Oxidative and modulatory effects of trace metals on metabolism of polycyclic aromatic hydrocarbons in the Antarctic fish Trematomus bernacchii

Aquatic Toxicology 85 (2007) 167–175

Oxidative and modulatory effects of trace metals on metabolism of polycyclic aromatic hydrocarbons in the Antarctic fish Trematomus bernacchii Maura Benedetti a , Giacomo Martuccio a , Daniele Fattorini a , Adriana Canapa a , Marco Barucca a , Marco Nigro b , Francesco Regoli a,∗ a

Istituto di Biologia e Genetica, Universit`a Politecnica delle Marche, Ancona, Via Ranieri Monte d’Ago, 60100 Ancona, Italy b Dipartimento di Morfologia Umana e Biologia Applicata, Universit` a di Pisa, Italy Received 16 July 2007; received in revised form 24 August 2007; accepted 28 August 2007

Abstract Biological interactions between various classes of pollutants are of great relevance for the Antarctic marine environment, where the naturally elevated bioavailability of metals like cadmium might indirectly influence sensitivity of endemic organisms toward other environmental pollutants, e.g. polycyclic aromatic hydrocarbons (PAHs). To further investigate reciprocal effects of different chemicals, the fish Trematomus bernacchii was exposed to trace metals (Cd, Cu, Hg, Ni, Pb) and benzo[a]pyrene (BaP, as a model PAH), dosed alone and in combinations. Co-exposures revealed that BaP did not influence the accumulation of metals, while these elements caused significant changes on tissue levels of the PAH. The marked EROD induction caused by BaP was completely suppressed by co-exposure with Cd and Cu, but no effects were observed with Ni, Hg and Pb. Similar results were confirmed at the protein level by Western blot analyses while CYP1A1 mRNA levels were reduced only during Cd co-exposures. Clear evidence of oxidative perturbations was observed in fish co-treated with Cd and BaP and the reduced capability to absorb peroxyl and hydroxyl radicals suggested some oxidative pathways by which this element might indirectly modulate the biotransformation efficiency of Cytochrome P450. Partly different and post-transcriptional mechanisms of action could be hypothesized for Cu, while moderate oxidative effects of Hg, Ni and Pb during co-exposures would confirm their limited influence on metabolism of PAHs. In general, the overall results revealed a complex pathway of interactions between different chemicals during co-exposures and the importance of oxidative status in modulating induction and expression of CYP1A1. © 2007 Elsevier B.V. All rights reserved. Keywords: Trace metals; Polycyclic aromatic hydrocarbons; Interactions; Bioaccumulation; Cytochrome P450; Oxidative stress; Antarctica; Trematomus bernacchii

1. Introduction Exposure of marine organisms to mixtures of xenobiotics represents an emerging environmental issue since interactions between chemicals and their multiple biological effects are largely unknown (Stohs and Bagchi, 1995; Carpenter et al., 2002). Different chemicals may have different mechanisms of action at the cellular level, and even a single molecule may exert its effects through more than one pathway. The toxicity and carcinogenicity of PAHs can be a direct consequence of metabolic activation by Cytochrome P450, or indirectly mod-

∗

Corresponding author. Tel.: +39 071 2204613; fax: +39 071 2204609. E-mail address: [email protected] (F. Regoli).

0166-445X/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2007.08.009

ulated by the increased production of reactive oxygen species (ROS) and transcriptional regulation of several gene systems (Morel and Barouki, 1998; Meyer et al., 2002). Similarly, trace metals are potent toxicants through a wide spectrum of mechanisms including the increased formation of ROS which damage proteins, DNA and lipids (Frenzilli et al., 2001; Regoli et al., 2004), enzyme inhibition, impairment of cell signalling and calcium homeostasis, changes in gene regulation and physiological alterations (Stohs and Bagchi, 1995; Elbekai and El-Kadi, 2005). Even more complex and difficult to predict are the biological effects caused by different classes of chemicals during co-exposures when reciprocal interactions, cascade and indirect mechanisms can both enhance or suppress the expected responses. In this respect, trace metals have been shown to affect the mutagenicity and carcinogenicity of benzo[a]pyrene

