Mutagenic activation of arylamines by subcellular fractions ofChamaelea gallina clams exposed to environmental pollutants

Environmental and Molecular Mutagenesis 41:55– 63 (2003)

Mutagenic Activation of Arylamines by Subcellular Fractions of Chamaelea gallina Clams Exposed to Environmental Pollutants Manuel Jose ´ Rodrı´guez-Ortega,1 Antonio Rodrı´guez-Ariza,1 Oscar Amezcua,2 and Juan Lo ´ pez-Barea1* 1

Department of Biochemistry and Molecular Biology, University of Co´rdoba, Co´rdoba, Spain 2 CICEM “El Torun ˜ o,” Ca´diz, Spain

Biochemical measurements in the sentinel clam Chamaelea gallina have been used as biomarkers of marine pollution. In this study, S9, cytosolic fractions (CF), and microsomal fractions (MF) prepared from unexposed clams and clams exposed to model pollutants were used to activate 2–aminoanthracene (2-AA) and 2–acetylaminofluorene (AAF) to mutagens in Salmonella typhimurium strain BA149, which overexpresses O-acetyltransferase. Arylamine activation was similar with subcellular fractions from unexposed and Aroclor 1254-exposed clams, but decreased with fractions from As(III)- and Cu(II)-exposed clams. Bioactivation of arylamines by CF was higher than by MF, and higher with NADH than with NADPH as the reducing agent. ␣-Naphthoflavone inhibited AAF activation by CF and MF, but increased 2-AA activation nearly twofold. In contrast to the results with arylamine activation, benzo[a]pyrene hydrox-

ylase (BPH) activity increased twofold in fractions from Aroclor 1254- and Cu(II)-exposed clams. Activation of 2-AA was also evaluated using S9 fractions from clams sampled at littoral sites with different pollutant levels. Compared to a reference site, lower 2-AA bioactivation and higher BPH activity were found in clams containing high levels of copper and organic contaminants, although the differences were not statistically significant. While these findings agree with the results of the model Cu(II) exposure, the effects of other pollutants cannot be ruled out. The results of the study demonstrate that arylamine activation by clams is not preferentially catalyzed by microsomal monooxygenases but by unknown cytosolic system(s), and that bioactivation of 2–AA and AAF appears to occur by different pathways. Environ. Mol. Mutagen. 41:55– 63, 2003. © 2003 Wiley-Liss, Inc.

Key words: mollusks; 2-aminoanthracene; acetylaminofluorene; pollution biomarkers; monooxygenases; promutagens

INTRODUCTION Effective monitoring of environmental pollution involves an integrated assessment of selected contaminants and the biological responses of living organisms. Bivalve mollusks are used extensively as sentinel organisms for monitoring marine ecosystems due to their ability to bioaccumulate contaminants, their sessile nature, and widespread distribution [Ringwood et al., 1998]. Several biochemical parameters, also known as biomarkers, are used to assess the responses of marine organisms to contaminants [Peakall, 1994; Ringwood et al., 1998]. Fourteen biomarkers have been recently evaluated in the clam Chamaelea gallina from southern Spanish sites with different pollutant burdens. They included the activities of several primary or ancillary antioxidant enzymes and glutathione-related enzymes and the levels of oxidative damage to lipids and reduced glutathione (GSH) [Rodrı´guez–Ortega et al., 2002]. The induction of biotransformation enzymes has been used as a biomarker in monitoring programs [Ringwood et al., 1998]. Bivalves have relatively low activities of micro© 2003 Wiley-Liss, Inc.

somal biotransforming enzymes, and these enzymes have characteristics that are quite different from their vertebrate homologs [Livingstone et al., 1989]. Recent studies with bivalve mollusks report the existence of a CYP1A–like protein, inducible by organic contaminants such as polyaromatic hydrocarbons (PAHs) and polychlorinated biphenyls

Grant sponsor: Ministerio de Educacio´n y Cultura, Spain; Grant number: project 1FD97-0610, Plan Andaluz de Investigacio´n, Consejerı´a de Educacio´n y Ciencia, Junta de Andalucı´a, Spain (group CVI-151); Grant sponsor: Consejerı´a de Agricultura y Pesca, Junta de Andalucı´a, Spain. *Correspondence to: Juan Lo´pez-Barea, Department of Biochemistry and Molecular Biology, University of Co´rdoba, Campus de Rabanales, Severo Ochoa Building, 2nd Floor, 14071 Co´rdoba, Spain. E-mail: bb1lobaj@ uco.es Received 23 March 2002; provisionally accepted 17 May 2002; and in final form 19 October 2002 DOI 10.1002/em.10130

56

Rodrı´guez-Ortega et al.

