone by prostaglandin (H) synthase: possible implications in o-phenylphenol carcinogenesis. Carcinogenesis 12, 145-149. (37) Morimoto, K., Fukuoka, M., Hasegawa, R., Tanaka, A., Takahashi, A., and Hayashi, Y. (1987) DNA damage in urinary bladder epithelium of male F344 rats treated with 2-phenyl-1,4-benzo-
PCB Adduct Formation
(38)
(39) (40) (41) (42)
(43)
(44) (45)
(46)
quinone, one of the non-conjugated urinary metabolites of sodium o-phenylphenate. Jpn. J. Cancer Res. (GANN) 78, 1027-1030. Morimoto, K., Sato, M., Fukuoka, M., Hasegawa, R., Takahashi, T., Tsuchiya, T., Tanaka, A., Takahashi, A., and Hayashi, Y. (1989) Correlation between the DNA damage in urinary bladder epithelium and the urinary 2-phenyl-1,4-benzoquinone levels from F344 rats fed sodium o-phenylphenate in the diet. Carcinogenesis 10, 1823-1827. Nagai, F., Ushiyama, K., Satoh, K., and Kano, I. (1990) DNA cleavage by phenylhydroquinone: the major metabolite of a fungicide o-phenylphenol. Chem.-Biol. Interact. 76, 163-179. Grether, T., Brunn, H., and Laib, R. J. (1989) 32P-postlabeling method as a sensitive indicator for analysis of genotoxicity of biphenyl derivatives. Arch. Toxicol. 63, 423-424. McLean, M. R., Bauer, U., Amaro, A. R., and Robertson, L. W. (1996) Identification of catechol and hydroquinone metabolites of 4-monochlorobiphenyl. Chem. Res. Toxicol. 9, 158-164. Anklam, E., and Asmus, K.-D. (1989) Radical cations of polychlorinated and polybrominated biphenyls in 1,2-dichlorethane: a pulse radiolysis study. J. Chem. Soc., Perkin Trans. 2, 15731576. Ho¨fler, F., Melzer, H., Mo¨ckel, J., Robertson, L. W., and Anklam, E. (1988) Relationship between liquid and gas chromatographic retention behavior and calculated molecular surface area of selected polyhalogenated biphenyls. J. Agric. Food Chem. 36, 961-965. Gupta, R. C., Reddy, M. V., and Randerrath, K. (1982) 32Ppostlabeling analysis of non-radioactive aromatic carcinogen-DNA adducts. Carcinogenesis 3, 1081-1092. Gupta, R. C. (1993) 32P-postlabeling analysis of bulky aromatic adducts. In Postlabeling Methods for Detection of DNA Adducts (Phillips, D. H., Castegnaro, M., and Bartach, H., Eds.) IARC, Lyon, France. NIH Guidelines for the Laboratory Use of Chemical Carcinogens (1981) NIH Publication No. 81-2385, U.S. Government Printing Office, Washington, D.C.
Chem. Res. Toxicol., Vol. 9, No. 1, 1996 171 (47) Lowry, O. H., Rosenbourgh, N. J., Farr, A. C., and Randall, R. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. (48) Gupta, R. C. (1985) Enhanced sensitivity of 32P-postlabeling analysis of aromatic carcinogen: DNA adducts. Cancer Res. 45, 5656-5662. (49) George, P. (1953) The chemical nature of the second hydrogen peroxide compound formed by cytochrome c peroxidase and horseradish peroxidase. Biochem. J. 54, 267-276. (50) Reddy, M. V., and Randerrath, K. (1986) Nuclease P1-mediated enhancement of sensitivity of 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis 7, 1543-1551. (51) Wood, P. M. (1992) A comparison of peroxidase and cytochrome P-450. Biochem. Soc. Trans. 20, 349-352. (52) Hollenberg, P. F. (1992) Mechanisms of cytochrome P450 and peroxidase-catalyzed xenobiotic metabolism. FASEB 6, 686-694. (53) Graham, D. G., and Jeffs, P. W. (1977) The role of 2,4,5trihydroxyphenylalanine in melanin biosynthesis. J. Biol. Chem. 252, 5729-5734. (54) Ator, M. A., David, S. K., and Ortiz de Montellano, P. R. (1987) Structure and catalytic mechanism of horseradish peroxidase. J. Biol. Chem. 262, 14954-14960. (55) Surh, Y. J., Park, K. K., and Miller, J. A. (1994) Inhibitory effect of vitamin C on the mutagenicity and covalent DNA binding of the electrophilic and carcinogenic metabolite, 6-sulfooxymethylbenzo[a]pyrene. Carcinogenesis 15, 917-920. (56) Pedersen, J. Z., and Finazzi-Agro`, A. (1993) Protein-radical enzymes. FEBS 325, 53-58. (57) Degen, G. H. (1993) Prostaglandin-H synthase containing cell lines as tools for studying metabolism and toxicity of xenobiotics. Toxicology 82, 243-256. (58) Silberhorn, E. M., Glauert, H. P., and Robertson, L. W. (1990) Carcinogenicity of polyhalogenated biphenyls: PCBs and PBBs. CRC Crit. Rev. Toxicol. 20, 439-496.
TX9500843