168

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

(BaP) and 2,3,7,8-tetrachlorodibenzo-p-dioxin by disrupting the expression of a battery of detoxifying genes (Tully et al., 2000; Korashy and El-Kadi, 2004). They can also alter the metabolism of the Cytochrome P450 (CYP450) system (Korashy and ElKadi, 2004); recent in vitro and in vivo studies in fish and mammalian cells demonstrated that some elements (e.g., Cr, Pb, Cu, Zn, and As) reduce the efficiency of biotransformation enzymes at the transcriptional level (Risso-de Faverney et al., 2000; Tully et al., 2000), protein and catalytic activity (Sandvik et al., 1997; Viarengo et al., 1997; Regoli et al., 2005a), while other authors reported stimulating effects of arsenic on CYP1A mRNA, protein and activity in rats (Seubert et al., 2002; Sorrentino et al., 2005) and of cadmium in fish co-exposed with benzo[a]pyrene (Lemaire-Gony et al., 1995). Investigations on the effects of chemical mixtures are particularly lacking for the Antarctic ecosystem (Regoli et al., 2005a), despite the constant increase of human activities and the great sensitivity of endemic marine species. Biological responses of Antarctic organisms are influenced by extreme environmental conditions, marked fluctuations of food availability and metabolic processes and by the naturally elevated levels of cadmium in the Ross Sea (Regoli et al., 2002, 2005a,b). The rock cod Trematomus bernacchii is a key species in the Ross Sea, widely distributed and commonly found within the first 200 m of depth. It represents more than 90% of the abundance and biomass of coastal fish fauna at Terra Nova Bay (Vacchi et al., 1996). Previous investigations on this species, characterized several cellular responses to pollutants including the inducibility of Cytochrome P450 (CYP1A) and metallothioneins, oxyradical metabolism and susceptibility to oxidative stress, onset of DNA damages and vitellogenin gene expression (Focardi et al., 1995; Miller et al., 1999; Regoli et al., 2005a; Canapa et al., 2007). These data demonstrated reciprocal interactions between the metabolism of organochlorine xenobiotics and cadmium suggesting endocrine effects also from chronic exposure to this element (Regoli et al., 2005a; Canapa et al., 2007). The interest for trace metals is of particular relevance in Antarctica, where the basal concentrations of these elements are often influenced by geological anomalies or specific local features, such as the upwelling phenomena responsible for the unusually high levels of cadmium in organisms of Terra Nova Bay, Ross Sea (Bargagli et al., 1996). Natural hepatic concentrations of cadmium in the liver of T. bernacchii range between approximately 9 and 24 ␮g/g, values approximately 10–50 fold higher than those typical of temperate species (Canapa et al., 2007). Although this natural enrichment of cadmium has no direct adverse consequences for Antarctic fishes, some interactions with the metabolism of other chemicals have been demonstrated (Regoli et al., 2005a; Canapa et al., 2007). The aim of this work was to extend our knowledge on the biological effects of chemical mixtures including different trace metals (Cd, Cu, Hg, Ni, Pb) and benzo[a]pyrene, chosen as a model for polycyclic aromatic hydrocarbons (PAHs) which can be potentially released at a local level from scientific bases, shipping or transport operations and accidental oil spills. Organisms were exposed to chemicals dosed alone and in combinations. Analyses of metals and BaP were integrated with a wide range

of biomarkers commonly used as diagnostic and prognostic tools in environmental studies. The induction and modulation of biotransformation efficiency of Cytochrome P450 was determined as the catalytic activity of ethoxyresorufin O-deethylase (EROD), as the protein content and mRNA expression. The metabolism of BaP was also evaluated by the content of aromatic BaP-like metabolites in the bile of exposed T. bernacchii. Considering the crucial role of oxyradical metabolism in adaptive strategies of Antarctic organisms (Regoli et al., 2005a,b) and the importance of ROS in mediating the toxicity of pollutants, several oxidative biomarkers were analysed in treated fish. Individual antioxidant defences, including the activities of catalase, Se-dependent and Se-independent glutathione peroxidases, glutathione reductase and glutathione S-transferases, and total glutathione, were integrated with the measurement of total oxyradical scavenging capacity (TOSC) toward peroxyl (ROO·) and hydroxyl (OH·) radicals. While individual antioxidants can be very sensitive in revealing a pro-oxidant stressor, TOSC values better reflect the overall susceptibility of tissues to oxidative stress conditions (Regoli and Winston, 1999; Gorbi and Regoli, 2004a). The results of this study were expected to provide an additional contribution to characterize the sensitivity of T. bernacchii to pollutants, to mechanisms of interactions between different classes of chemicals and to the role of oxidative responses in modulating such biological effects.

2. Materials and methods 2.1. Organisms and laboratory exposures Sexually mature specimens of Trematomus bernacchii were sampled from Gerlache Inlet (Terra Nova Bay, Ross Sea, Antarctica) during the XIX Italian Antarctic Expedition (austral summer 2003–2004). Organisms were acclimatized to laboratory conditions 1 one week in aquaria with running, unfiltered seawater at a controlled temperature of −1 ± 0.5 ◦ C. A total of 120 fish were randomly divided in 12 groups of treatments (each consisting of 10 specimens) and i.p. injected with: corn oil (control), BaP (10 ␮g/g dissolved in corn oil) or the following trace metals dissolved in physiological saline solution: Cd (2 ␮g/g as CdCl2 ), Cu (2 ␮g/g as Cu Cl2 ), Hg (0.2 ␮g/g as HgCl2 ), Ni (2 ␮g/g as NiCl2 ), Pb (2 ␮g/g as PbCl2 ). Additional five groups of fish were injected with trace metals (Cd, Cu, Hg, Ni, Pb as described above) and after 4 h, with BaP. Despite the fact that exposure concentrations might appear relatively high, they are still environmentally realistic for polluted areas and selected to simulate an acute, short-term exposure. Injected volumes were 100 ␮l/100 g fish weight for corn oil solution and 10 ␮l per 100 g fish for saline solution; the latter, always administered as separate injection, has already been shown to not influence biochemical responses in fish (Regoli et al., 2005a). No mortality was observed during the experiments and organisms were sacrificed after 7 days from injection; livers and gall bladder were rapidly dissected, frozen in liquid nitrogen, and stored at −80 ◦ C until analyses.