(PCBs) [Peters et al., 1998]. Biochemical biomarkers, including the capacity to activate or detoxify some mutagens, have been used as sensitive indicators of exposure to pollutants in the marine environment [Bagnasco et al., 1991; Rodrı´guez–Ariza et al., 1994, 1995; Dı´az–Me´ndez et al., 1998]. The study of this activation capability may bring new insights to our understanding of the fate and the effects of carcinogens in the marine environment [Britvic and Kurelec, 1986]. N–Hydroxylation, the key biochemical step in arylamine activation, is catalyzed by microsomal enzymes, such as cytochrome P450-dependent monooxygenases [Steele and Ioannides, 1986] and FAD-dependent monooxygenases [Pelroy and Gandolfi, 1980]. The resulting N-hydroxylamines can undergo cytosolic N-acetylation and N,Oacetyltransfer reactions to generate the ultimate mutagen(s) [Yamazoe et al., 1989]. Several studies of mammalian biotransformation have suggested that a cytosolic system can also activate arylamines to genotoxins [Forster et al., 1981; Traynor et al., 1991; Leist et al., 1992; Marczylo and Ioannides, 1994, 1997, 1998]. Studying the activation of arylamines to bacterial mutagens by S9 fractions from bivalve mollusks was initially hindered by their low biotransformation capability [Michel et al., 1993]. The use of a Salmonella typhimurium strain overexpressing O-acetyltransferase [Jurado et al., 1994] facilitated the study of the mutagenic activation of arylamines by S9 fractions of bivalve mollusks and a cytosolic location was suggested for the responsible enzymatic system [Dı´az-Me´ndez et al., 1998]. In order to evaluate the effects of pollutants on the activation of arylamines to mutagens, we exposed Chamaelea gallina clams to four model pollutants, Aroclor 1254, Cu(II), tributyl-Sn, and arsenic (III) oxide. Subcellular fractions prepared from these bivalves were used to activate 2-aminoanthracene (2-AA) and 2-acetylaminofluorene (AAF) to bacterial mutagens. The activation of 2-AA and AAF by clams was studied further in terms of the subcellular location of the activating system, the effect of cofactors and metabolic inhibitors, and its relationship to benzo[a]pyrene hydroxylase (BPH) activity. Finally, the biotransformation capacity of exposed clams in these model experiments was compared with results obtained with clams from Andalusian littoral sites with different contaminant levels. MATERIALS AND METHODS Chemicals AAF, benzo[a]pyrene (B[a]P), NADPH, ␣- and ␤-naphthoflavone (ANF, BNF), and methimazole (MZ) were purchased from Sigma (St. Louis, MO). 2-AA was obtained from Aldrich (St. Louis, MO), and 3-hydroxybenzo[a]pyrene (3-OH-B[a]P) from the Community Bureau of Reference (BCR; Brussels, Belgium). Aroclor 1254 was from Chem Service (West Chester PA), arsenic (III) oxide from Panreac (Barcelona,

Spain), and cupric chloride and tributyltin chloride from Merck (Darmstadt, Germany).

Exposure Experiments and Samples From Natural Environments Exposure to model pollutants was carried out at the Aquaculture Research and Development Center “El Torun˜o” (Puerto Santa Marı´a, Ca´diz, Spain). Chamaelea gallina specimens, obtained from a natural bank at an unpolluted site of Ca´diz Bay, were placed in baskets (⬃ 400 clams/m2) and loaded into tanks (400 L, 105 cm diameter). They were maintained at 18°C with daily water changes for at least 1 week before the treatments began and fed daily with 1 L/tank of a microalgae mixture (Chaetoceros gracilis/ Isocrysis galvana, 1.8 ⫻ 107 cells/ml). The animals were exposed for 7 days to model pollutants dissolved in the water at the doses indicated: Aroclor 1254 (10, 100, 1,000 ppb), CuCl2 (100, 1,000, 5,000 ppb), tributyltin (0.3, 1, 3 ppb), and As2O3 (100, 1,000, 10,000 ppb); separate tanks were used for each compound and the control. After exposure, the clams were collected, their shells were discarded, and whole bodies were pooled and frozen at ⫺80°C. The frozen clams were transported to the University of Co´rdoba and subsequently ground in a mortar with liquid N2 and stored at ⫺80°C until analyzed. Clams were also sampled from two littoral sites having different pollutant levels. Punta Umbrı´a animals (Huelva, Spain) are chronically exposed to the effluents of the Huelva Estuary that contains metals and organic pollutants derived from mining and industrial sources and also agricultural plaguicides. In contrast, Almerimar clams (Almerı´a, Spain) are exposed to lower metal and organic pollutant levels, although they may be affected by a large marina located nearby [Rodrı´guez–Ortega et al., 2002]. Clams from those two sites were frozen, ground, and stored as above.

Preparation of Subcellular Fractions and Metabolic Activation Mixtures All steps were carried out at 4°C using sterile solutions. S9 fractions were obtained as described by Maron and Ames [1983] with minor modifications. Pooled ground animals were weighed and homogenized in ice-cold 20 mM Tris-HCl buffer, pH 7.6, containing 0.5 M sucrose and 0.15 M KCl (1.5 ml/g tissue) using a Potter-Elvehjem apparatus with teflon pestle. Homogenates were centrifuged at 9,000 g for 10 min (Beckman, J2-21). Supernatants (S9 fractions) were decanted to avoid the superficial lipid layer, and 1 ml aliquots were stored at ⫺80°C in polystyrene tubes. Freshly prepared S9 fractions were used for the preparation of cytosolic fractions (CFs) and microsomal fractions (MFs) by ultracentrifugation at 105,000 g for 1 h (Beckman, L8-80M). Supernatants were used as CFs; microsomal pellets were resuspended in an equal volume of 1.15% KCl and repelleted at 105,000 g for 30 min. The washed microsomes were resuspended in 0.1 M potassium phosphate buffer (pH 7.5, 0.05 ml/g of initial tissue) containing 20% glycerol and 0.1 mM EDTA and used as MFs. Subcellular fractions were also stored at ⫺80°C until used. Metabolic activation mixtures contained 0.1 M sodium phosphate buffer, pH 7.4, 33 mM KCl, 5 mM glucose– 6 –phosphate, 4 mM NADP⫹, 8 mM MgCl2, and the corresponding subcellular fraction, in a total volume of 0.5 ml. The protein content of S9 fractions was ⬃ 20 mg/ml; that of CFs, ⬃ 5 mg/ml; and that of MFs, ⬃ 25 mg/ml. The activation mixture was prepared immediately before each experiment, sterilized by filtration (Millipore, 0.22 ␮m), and maintained at 4°C until used.

Metabolic Activation Assays Salmonella typhimurium strain BA149 [Jurado et al., 1994] was used for arylamine activation assays following the preincubation protocol described by Maron and Ames [1983] with minor modifications. These bacteria carry

Mutagenic Activation of Arylamines

57

found with higher and lower concentrations of S9 protein. This ensures that if activation capacity increased due to natural or experimental exposure, a greater mutant frequency would result; conversely, if any pollutant inhibited the activation capacity, a lower mutagenic response would result.

Benzo[a]pyrene Hydroxylase Assay BPH activity was measured in a Perkin Elmer LS 50B spectrofluorimeter as described by Dehnen et al. [1973] and Michel et al. [1994]. Proteins (0.25 mg) from MFs were incubated at 30°C for 15 min in a final volume of 0.9 ml containing 50 mM Tris-HCl buffer, pH 7.3, 2 mg/ml bovine serum albumin, 0.74 mM NADPH, and 0.07 mM B[a]P in acetone. The reaction was stopped by adding 0.1 ml of 10% Triton X100 in triethylamine. In time 0 assays, B[a]P was added after stopping the reaction. Incubations were run in triplicate, and standard curves were prepared by adding known quantities of 3-OH-B[a]P to time 0 assays. Fluorescence was determined using an excitation wavelength of 435 nm and calculated as the difference between 521 and 457 nm emission wavelengths. BPH activity was expressed as pmol of 3-OH-B[a]P equivalent produced per min per mg of protein.

Protein Determination Protein concentration was determined by the bicinchoninic acid method [Smith et al., 1985] using bovine serum albumin as the standard.

Statistical Analysis Fig. 1. Top: Effect of 2-AA and AAF concentrations on metabolic activation to genotoxins by S9 fractions of C. gallina (1 mg protein/plate). Bottom: Effect of C. gallina S9 protein concentration on metabolic activation of 2-AA (6 ␮g/plate) and AAF (50 ␮g/plate). Arrows indicate the promutagen dose (2-AA, 3 ␮g/plate; AAF, 50 ␮g/plate) and the S9 protein concentration (0.75 mg/plate) used for all subsequent assays.

The statistical significance of the results was evaluated by the Students’ t-test using InStat software (Graphpad, San Diego, CA). The normality of the data and the homogeneity of variance were ensured before testing. The statistical significance of the results is shown as follows: asterisk, P ⬍ 0.05; double asterisk, P ⬍ 0.01; and triple asterisk, P ⬍ 0.001.

RESULTS the hisD3052 allele for detection of His⫹ revertants [Maron and Ames, 1983] and the plasmid pYG219 for overproduction of O-acetyltransferase [Watanabe et al., 1990]. A bacterial colony from a master plate was inoculated into 25 ml of liquid Luria-Bertani medium containing 25 ␮g/ml ampicillin and 6.25 ␮g/ml tetracycline and grown at 37°C in 250 ml flasks for 16 hr without shaking, followed by a 2-hr incubation with shaking (90 rpm). The cultures were centrifuged at 2,000 g for 20 min and resuspended in 0.1 M sodium phosphate buffer, pH 7.4, to obtain a final density of 109 cells/ml. Preincubations were carried out in glass tubes (8 ⫻ 100 mm) by adding, in the following order, 0.5 ml of the metabolic activation mixture (buffer in controls), 0.1 ml of bacterial suspension in buffer (108 cells), and 0.03 ml of promutagen (AAF or 2-AA), dissolved in DMSO. The tubes were incubated for 20 min at 37°C with orbital shaking (90 rpm); 2 ml of molten top agar were then added to each tube and the contents were mixed and poured onto selective plates [Maron and Ames, 1983]. Total colonyforming units were determined by seeding Luria-Bertani plates with appropriate dilutions of the bacterial suspension. Plates were incubated upside-down at 37°C and colonies were counted after 1 day (total viable cells) or 2 days (His⫹ revertants). The doses of promutagens and the concentrations of S9 proteins were experimentally optimized. As shown in the top part of Figure 1, with 1 mg of S9 protein/plate, the mutagenic response was saturated with 3 ␮g/plate of 2-AA and 50 ␮g/plate of AAF, doses that were selected for all subsequent assays. The above-mentioned saturation was due to limiting metabolic activity, and increased activity resulted in higher mutagenic responses. As shown in the bottom part of Figure 1, a linear relationship was

Using an O-acetyltransferase overproducing strain of S. typhimurium, we had previously shown that S9 fractions from C. gallina efficiently activated arylamines to bacterial mutagens and that this activation capacity was found both in microsomes and cytosol [Dı´az-Me´ndez et al., 1998]. To test the effect of environmental contaminants on this activation, clams were exposed to water containing four model pollutants and their biotransformation capability was assessed in different subcellular fractions. Figure 2 shows the results of this experiment. Exposure to Aroclor 1254 (10 –1,000 ppb, lower scale) did not alter significantly the capacity of S9, CFs, or MFs to activate 2-AA or AAF, particularly in animals exposed to high pollutant doses. In contrast, significantly lower activation of both aromatic amines was found in those clams exposed to copper or arsenic. Thus, S9 of Cu(II)-exposed (100 –5,000 ppb, upper scale) clams activated 2-AA and AAF with 10- and 7-fold lower efficiency, respectively, than S9 from unexposed clams. The maximum effect was seen with 100 ppb Cu(II), the lowest concentration tested. The activation capacity of Cu(II)-exposed clams declined in incubations with CFs and MFs to the same extent as with S9. All fractions from As(III)-exposed animals (100 –10,000 ppb, upper scale) also had less activation

58

Rodrı´guez-Ortega et al.

Fig. 2. Metabolic activation of 2-AA (3 ␮g/plate) and AAF (50 ␮g/plate) by S9 (0.75 mg protein/plate), CF (0.75 mg protein/plate), and MF (0.1 mg protein/plate) from clams exposed 7 days to the indicated concentrations of Aroclor 1254 (filled circle, lower scale), copper (open circle, upper scale), TBT (filled triangle, lower scale), or As(III) (open triangle, upper scale).