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

2.2. Chemical analyses Trace metals were analysed according to previously described methods (Regoli et al., 2004). Fish livers were dried at 70 ◦ C to a constant weight and digested under pressure with nitric acid and hydrogen peroxide (7:1) in a microwave digestor system (CEM, Mars System). Quality assurance and quality control was assessed by processing blank samples and reference standard material (Mussel Tissue Standard Reference Material SRM 2977, National Institute of Standards and Technology). Metals (cadmium, copper, nickel and lead) were analysed by atomic absorption spectrophotometry with flame (Varian, Spectraa 220FS) and flameless atomization (Varian Spectraa 300 Zeeman). The mercury content was quantified by cold vapour atomic absorption spectrometry (Varian, VGA-76, Vapour Generator Accessory). Concentrations obtained for standard reference materials were always within the 95% confidence interval of certified values. Benzo[a]pyrene was determined in hepatic tissues after methanolic extraction (1:10, w/v) in a microwave digestor (150 W for 10 min). Samples were centrifuged at 600 × g for 5 min, methanolic solutions were concentrated in speedvac (RC1009, Jouan) and finally purified with solid phase extraction (Octadecyl C18, 500 mg × 6 ml, Bakerbond). A final volume of 0.5 ml was recovered with acetonitrile and HPLC analyses were carried out using a water–acetonitrile gradient and fluorometric detection. BaP was identified by the retention time of appropriate pure standard solutions and the Quality Assurance and Quality Control (QA/QC) of extraction and analytical procedures were tested processing blank and reference samples (Mussel Tissues Standard SRM 2977, NIST) (Regoli et al., 2005a). 2.3. Activity of CYP1A Ethoxyresorufin O-deethylase (EROD) was assayed spectrofluorometrically as described in Regoli et al. (2005a). Individual livers were homogenized (1:5, w/v) in 100 mM Kphosphate buffer pH 7.5, containing 0.15 M KCl and 1 mM ethylendiaminetetraacetic acid (EDTA). After centrifugation at 12,000 × g for 15 min, the resulting supernatants (S9) were immediately used for determination of enzyme activities. Incubations were carried out at 25 ◦ C in a final volume of 1 ml containing 100 mM K-phosphate buffer pH 7.5, 4 ␮M 7ethoxyresorufin and 0.25 mM NADPH for 3 min, before the addition of 2 ml acetone to stop the reaction. Incubation mixtures as described above, but stopped at time zero were used as blank values and subtracted from the sample fluorescence. Fluorometric analyses (535/585 nm) were quantified by reference to resorufin standards (0.01–1 ␮M).

169

PVDF membrane for 1 h at constant voltage, and blocked in milk (5%, w/v). After incubation with primary polyclonal antirat CYP1A1 antibodies (Santa Cruz) (1:500 dilution), blots were washed three times and incubated with secondary antibody (peroxidase-linked goat anti-rabbit IgG). The detection of antibody was performed by chemiluminescence and the intensity of bands was quantified by image analysis (1D Kodak software). 2.5. Analysis of P450 1A mRNA Total RNA was isolated from individual livers with RNeasy mini Kit (Qiagen). Following treatment with DNase I Amplification Grade (Sigma), an aliquot of each extract was used for gel electrophoresis and spectrophotometry to determine RNA quality and concentration. RT-PCR was performed using iScript cDNA synthesis kit (Bio-Rad) according to the manufacturer’s instructions. The quantitative determinations of P450 1A gene expression were performed by Real-Time PCR. Forward primer 5 -AGACATCCAGATGTCAGACG-3 and reverse primer 5 -TCTTGATACAGCCTCTCCTG-3 for P450 1A were designed on the partial T. bernacchii CYP1A sequence obtained from Miller et al. (1999). The ␤-actin, used to normalize data, was amplified using forward 5 CAGGGAGAAGATGACCCAGAT-3 (Scholz and Gutzeit, 2000) and reverse 5 -CCATCACCGGAATCCATGAC-3 . Amplified sequences of P450 1A and ␤-actin were, respectively, 144 bp and 130 bp long. The real-time PCR was performed on an iQTM multicolour Real-Time PCR Detection System (Bio-Rad) with iQTM SYBR Green Supermix (BioRad). Samples were run in triplicate and cycling parameters were as follows: 3 min at 95 ◦ C, then 40 cycles for 10 s at 95 ◦ C, 20 s at 55 ◦ C, 20 s at 72 ◦ C followed by melt curve analysis to ensure that only a single PCR product was amplified. Relative quantification of the gene P450 1A expression was performed with reference gene (␤-actin) using the CT method. 2.6. Analysis of biliary BaP metabolites by fixed wavelength fluorescence (FF) The analysis of aromatic metabolites in the bile was semi-quantitatively assessed by fixed fluorescence (FF) spectrofluorometry as the sum of benzo[a]pyrene-like metabolites (Lin et al., 1996). After thawing, gall bladders were punctured, and bile was diluted in 48% ethanol (1:2000 to 1:8000, v/v ratio) to obtain a linear response of fluorescent readings. Fluorescence was analysed at the fixed wavelength pairs 380/430 using 1-hydroxy pyrene (1-OH-Pyrene) as reference standard for BaP-type metabolites (Gorbi and Regoli, 2004b).

2.4. Gel electrophoresis and CYP1A immunoblotting

2.7. Antioxidant defences

CYP1A1 proteins were analysed by immunoblotting in hepatic S9 fractions containing constant amounts of proteins (80 ␮g protein). Proteins were separated by SDS-PAGE (3% acrylamide stacking gel and 7.5% acrylamide separating gel) in a Trisglycine buffer systems (Towbin et al., 1979), blotted onto a

Enzymatic antioxidants were measured in liver samples homogenized (1:5, w/v ratio) in 100 mM K-phosphate buffer (pH 7.5), 0.1 mM phenylmethylsulphonyl fluoride (PMSF), 0.1 mg/ml bacitracin, 0.008 TIU/ml aprotinin, NaCl 1.8%, and centrifuged at 110,000 × g for 1 h at 4 ◦ C. Measurements were

170

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

performed with a Varian (model Cary 3) spectrophotometer at a constant temperature of 18 ◦ C; all the assay conditions have been detailed elsewhere (Regoli et al., 2005a). Catalase (EC 1.11.1.6) activity was measured by the decrease in absorbance at 240 nm (ε = 0.04/mM cm) due to H2 O2 consumption (12 mM H2 O2 in 100 mM Na-phosphate buffer pH 7.0). Glutathione peroxidase (GPx) activities were measured in a coupled enzyme system where NADPH is consumed by glutathione reductase which converts the formed GSSG to its reduced form. The decrease of absorbance was monitored at 340 nm (␧ = 6.22/mM cm) in 100 mM K-phosphate buffer pH 7.5, 1 mM EDTA, 1 mM dithiothreitol (DTT), 1 mM NaN3 (for hydrogen peroxide assay), 2 mM GSH, 1 unit glutathione reductase, 0.24 mM NADPH and 0.5 mM hydrogen peroxide or 0.8 mM cumene hydroperoxide as substrates, respectively, for the selenium-dependent and for the sum of Se-dependent and Se-independent forms. The rate of the blank reaction was subtracted from the total rate. Glutathione S-transferase (GST) (EC 2.5.1.18) activities were determined at 340 nm using 1-chloro-2,4-dinitrobenzene (CDNB) as the substrate (ε = 9.6/mM cm). The assay was carried out in 100 mM Na-phosphate buffer pH 6.5, 1.5 mM CDNB, 1 mM GSH. Glutathione reductase (EC 1.6.4.2) activity was measured by following the oxidation of NADPH at 340 nm during the reduction of GSSG (ε = 6.22/mM cm). The assay conditions were 100 mM Na-phosphate buffer pH 7.0, 1 mM GSSG and 60 ␮M NADPH. Levels of total glutathione in the liver were measured after homogenization (1:5, w/v ratio) in 5% sulphosalicilic acid with 4 mM EDTA. Samples were maintained for 45 min on ice and centrifuged at 37,000 × g for 15 min. The resulting supernatants were enzymatically assayed as reported in Regoli et al. (2005a). Protein concentrations were determined by the Lowry method with Bovine Serum Albumin (BSA) as standard. 2.8. Total oxyradical scavenging capacity For the measurement of the Total Oxyradical Scavenging Capacity (TOSC), livers were homogenized as previously described for other enzymatic analyses except that PMSF and bacitracin were not added in the buffer. The TOSC assay measures the capability of cellular antioxidants to inhibit the oxidation of 0.2 mM ␣-keto-␥-methiolbutyric acid (KMBA) to ethylene gas in the presence of different forms of oxyradicals, artificially generated at constant rate (Regoli and Winston, 1999). Peroxyl radicals (ROO·) were generated by the thermal homolysis of 20 mM 2-2 -azo-bis-(2 methylpropionamidine)dihydrochloride (ABAP) in 100 mM potassium phosphate buffer, pH 7.4. Hydroxyl radicals were generated from the Fenton reaction of iron-EDTA (1.8 ␮M Fe3+ , 3.6 ␮M EDTA) plus ascorbate (180 ␮M) in 100 mM potassium phosphate buffer. Ethylene formation in control and sample reactions was analysed at 10–12 min time intervals by gas-chromatographic analyses according to Regoli and Winston (1999). TOSC values were quantified from the equation: TOSC = 100−(ʃSA/ʃCA × 100), where ʃSA and ʃCA are the integrated areas calculated under the kinetic curve produced during the reaction course for respective sample (SA) and control (CA) reactions. For all

the samples, a specific TOSC (normalized to the content of protein) was calculated by diving the experimental TOSC values by the relative protein concentration contained in the assay. 2.9. Statistical analyses Bioaccumulation and biological responses in organisms exposed to chemicals, dosed alone or in combination, were compared by analysis of variance, ANOVA (level of significance at p < 0.05); homogeneity of variance was tested by Cochran C, and post hoc comparison (Newman-Keuls) was used to discriminate between means of values. 3. Results Trace metals significantly increased in livers of T. bernacchii injected with these elements (Fig. 1). Measured concentrations were 155 ± 58.1 ␮g/g for Cd, 86.9 ± 47.0 ␮g/g for Cu, 3.44 ± 1.51 ␮g/g for Hg, 11.1 ± 6.35 ␮g/g for Ni and 8.57 ± 2.42 ␮g/g for Pb. Co-exposures revealed that BaP did not influence the levels of metals and similar concentrations were measured in organisms exposed only to metals and co-treated with the PAH (Fig. 1). Also benzo[a]pyrene markedly increased in livers of exposed fish (33.3 ± 13.5 ␮g/g) but significant effects of various metals were observed during co-exposures; concentrations of BaP were enhanced in organisms co-treated with cadmium, while reduced by mercury, lead, and markedly by nickel and copper (Fig. 1). Clear interactions of trace metals on the metabolism of BaP were obtained on the biotransformation pathway of Cytochrome P450. Metals dosed alone did not cause any significant variation of hepatic EROD activity (not shown) which was significantly induced by BaP (Fig. 2A). EROD activation after BaP injection was suppressed by more than 80% in organisms co-exposed with cadmium and copper, while those treated with BaP and other metals (Hg, Ni, Pb) showed an EROD activity comparable to organisms exposed only to BaP (Fig. 2A). Compared to the marked increase of CYP1A1 protein content caused by BaP, western blotting and image analyses showed a reduction of approximately 45% of immunoreactive proteins only in organisms co-exposed with cadmium and copper and no significant effects during co-exposures with Hg, Ni, and Pb (Fig. 2B). Analyses at the transcriptional level were carried out only for organisms exposed to BaP and co-exposed with Cd or Cu. While BaP induced a 140-fold increase of hepatic CYP1A mRNA (Fig. 3), this effect was reduced by 30% in organisms co-exposed with cadmium. On the contrary, copper did not significantly influence CYP1A gene expression, and levels of mRNA were similar to those induced only by BaP (Fig. 3). Biliary BaP type metabolites were detectable in fish exposed to BaP, dosed alone or in combination with various metals, and without significant differences between these groups of exposure (Fig. 4). The antioxidant status was assessed as individual antioxidants and total oxyradical scavenging capacity (TOSC) toward