The number of His⫹ revertants per plate without metabolic activation was 106 ⫾ 18 (2-AA) and 20 ⫾ 4 (AAF), respectively. Spontaneous control values (with metabolic activation system but no test chemical) averaged 18 ⫾ 5 His⫹ revertants per plate. Data are mean ⫾ SD of three replicates.

capability than unexposed clams, particularly for 2-AA activation. Although less marked than that of Cu(II), the As(III) effect was dose-dependent in all fractions. Exposure to tributyltin (TBT) (0.3–3 ppb, lower scale) had diverging effects on the activation of the promutagens. In comparison with unexposed clams, TBT-exposed animals had lower 2-AA activation but higher AAF activation capacity, with both promutagens showing dose-response relationships. While Cu(II) affected the arylamine activation capacity of all subcellular fractions to similar extents, the strongest effects of As(III)- and TBT-exposure were found with microsomes. To characterize the enzyme(s) responsible for arylamine activation, the reducing cofactor requirement was studied in subcellular fractions isolated from S9 (Fig. 3). CFs and MFs differed in the activation of each arylamine and in the reductant requirements. CFs promoted a low 2–AA activation, which increased 4.4-fold in the presence of glucose6-phosphate and NADP⫹, and 8-fold when glucose-6-phosphate dehydrogenase (G-6PDH) was also included. NADPH alone provided a similar level of 2-AA activation,

7.5-fold over the control level, but NADH was the reductant preferred for activation of this arylamine, raising the activity to 9.5-fold over the control value. In contrast, activation of 2-AA by MFs was nearly independent of electron donors, since mutagenicity increased 1.4-fold in the presence of glucose-6-phosphate, NADP⫹, and G– 6PDH, and increases of 1.3- and 1.5-fold, respectively, were observed when NADPH or NADH was included in the activation mixtures. AAF was activated less efficiently than 2-AA, although a similar pattern was observed in the activation of both arylamines. AAF activation by CF depended moderately on reducing cofactors, NADH again being the most effective electron donor, while MF activated AAF to a lower extent and almost independently of cofactor presence. Approximately 50% of the arylamine activation capacity detected in S9 fractions of nontreated C. gallina was recovered in CF and MF, with the greater amount of activation capacity being present in CF. On average, we estimate that the S9 prepared from 1 g of C. gallina tissue would activate 3 ␮g/plate of 2-AA to yield 110,000 His⫹ revertants, while the CF prepared from this tissue would yield 60,000 rever-

Mutagenic Activation of Arylamines

Fig. 3. Metabolic activation of 2-AA(3 ␮g/plate) and AAF (50 ␮g/plate) by CF (0.75 mg protein/plate) and MF (0.1 mg protein/plate) from unexposed clams without cofactors (A) or with glucose-6-phosphate and NADP⫹ (B), glucose-6-phosphate, NADP⫹, and glucose-6-phosphate dehydrogenase (C), 8 mM NADPH (D), or 8 mM NADH (E). The number of His⫹ revertants per plate without metabolic activation was 98 ⫾ 11 (2–AA) and 17 ⫾ 5 (AAF), respectively. Spontaneous control values averaged 16 ⫾ 4 His⫹ revertants per plate. Data are mean ⫾ SD of three replicates.

tants (54.5% of that obtained with S9). In contrast, the MF prepared from this amount of tissue would yield 8,000 revertants (7.3% of that obtained with S9). Activation of AAF was less efficient than that of 2-AA and was localized mainly in the cytosol. We estimate that the S9 obtained from 1 g of clam tissue would activate AAF to yield 31,000 His⫹ revertants, while the corresponding CF would yield 14,000 revertants (45.2% of that obtained with S9) and the MF would yield 500 revertants (1.6% of that obtained with S9). The effects of compounds known to interact with several monooxygenases were studied to characterize arylamine activation by CF or MF from clams (Fig. 4). The activation of 2-AA or AAF was assayed with both fractions in the absence or the presence of ANF, a well known inhibitor of CYP1A-linked activity; BNF, a structural analog of ANF [Stegeman and Kloepper-Sams, 1987]; or MZ, a nonphysiological substrate of flavin monooxygenase that competitively inhibits its activity [Ziegler, 1988]. These compounds produced clearly diverging effects in the activation of each arylamine to a mutagen. Activation of AAF by MF and CF was inhibited (to 65%–74% of control) by ANF, BNF, and MZ, in general to a significant extent. In contrast, although inhibited by MZ (to 68%–75%), 2-AA activation by MF or CF increased significantly by nearly twofold when assayed in the presence of ANF. The increases were not as great (especially with CF) when its structural analog, BNF, was used.

59

Fig. 4. Effect of ANF, BNF, and MZ on metabolic activation of 2-AA and AAF by MF (0.1 mg protein/plate) and CF (0.75 mg protein/plate) from unexposed clams. The number of His⫹ revertants per plate without metabolic activation was 114 ⫾ 19 (2–AA) and 22 ⫾ 5 (AAF), respectively. Spontaneous control values averaged 19 ⫾ 4 His⫹ revertants per plate. Data are mean ⫾ SD of three replicates.