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

171

Fig. 1. Concentrations of trace metals and BaP (␮g/g dry weight) in liver of Trematomus bernacchii under various experimental conditions (mean value ± standard deviations, n = 10). The p-values are reported for significant variations, and different letters indicate significant differences between groups of means (post hoc comparison); b.d.l.: below detection limit.

hydroxyl radicals and peroxyl radicals (Table 1). Significant variations of oxidative parameters were observed in almost all experimental groups, but the magnitude and number of such changes greatly differed for exposures and co-exposures to various chemicals. Moderate oxidative effects were observed for BaP dosed alone, which only caused a significant increase of glutathione peroxidases. Fish exposed to cadmium exhibited more evident variations of glutathione metabolism with significant alterations of levels of total glutathione, glutathione peroxidases and glutathione S-transferases. These effects, however, did not reflect a more general imbalance of oxyradical metabolism since the scavenging capacity toward hydroxyl and peroxyl radicals remained almost similar to control organisms.

Fig. 2. Enzymatic activity of ethoxyresorufin O-deethylase (EROD) (A) and the levels of Cytochrome P4501A (CYP1A1) proteins (B) in the liver of Trematomus bernacchii under various experimental conditions. Results are given as the mean values ± standard deviation (n = 10). The p-values are reported for significant variations, and different letters indicate significant differences between groups of means (post hoc comparison).

Fig. 3. Levels of CYP1A1 expression in the liver of Trematomus bernacchii under various experimental conditions. Results are expressed as ratio between intensity of amplification products of CYP1A1 and ␤-actin mRNA (mean value ± standard deviations, n = 10). The p-values are reported for significant variations, and different letters indicate significant differences between groups of means (post hoc comparison).

Fig. 4. Levels of biliary BaP-type metabolites in the liver of Trematomus bernacchii under various experimental conditions. Results are given as the mean value ± standard deviation (n = 10). The p-values are reported for significant variations, and different letters indicate significant differences between groups of means (post hoc comparison); b.d.l: below detection limit.

Asterisks indicate values significantly different from control group (p < 0.05); dots indicate values significantly different from benzo[a]pyrene exposure group; letter “a” indicate values (for co-exposed only) significantly different from metals exposure group.

28.6 37.3 24.2 48.4*·a 20.2 58.6 42.6 16.3 38.2 32.8 28.3 26.1 ± ± ± ± ± ± ± ± ± ± ± ± 385 447 377 275 407 343 372 349 349 452 413 375 10.6 20.6 21.2 6.86*· 18.7 6.26* 2.79 1.87 14.6 8.90 8.46 8.81·a ± ± ± ± ± ± ± ± ± ± ± ± 67.1 76.4 52.2 37.4 56.3 35.8 56.1 53.6 41.7 53.6 65.4 46.4 1.05 1.15 2.13 0.97*· 1.96 1.40 0.67 0.12 0.16 0.37 0.07 1.34 ± ± ± ± ± ± ± ± ± ± ± ± 2.80 3.09 3.24 6.03 5.04 2.72 3.22 2.36 1.74 2.59 2.73 4.76 26.3 37.7 32.2* 20.3* 24.3 13.9*·a 23.5* 18.7* 21.9* 22.3*· 22.4 58.8 ± ± ± ± ± ± ± ± ± ± ± ± 213 173 146 155 177 125 160 138 125 107 197 212 13.5 39.1* 65.3* 20.2·a 10.2 15.8 38.9* 39.1* 9.31· 26.6*a 27.5* 16.4* ± ± ± ± ± ± ± ± ± ± ± ± 72.8 146 237 73.7 80.1 93.3 169 169 52.6 114 147 101 0.09 0.40 0.42*· 0.29*·a 0.25 0.28 0.24 0.11 0.22* 0.55 0.19 0.02 ± ± ± ± ± ± ± ± ± ± ± ± 0.59 0.58 2.27 1.08 0.65 0.54 0.86 0.62 1.02 1.02 0.70 0.43 Ctrl BaP Cadmium Cd-BaP Copper Cu-BaP Mercury Hg-BaP Nickel Ni-BaP Lead Pb-BaP

TOSC-OH• (U TOSC/mg prot) Catalase (␮mol/min/mg prot) Glutathione reductase (nmol/min/mg prot) Glutathione S-transferases (nmol/min/mg prot) Glutathione peroxidases CHP (nmol/min/mg prot) Total glutathione (␮mol/g tissue) Exposure

Table 1 Biomarkers of oxidative stress in Trematomus bernacchii after laboratory exposures to different classes of contaminants (mean value ± standard deviations, n = 10)

355 ± 16.9 416 ± 24.4 292 ± 41.8· 273 ± 28.9*· 403 ± 15.5 354 ± 25.9 279 ± 19.7 335 ± 23.1 295 ± 12.0 274 ± 19.1 301 ± 22.9 232 ± 7.33*

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175 TOSC-ROO• (U TOSC/mg prot)

172

Fish co-exposed to Cd and BaP demonstrated oxidative interactions between these chemicals on glutathione metabolism parameters which, with the exception of increased glutathione reductase activity, were generally decreased/depleted in comparison to those obtained after exposures to cadmium and/or BaP alone: a significant inhibition of antioxidative capacities in organisms co-exposed to Cd and BaP was further supported by the reduced activity of catalase and decreased scavenging capacity toward both HO• and ROO• . Organisms exposed to copper did not exhibit any significant variation of oxidative parameters compared to control organisms, while a significant inhibition of glutathione S-transferases and catalase was observed after co-exposure with BaP. Mercury, nickel and lead caused a limited perturbation of oxidative status, with an increase of glutathione peroxidases and/or depletion of glutathione S-transferases, and these results were not generally influenced by co-exposure with BaP. 4. Discussion The aim of this work was to evaluate whether different trace metals can interact with metabolism of BaP, chosen as a model for polycyclic aromatic hydrocarbons. Despite the fact that i.p. injections do not reflect natural routes of exposure as either dietary or waterborne exposures, they still represent a fundamental approach to collect baseline ecotoxicological data in Antarctic species. Obtained results cannot be used to extrapolate dose-effect relationships in environmental conditions but they can be useful to investigate the mechanisms of action of toxicants at cellular levels and interactions between different classes of chemicals. In this work, concentrations of cadmium, copper, mercury, lead and nickel significantly increased in liver of exposed T. bernacchii. At the moment, there is no evidence of highly polluted areas in Terra Nova Bay, but measured tissue burdens can be considered realistic if compared with baseline concentrations of field-collected organisms. Reported data for liver of various fish species range from 2 to 50 ␮g/g (Cd), from 3.3 to 92 ␮g/g (Cu), from 0.1 to 1.51 ␮g/g (Hg) (Bargagli et al., 1996; Marquez et al., 1998; Bustamante et al., 2003). Concentrations of metals measured in organisms co-treated with BaP were similar to those observed in fish injected only with these elements. Previous investigations on T. bernacchii and European eel Anguilla anguilla had revealed that co-exposure by i.p. injection with 2,3,7,8-tetrachlorobibenzo-p-dioxin greatly enhanced bioaccumulation of cadmium; the subcellular association of this element with microsomal fraction suggested an active storage of the metal within the TCDD-induced proliferating endoplasmic reticulum (Benedetti et al., 2005; Regoli et al., 2005a). In this respect, the different effects of TCDD and BaP on metal accumulation could be due to the different affinity of these ligands toward the aryl hydrocarbon receptor (AhR); the slower kinetic activation of CYP1A1 caused by BaP in comparison with TCDD, and the consequent lower proliferation of endoplasmic reticulum would reduce the possibility to compartmentalize metals within this cellular fraction (Reddy et al., 1990; Connell et al., 1998).

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

In this study all the metals influenced metabolism of BaP, although significantly different effects were caused by various elements on bioaccumulation and biotransformation rate of BaP. Compared to organisms treated only with BaP, concentrations of this molecule increased in fish co-exposed with Cd, slightly decreased with Hg and Pb, while markedly lower values were measured after co-exposures with Ni and Cu. The induction of EROD activity, caused by benzo[a]pyrene, was greatly suppressed only by cadmium and copper, while no significant effects were observed in organisms co-exposed with Hg, Pb and Ni. Independent of the differences in EROD activities, all the treatments with BaP and metals revealed similar levels of biliary BaP-type metabolites; these results confirmed that bile metabolites can indicate a certain metabolization of PAHs even at low levels of EROD induction, but also the limited capability of this screening method to quantitatively compare different groups of exposure (Gorbi and Regoli, 2004b; Gorbi et al., 2005). Inhibitory effects of different metals, including Cd and Cu on the biotransformation pathway of organic chemicals have already been reported both in in vitro and in vivo models (Viarengo et al., 1997; Carpenter et al., 2002; Korashy and ElKadi, 2004; Oliveira et al., 2004; Regoli et al., 2005a). In vivo studies on fish (Dicentrarchus labrax and Microgradus tomcod) have demonstrated significantly reduced EROD activity after exposure to PAHs combined with simultaneous or delayed administration of these metals (Oliveira et al., 2004; Sorrentino et al., 2005), but a few contrasting results have also been reported for cadmium which increased the induction of EROD activity in the fish Dicentrarchus labrax co-exposed with benzo[a]pyrene (Lemaire-Gony et al., 1995). In our study, the inhibition of EROD in T. bernacchii coexposed to BaP and Cd was confirmed by lower levels of CYP1A-immunoreactive proteins and of CYP1A mRNAs; compared to organisms exposed only to BaP, the inhibitory effect of cadmium was 80%, 45% and 30% for the catalytic, protein and transcriptional levels, respectively. Transcriptional effects of cadmium have been observed in primary hepatocytes of black seabream and rainbow trout, with a marked inhibition of CYP1A mRNA induced by 3methylcholanthrene (Risso-de Faverney et al., 2000). In vivo experiments further demonstrated the potential of this element to influence the activity of transcription factors like nuclear factor I (NFI) and nuclear factor kB (NF-kB) with downregulation of CYP gene expression (Morel and Barouki, 1998; Barouki and Morel, 2001; Waisberg et al., 2003). Our results suggest that prooxidant mechanisms may have an important role in modulating the inhibitory effects of cadmium on CYP system. During co-exposures to this metal and BaP, glutathione reductase activity was significantly higher while other parameters of glutathione metabolism were generally lowered compared to treatments with chemicals dosed alone. Furthermore, the significant inhibition of catalase activity and the reduced oxyradical scavenging capacity toward peroxyl and hydroxyl radicals, indicated these organisms as more susceptible to oxidative damages. CYP1A genes can be responsive to oxidative stress and accumulation of mRNA is prevented by reactive oxygen species at the transcriptional level (Barker et