TABLE I. Effects of Reducing Cofactors and Compounds Interacting With Cytochromes P450 on Microsomal BPH Activity of Chamaelea gallina Incubation mixture Completea Boiledb ⫺NADPH ⫹ NADH ⫹ ANF 100 ␮M a

BPH (pmol/min/mg)

%

1.462 ⫾ 0.219 0.158 ⫾ 0.014 0.583 ⫾ 0.254 0.524 ⫾ 0.370 3.713 ⫾ 0.556

100 11 40 36 254

The assay was carried out as described in text. Microsomes were boiled 10 min before addition to the incubation mixture.

b

The cytosolic location of arylamine activation and its lack of inhibition by ANF led us to compare this system with clam microsomal BPH activity, clearly linked in vertebrates with CYP1A forms [Stegeman and Kloepper-Sams, 1987; Stegeman and Hahn, 1994]. As shown in Table I, the BPH activity of C. gallina was extensively denatured after 10min boiling. In addition, BPH activity required NADPH as a reductant, decreasing to 40% in its absence, while NADH was an inefficient substitute. Finally, BPH activity increased 2.5-fold in the presence of 100 ␮M ANF. Possible effects of model pollutants on BPH activity were assayed in the MF of clams exposed to Aroclor, Cu(II), As(III), or TBT (Fig. 5). Exposure to Aroclor or cupric ions significantly increased BPH activity, 2.3-fold by Aroclor and 2.2-fold by Cu(II). In contrast, no significant alterations of this microsomal mono-

60

Rodrı´guez-Ortega et al. TABLE II. Contaminant Content and Biotransformation Capability in Clams Collected From Two Different Andalusian Littoral Sites Sampling area

Contaminantsa Cub Znb Cdb ⌺PCBsc DDTsc Biotransformation capability 2-AA activationd BPHe

Almerimar (reference)

Punta Umbrı´a (contaminated)

Ratio (C/R)

3.8 38.0 2.1 0.2 1.1

54.4 114.3 3.4 1.4 4.9

14.3 3.0 1.6 7.0 4.5

1,407 ⫾ 155 0.799 ⫾ 0.170

1,087 ⫾ 212 1.384 ⫾ 0.245

0.8 1.7

a

Data on contaminants from Rodrı´guez-Ortega et al. [2002]. ␮g per gram of dry weight (ppm). c ng per gram of dry weight (ppb). d Induced His⫹ revertants/mg of S9 protein; 3 ␮g of 2-AA and 0.75 mg of protein were added per plate. The number of His⫹ revertants per plate in the absence of metabolic activation was 86 ⫾ 13. Spontaneous control values (with metabolic activation system but without test chemical) averaged 20 ⫾ 4 His⫹ revertants per plate. Data are expressed as mean ⫾ SD of three replicates. e⫺ BPH activity expressed as pmol/min/mg of microsomal protein. Data are expressed as mean ⫾ SD of three replicates. b

Fig. 5. Effect of exposure to Aroclor 1254, Cu(II), TBT, and As(III) on BPH activity in C. gallina microsomes.

oxygenase activity were found in clams exposed to As(III) or TBT. The effects of exposure to model pollutants were also compared with the results obtained in animals from natural areas. To this end, 2-AA activation by S9 fractions and microsomal BPH activity were determined in C. gallina samples collected from littoral areas with different pollution levels (Table II). Clams from a Punta Umbrı´a littoral site, affected by the Huelva Estuary, contained more metals and organic pollutants than clams from Almerimar in the Almerı´a littoral. Compared to this reference, clams from Punta Umbrı´a had a much higher content of three transition metals, Cu, Zn, and Cd, of several PCB congeners, and of DDT and its metabolites. Table II also shows two independent assays of the biotransformation capability of C. gallina collected at these two sites. The 2-AA activation capacity assayed in S9 fractions from Punta Umbrı´a clams was lower than in reference animals, but the differences were not significant (0.8-fold; P ⫽ 0.102). In contrast, significantly higher BPH activity (1.7-fold; P ⫽ 0.027) was detected in microsomes from clams of the more polluted Punta Umbrı´a site than in those living in the reference area. DISCUSSION Arylamine activation in C. gallina was previously described by our group [Dı´az-Me´ndez et al., 1998]. Here we

show that 87% of the 2-AA S9 activation and 95% of the AAF S9 activation capacities are localized in the cytosol of this clam, indicating that these systems cannot be attributed to microsomal cross-contamination of cytosol. This contrasts with results found in untreated rats, in which 2%– 4% of the arylamine activation capacity of S9 was recovered in the CF [Traynor et al., 1991]. In rats, both cytosolic and microsomal systems were needed for maximal expression of mutagenicity, although each of them could activate arylamines independently [Traynor et al., 1991]. A cytosolic, NADPH-dependent oxygenase system that bioactivates aromatic amines, perhaps through N-hydroxylation, was previously described in rats and mice [Ayrton et al., 1992; Marczylo and Ioannides, 1994]. In rodents, this system is only induced by Aroclor 1254 [Leist et al., 1992], through an Ah receptor-linked pathway, as shown by using mouse strains unresponsive to Ah-mediated chemicals [Marczylo and Ioannides, 1997]. In contrast to mammals, our present study shows that arylamine activation capacity is unaltered in C. gallina exposed to high concentrations of Aroclor 1254, a mixture of PCB congeners that this clam concentrates over 50 –fold [Rodrı´guez-Ariza et al., 2002]. The insensitivity of the arylamine activation system of C. gallina to PCBs supports the present doubts concerning the existence of a functional Ah receptor homolog in invertebrates [Hahn, 1998]. Although small dioxin-binding proteins have been reported in the cytosol of several invertebrate species, their functional relationship to the vertebrate Ah receptor and their possible role in toxic responses are still unknown [Brown et al.,