173

al., 1994); ROS and in particular hydrogen peroxide (H2 O2 ), are also known to down-regulate CYP1A expression acting in the signalling pathway through the stress response transcription factor, NF-kB and NFI (Morel and Barouki, 1998, 1999; Barouki and Morel, 2001). The protein content and the catalytic activities of CYP enzymes can be reduced when ROS act as second messengers leading to activation of kinases with phosphorylation of CYP450 proteins or when they interact with the iron of heme groups (Lowe et al., 1998). Due to its affinity for SH-residues, indirect prooxidant effects of cadmium can further influence the assemblage, structural integrity and functions of CYP enzymes, e.g. binding to the thiol group of CYP cysteines or to glutathione. The induction of heme oxygenase-1 (HO-1), responsible for the degradation of heme to the antioxidant biliverdin, reduces the bioavailability of this prosthetic group and has been indicated as an important pathway for decreased levels of CYP1A proteins (Elbekai and El-Kadi, 2004; Korashy and El-Kadi, 2004). Although the activation of HO-1 was not measured in our study, this mechanism is worthy of future investigations. Compared to cadmium, copper also significantly inhibited EROD activity and CYP1A protein content during co-exposures with BaP, but no effects were observed at the transcriptional level. Although copper induced quite limited oxidative changes, worthy to note was the significant inhibition of catalase activity in organisms co-treated with BaP and Cu. In this respect the reduced protein content and enzymatic activity of CYP450 could be ascribed to post-transcriptional effects caused by higher levels of hydrogen peroxide, and/or the capability of Cu to bind to –SH groups (Viarengo, 1997; Kim et al., 2002; Oliveira et al., 2004), and/or oxidative mechanisms limiting the availability of heme for both catalase and CYP450 proteins (Korashy and El-Kadi, 2004). Mercury, nickel and lead appeared less bio-active than Cd and Cu, suggesting a different metabolic pathway, i.e. their storage and elimination through lysosomes. Oxidative effects of these elements were limited to the induction of glutathione peroxidases or inhibition of glutathione S-transferases and, with a few exceptions, oxidative perturbations were similar after co-exposures with BaP. The lack of inhibitory effects on the efficiency of Cytochrome P450 would further suggest the role of synergistic prooxidant mechanisms during co-exposures. The overall results of this study demonstrated significant interactions between metabolism of trace metals and aromatic xenobiotics, with implications for monitoring both bioaccumulation and biological effects in key sentinel species. Oxidative pathways would be determinant in modulating biotransformation of aromatic xenobiotics at transcriptional, translational and catalytic levels, with different mechanisms for various elements responding to specific signals, interactions, and indirect or cascade effects difficult to predict for various chemicals or mixtures. In particular, the evident effects of cadmium are of great importance for the Antarctic area of Terra Nova Bay, due to the natural enrichment of this element. The increase of BaP concentrations and the limited metabolism of this aromatic compound in coexposed fish, suggest that local environmental features at Terra Nova Bay can influence the susceptibility of fish to bioaccumulation and toxic effects of PAHs potentially released from local

174

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175

anthropogenic activities, accidental oil spills, or atmospheric deposition. Acknowledgment This study was financially supported by the Italian National Program on Antarctic Research (PNRA). References Bargagli, R., Nelli, L., Ancora, S., Focardi, S., 1996. Elevated cadmium accumulation in marine organisms from Terra Nova Bay (Antarctica). Polar Biol. 16, 513–520. Barker, C.W., Fargan, J.B., Pasco, D.S., 1994. Down-regulation of P450 1A1 and P450 1A2 in isolated hepatocytes by oxidative stress. J. Biol. Chem. 269, 3985–3990. Barouki, R., Morel, Y., 2001. Repression of cytochrome P450 1A1 gene expression by oxidative stress: mechanisms and biological implications. Biochem. Pharmacol. 61, 511–516. Benedetti, M., Gorbi, S., Bocchetti, R., Fattorini, D., Notti, A., Martuccio, G., Nigro, M., Regoli, F., 2005. Characterization of cytochrome P450 in the Antarctic key sentinel species Trematomus bernacchii. Pharmacologyonline 3 (1-8), ISSN-1827-8620. Bustamante, P., Bocher, P., Ch´erel, Y., Miramand, P., Caurant, F., 2003. Distribution of trace elements in the tissues of benthic and pelagic fish from the Kerguelen Islands. Sci. Total Environ. 313, 25–39. Canapa, A., Barucca, M., Gorbi, S., Benedetti, M., Zucchi, S., Biscotti, M.A., Olmo, E., Nigro, M., Regoli, F., 2007. Vitellogenin gene expression in males of the Antarctic fish Trematomus bernacchii from Terra Nova Bay (Ross Sea): a role for environmental cadmium? Chemosphere 66, 1270– 1277. Carpenter, D.O., Arcaro, K., Spink, D.C., 2002. Understanding the human health effects of chemical mixtures. Environ. Health Perspect. 110, 25–42. Connell, B.J., Singh, A., Chu, I., 1998. Interaction of PCB congeners and 2,3,7, 8-TCDD in the rat liver: an electron microscope study. J. Submicrosc. Cytol. Pathol. 30, 157–163. Elbekai, R.H., El-Kadi, A.O.S., 2004. Modulation of aryl hydrocarbon receptorregulated gene expression by arsenite, cadmium, and chromium. Toxicology 202, 249–269. Elbekai, R.H., El-Kadi, A.O.S., 2005. The role of oxidative stress in the modulation of aryl hydrocarbon receptor-regulated genes by As3+ , Cd2+ , and Cr6+ . Free Radic. Biol. Med. 39, 1499–1511. Focardi, S., Fossi, M.C., Lari, L., Casini, S., Leonzio, C., Meidel, S.K., Nigro, M., 1995. Induction of MFO activity in the Antarctic fish Pagothenia bernacchii: preliminary results. Mar. Environ. Res. 39, 97–100. Frenzilli, G., Nigro, M., Scarcelli, V., Gorbi, S., Regoli, F., 2001. DNA integrity and total oxyradical scavenging capacity in the Mediterranean mussel Mytilus galloprovincialis: a field study in a highly eutrophicated coastal lagoon. Aquat. Toxicol. 53, 19–32. Gorbi, S., Baldini, C., Regoli, F., 2005. Seasonal variability of metallothioneins, cytochrome P450, bile metabolites and oxyradical metabolism in the European eel Anguilla anguilla L. (Anguillidae) and striped mullet Mugil cephalus L. (Mugilidae). Arch. Environ. Contam. Toxicol. 49, 62–70. Gorbi, S., Regoli, F., 2004a. Review. Total Oxyradical scavenging capacity as an index of susceptibility to oxidative stress in marine organisms. Comment Toxicol. 9, 303–322. Gorbi, S., Regoli, F., 2004b. Induction of cytochrome P4501A and biliary PAH metabolites in European eel Anguilla anguilla: seasonal, dose-and timeresponse variability in field and laboratory conditions. Mar. Environ. Res. 58, 511–513. Kim, J.S., Ahn, T., Yim, S.K., Yun, C.H., 2002. Differential effect of copper (II) on the cytochrome P450 enzymes and NADPH-cytochrome P450 reductase: inhibition of cytochrome P450-catalyzed reactions by copper (II) ion. Biochem-US 41, 9438–9447. Korashy, H.M., El-Kadi, A.O.S., 2004. Differential effects of mercury, lead and copper on the constitutive and inducible expression of aryl hydrocarbon