Mutagenic Activation of Arylamines

1997]. The wide evolutionary distance between bivalves and mammals could explain the diverging properties of their arylamine activation systems, as indicated by the preference for NADH in the mollusk arylamine activation system, while those of mammals and fish are absolutely dependent on NADPH [Leist et al., 1992; Rodrı´guez-Ariza et al., 1994]. Our present study further shows that arylamine activation capacity diminishes in clams exposed to copper or arsenic. Cu(II) is a redox-active metal that uses the Fenton reaction to catalyze production of reactive oxygen species that damage lipids, nucleic acids, and proteins [Sies, 1986]. Arsenic is an environmental pollutant of toxicological concern. As(V) undergoes a two-electron reduction to As(III) by reaction with GSH, direct or enzyme-mediated, followed by methylation [Thomas et al., 2001]. The mechanism for As(V) toxicity, replacement of phosphate leading to ATP depletion [Dixon, 1997], is not well known. Trivalent arsenicals, including methylated forms, are more toxic than As(V). They react with dithiols of pyruvate dehydrogenase, glutathione or thioredoxin reductases, and redox-active proteins such as thioredoxins or glutaredoxins. These reactions can drastically alter the cellular redox status leading to an increased sensitivity to oxidative stress [Gebel, 2000; Hughes, 2002]. Arsenic is a human carcinogen [International Agency for Research on Cancer, 1990], whose proposed mechanisms of action include genotoxicity, gene amplification, transformation, altered DNA repair or methylation, cell proliferation, cocarcinogenesis, and tumor promotion [Mass et al., 2001; Hughes, 2002]. Short-term exposure (3–24 hr) of cultured human fibroblasts to 0.1–5 ␮M As(III) upregulates expression of c-Fos and c-Jun, thioredoxin, and Ref-1 and activates AP-1 and NF-␬B binding, while chronic exposure, as in our controlled experiment, decreases such effects [Xu et al., 2002]. Evidence of As-induced oxidative stress in mammals has been reported in vivo [Flora, 1999]. The expression of CYP1A1 and some housekeeping genes is repressed under oxidative stress, contributing to an improved cellular homeostasis [Barouki and Morel, 2001]. Alternatively, loss of arylamine activation in Cu(II)- or As(III)-exposed clams could be due to oxidative damage of the enzymes involved in 2-AA or AAF biotransformation, as shown in Cu(II)-treated fish [Rodrı´guez-Ariza et al., 1995]. TBT is an ubiquitous contaminant that inhibits several biotransforming systems and is toxic to aquatic organisms, particularly mollusks, leading to masculinization of female gastropods and shell anomalies in oysters [Alzieu, 2000]. Masculinization has been attributed to TBT interaction with cytochrome P450-catalyzed steroid metabolism. Significant increases in testosterone levels were reported in clams (Ruditapes decussata) exposed to TBT due to inhibition of the P450 aromatase that converts testosterone to estradiol 17-␤ [Morcillo et al., 1998]. Organotins are also potent inhibitors of glutathione transferase-mediated conjugation and excre-

61

tion in mammals and fish [Henry and Byington, 1976; Al-Ghais and Ali, 1999], and this could also be involved in mollusk masculinization [Ronis and Mason, 1996]. Preferential inhibition by TBT of the conjugation/detoxification of AAF metabolites, as compared to 2-AA derivatives, could explain the increased AAF activation in clams exposed to this organotin. A differential interaction of TBT with the 2-AA and AAF activation systems cannot be ruled out. Although the reductant requirements for activation of each arylamine were similar, NAD(P)H increased 2-AA activation more than AAF activation. The twofold increase by ANF of 2-AA activation, but not of AAF activation, suggests the involvement of different pathways in the bioactivation of each arylamine. In vertebrates, ANF inhibits CYP1A-dependent activities, although some forms of cytochromes P450, such as those of the CYP3A subfamily, can be activated by this compound [Shou et al., 1994]. Since most known cytochromes P450 are membrane-bound, both the cytosol location of the 2-AA activation system and its marked activation by ANF argue against significant involvement of CYP1A in the activation of this arylamine. The potentiation by ANF of 2-AA activation could be explained if ANF increased the activity of enzymes converting 2-AA into genotoxins, or ANF inhibited enzymes transforming 2-AA into nonmutagenic chemicals, such as ring-hydroxylated derivatives [Tong et al., 1986]. Model pollutants had different effects on clam BPH activity and on their ability to activate arylamines, suggesting that different catalysts were involved in these two processes. The higher microsomal BPH activity after Aroclor exposure agrees with previously reported increases in CYP1A-immunoreactive protein and related activities in PCB-exposed Mytilus edulis [Peters et al., 1998], although the participation of forms different from CYP1A in the BPH activity of C. gallina cannot be excluded. After Cu(II) exposure, clam microsomes also showed a twofold increase in BPH activity. This is an unanticipated result since many Cu(II) effects are usually attributed to oxidative stress, which promote lower CYP1A-related activities in vertebrates [Barouki and Morel, 2001]. Alternatively, Cu(II) could enhance the levels of cortisol, a general stress hormone in mollusks, as shown in fish exposed to sublethal Cd stress [Fu et al., 1990]. The higher cortisol levels could increase the expression of some other glucocorticoid-induced cytochrome P450s, such as CYP3A [Okey, 1990]. In fact, microsomal BPH activity of C. gallina is activated by ANF, a compound that also stimulates B[a]P metabolism by CYP3A forms [Shou et al., 1994]. An additional aim of this study was to compare the effects of pollutants on arylamine biotransformation using clams exposed under controlled conditions and clams from natural areas with different contaminant levels. The results of both analyses fit reasonably well, since clams from the polluted site, with much higher levels of copper and other metals, showed a lower capacity to activate 2-AA and

62

Rodrı´guez-Ortega et al.

higher microsomal BPH activity, as was observed in Cu(II)exposed clams. Nevertheless, some pollutant(s) different from copper could be responsible for the lower arylamine activation capability and/or higher BPH activity. In support of this contention, high concentrations of PCBs and of DDT and its metabolites were also found in bivalves from the more polluted site (Punta Umbrı´a), an area chronically exposed to contaminants from mining as well as industrial and agricultural activities [Rodrı´guez–Ortega et al., 2002]. In summary, the cytosol of C. gallina clams has an enzyme system that catalyzes arylamine bioactivation to genotoxins and preferentially uses NADH as an electron donor. The activity of this system is not altered by PCB exposure. Arylamine bioactivation, however, is significantly decreased after exposure of clams to As(III) and Cu(II), treatments that significantly enhance the microsomal BPH activity. These results suggest that arylamine activation in bivalves is not catalyzed by microsomal monooxygenases, but by other soluble system(s) yet unknown. Further research is needed to identify the responsible catalyst(s) and to understand the effects of metal exposure on biotransformation capability in bivalves. The results of this study also suggest that the ability of clams to activate arylamine to mutagens may be useful as a biomarker of marine pollution. REFERENCES Al-Ghais SM, Ali B. 1999. Inhibition of glutathione S–transferase catalyzed xenobiotic detoxication by organotin compounds in tropical marine fish tissues. Bull Environ Contam Toxicol 62:207–213. Alzieu C. 2000. Impact of tributyltin in marine invertebrates. Ecotoxicology 9:71–76. Ayrton AD, Neville S, Ioannides C. 1992. Cytosolic activation of 2–aminoanthracene: implications in its use as diagnostic mutagen in the Ames test. Mutat Res 104:43– 48. Bagnasco M, Camoirano A, De Flora S, Melodia S, Arillo A. 1991. Enhanced liver metabolism of mutagens and carcinogens in fish living in polluted sea water. Mutat Res 262:129 –137. Barouki R, Morel Y. 2001. Repression of cytochrome P4501A1 gene expression by oxidative stress: mechanisms and biological implications. Biochem Pharmacol 61:511–516. Britvic S, Kurelec B. 1986. Selective activation of carcinogenic aromatic amines to bacterial mutagens in the marine mussel Mytilus galloprovincialis. Comp Biochem Physiol 85C:111–114. Brown DJ, Clarke GC, Van Beneden RJ. 1997. Halogenated aromatic hydrocarbon– binding proteins identified in several invertebrate marine species. Aquat Toxicol 37:71–78. Dehnen W, Tomingas R, Roos J. 1973. A modified method for the assay of benzo[a]pyrene hydroxylase. Anal Biochem 53:373–383. Dı´az-Me´ndez FM, Rodrı´guez-Ariza A, Usero-Garcı´a J, Pueyo C, Lo´pezBarea J. 1998. Mutagenic activation of aromatic amines by molluscs as a biomarker of marine pollution. Environ Mol Mutagen 31:282–291. Dixon HBF. 1997. The biochemical action of arsonic acids especially as phosphate analogs. Adv Inorg Chem 44:191–227. Flora SJ. 1999. Arsenic-induced oxidative stress and its reversibility following combined administration of N–acetylcysteine and meso 2,3– dimercaptosuccinic acid in rats. Clin Exp Pharmacol Physiol 26:865– 869. Forster R, Green MHL, Priestley A, Gorrod JW. 1981. Apparent activation

of 2–acetylaminofluorene and some aromatic amines by cytosolic preparations. Mutat Res 85:187–194. Fu H, Steinebach OM, van der Hamer CJA, Balm PHM, Lock RAC. 1990. Involvement of cortisol and metallothionein-like proteins in the physiological response of tilapia (Oreochromis mossambicus) to sublethal cadmium stress. Aquat Toxicol 16:257–269. Gebel T. 2000. Confounding variables in the environmental toxicology of arsenic. Toxicology 144:155–162. Hahn ME. 1998. The aryl hydrocarbon receptor: a comparative perspective. Comp Biochem Physiol 121C:23–53. Henry RA, Byington KH. 1976. Inhibition of glutathione S–aryl transferase from rat liver by organogermanium, lead and tin compounds. Biochem Parmacol 25:2291–2295. Hughes MF. 2002. Arsenic toxicity and potential mechanisms of action. Toxicol Lett 133:1–16. International Agency for Research on Cancer. 1990. IARC monographs on the evaluation of carcinogenic risks of chemicals to humans. Lyon, France: IARC Press. p 39 –141. Jurado J, Alejandre-Dura´n E, Pueyo C. 1994. Mutagenicity testing in Salmonella typhimurium strains possessing both the His reversion and Ara forward mutation systems and different levels of classical nitroreductase or O-acetyltransferase activities. Environ Mol Mutagen 23:286 –293. Leist M, Ayrton AD, Ioannides C. 1992. A cytosolic oxygenase activity involved in the bioactivation of 2–aminofluorene. Toxicology 71: 7–20. Livingstone DR, Kirchin MA, Wiseman A. 1989. Cytochrome P450 and oxidative metabolism in molluscs. Xenobiotica 19:1041–1062. Marczylo T, Ioannides C. 1994. Bioactivation of 6 –aminochrysene by animal and human hepatic preparations: contributions of microsomal and cytosolic enzyme systems. Mutagenesis 9:233–239. Marczylo T, Ioannides C. 1997. Induction of the rat hepatic cytosolic arylamine oxidase by a series of polychlorinated biphenyls: association with the Ah locus. Toxicol Lett 92:81–91. Marczylo T, Ioannides C. 1998. The substrate specificity of the rat hepatic cytosolic arylamine oxidase catalyzing the bioactivation of aromatic amines. Cancer Lett 127:141–146. Maron DM, Ames BM. 1983. Revised methods for the Salmonella mutagenicity test. Mutat Res 113:173–215. Mass MJ, Tennant A, Roop BC, Cullen WR, Styblo M, Thomas DJ, Kligerman AD. 2001. Methylated trivalent arsenic species are genotoxic. Chem Res Toxicol 14:355–361. Michel XR, Cassand PM, Narbonne JF. 1993. Activation of benzo[a]pyrene and 2–aminoanthracene to bacteria mutagens by mussel digestive gland postmitochondrial fraction. Mutat Res 301:113– 119. Michel X, Salau¨n JP, Galgani F, Narbonne JF. 1994. Benzo[a]pyrene hydroxylase activity in the marine mussel Mytilus galloprovincialis: a potential marker of contamination by polycyclic aromatic hydrocarbon–type compounds. Mar Environ Res 38:257–273. Morcillo Y, Ronis MJJ, Sole´ M, Porte C. 1998. Effects of tributyltin on the cytochrome P450 monooxygenase system and sex steroid metabolism in the clam Ruditapes decussata. Mar Environ Res 46:583– 586. Okey AB. 1990. Enzyme induction in the cytochrome P450 system. Pharmac Ther 45:241–298. Peakall DB. 1994. Biomarkers: the way forward in environmental assessment. Toxicol Ecotoxicol News 1:55– 60. Pelroy RA, Gandolfi AJ. 1980. Use of mixed function amine oxidase for metabolic activation in the Ames/Salmonella assay system. Mutat Res 72:329 –334. Peters LD, Shaw JP, Nott M, O’Hara SCM, Livingstone DR. 1998. Development of cytochrome P450 as a biomarker of organic pollution in Mytilus sp.: field studies in United Kingdom (“sea empress” oil spill) and the Mediterranean Sea. Biomarkers 4:425– 441.

Mutagenic Activation of Arylamines Ringwood AH, Hammeedi MJ, Lee RF, Brower M, Peters EC, Scott GI, Luoma SN, DiGiulio RT. 1998. Bivalve biomarkers workshop: overview and discussion group summaries. Biomarkers 4:391–399. Rodrı´guez-Ariza A, Dorado G, Navas JI, Pueyo C, Lo´pez-Barea J. 1994. Promutagen activation by fish liver as biomarker of littoral pollution. Environ Mol Mutagen 24:116 –123. Rodrı´guez-Ariza A, Dı´az–Me´ndez FM, Navas JI, Pueyo C, Lo´pez-Barea J. 1995. Metabolic activation of carcinogenic aromatic amines by fish exposed to environmental pollutants. Environ Mol Mutagen 25:50 – 57. Rodrı´guez-Ariza A, Rodrı´guez-Ortega MJ, Marenco JL, Amezcua O, Alhama J, Lo´pez-Barea J. 2002. Uptake and clearance of PCB congeners in Chamaelea gallina: response of oxidative stress biomarkers. Comp Biochem Physiol C (in press). Rodrı´guez-Ortega MJ, Alhama J, Funes V, Romero-Ruiz A, Rodrı´guezAriza A, Lo´pez-Barea J. 2002. Biochemical biomarkers of pollution in the clam Chamaelea gallina from South-Spanish littoral. Environ Toxicol Chem 21:542–549. Ronis MJJ, Mason AZ. 1996. The metabolism of testosterone by the periwinkle (Littorina littorea) in vitro and in vivo: effects of tributyltin. Mar Environ Res 42:161–166. Shou M, Grogan J, Mancewicz JA, Krausz KW, Gonza´lez FJ, Gelboin HV, Korzekwa KR. 1994. Activation of CYP3A4: evidence for the simultaneous binding of two substrates in a cytochrome P450 active site. Biochemistry 33:6450 – 6455. Sies H. 1986. Biochemistry of oxidative stress. Angew Chem Int Ed Engl 25:1058 –1071. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC. 1985. Measurement of protein using bicinchoninic acid. Anal Biochem 150: 76 – 85. Steele CM, Ioannides C. 1986. Induction of hepatic mixed function oxidases by aromatic amines and its relationship to their bioactivation to mutagens. Mutat Res 162:41– 46.

63

Stegeman JJ, Kloepper-Sams P. 1987. Cytochrome P-450 isoenzymes and monooxygenase activity in aquatic animals. Environ Health Perspect 71:87–95. Stegeman JJ, Hahn ME. 1994. Biochemistry and molecular biology of monooxygenases: current perspectives on forms, functions, and regulation of cytochrome P450 in aquatic species. In: Malins DC, Ostrander GK, editors. Aquatic toxicology: molecular, biochemical and cellular perspectives. Boca Raton: CRC Press. p 87–206. Thomas DJ, Styblo M, Lin S. 2001. The cellular metabolism and systemic toxicity of arsenic. Toxicol Appl Pharmacol 176:127–144. Tong S, Smith J, Manson D, Gorrod JW, Ioannides C. 1986. The metabolic activation of 2–naphtylamine to mutagens in the Ames test. Anticancer Res 6:1107–1112. Traynor CA, Shane BS, Hajos AKD, Winston GW. 1991. Arylamine activation following chronic ethanol ingestion by rats: studies on the liver S9, microsomal and cytosolic fractions and comparison with Aroclor 1254 pretreatment. Mutat Res 247:153–166. Watanabe M, Ishidate M Jr, Nohmi T. 1990. Sensitive method for the detection of mutagenic nitroarenes and aromatic amines: new derivatives of Salmonella typhimurium tester strains possessing elevated O–acetyltransferase levels. Mutat Res 234:337–348. Xu Y, Jin X, Snow ET. 2002. Effect or arsenic on transcription factor AP-1 and NF-␬B binding activity and related gene expression. Toxicol Lett 133:33– 45. Yamazoe Y, Abu-Zeid M, Gong D, Staiano N, Kato R. 1989. Enzymatic acetylation and sulfation of N-hydroxylarylamines in bacteria and rat liver. Carcinogenesis 10:1675–1679. Ziegler DM. 1988. Flavin-containing monooxygenase: catalytic mechanism and substrate specificities. Drug Metab Rev 19:1–32.

Accepted by— E. Zeiger