receptor (AHR)-regulated genes in cultured hepatoma Hepa 1c1c7 cells. Toxicology 201, 153–172. Lemaire-Gony, S., Lemaire, P., Pulsford, A.L., 1995. Effects of cadmium and benzo(a)pyrene on the immune system, gill ATPase and EROD activity of European sea bass Dicentrarchus labrax. Aquat. Toxicol. 31, 297–313. Lin, L.C., Cormier, S.M., Torsella, J.A., 1996. Fish biliary polycyclic aromatic hydrocarbon metabolites estimated by fixed-wavelength fluorescence: comparison with HPLC-fluorescent detection. Ecotox. Environ. Saf. 35, 16–23. Lowe, G.M., Hulley, C.E., Rhodes, E.S., Young, A.J., Bilton, R.F., 1998. Free radical stimulation of tyrosine kinase and phosphatase activity in human pheripheral blood mononuclear cells. Biochem. Biophys. Res. Commun. 245, 17–22. Marquez, M., Vodopivez, C., Casaux, R., Curtosi, A., 1998. Metal (Fe, Zn, Mn and Cu) levels in the Antarctic fish Notothenia coriiceps. Polar Biol. 20, 404–408. Meyer, R.P., Podvinec, M., Meyer, U.A., 2002. Cytochrome P450 CYP1A1 accumulates in the cytosol of kidney and brain and is activated by heme. Mol. Pharmacol. 62, 1061–1067. Miller, H.C., Mills, G.N., Bembo, D.G., Macdonald, J.A., Evans, C.W., 1999. Induction of cytochrome P4501A (CYP1A) in Trematomus bernacchii as an indicator of environmental pollution in Antarctica: assessment by quantitative RT-PCR. Aquat. Toxicol. 44, 183–193. Morel, Y., Barouki, R., 1998. Down-regulation of cytochrome P4501A1 gene promoter by oxidative stress-critical contribution of nuclear factor 1. J. Biol. Chem. 273, 26969–26976. Morel, Y., Barouki, R., 1999. Repression of gene expression by oxidative stress. Biochem. J. 342, 481–496. Oliveira, M., Santos, M.A., Pacheco, M., 2004. Glutathione protects heavy metal-induced inhibition of hepatic microsomal ethoxyresorufin Odeethylase activity in Dicentrarchus labrax L. Ecotoxicol. Environ. Saf. 58, 379–385. Reddy, V., Houser, W., Bresnick, E., 1990. Binding characteristics of 4S PAHbinding protein and Ah receptor from rats and mice Raha A. J. Toxicol. Environ. Health 29, 339–355. Regoli, F., Frenzilli, G., Bocchetti, R., Annarumma, F., Scarcelli, V., Fattorini, D., Nigro, M., 2004. Time-course variations of oxyradical metabolism, DNA integrity and lysosomal stability in mussels, Mytilus galloprovincialis, during a field translocation experiment. Aquat. Toxicol. 68 (2), 167– 178. Regoli, F., Nigro, M., Benedetti, M., Gorbi, S., Pretti, C., Gervasi, P.G., Fattorini, D., 2005a. Interactions between metabolism of trace metals and xenobiotics agonists of the Ah receptor in the Antarctic fish Trematomus bernacchii: environmental perspectives. Environ. Toxicol. Chem. 24 (6), 1475– 1482. Regoli, F., Nigro, M., Benedetti, M., Fattorini, D., Gorbi, S., 2005b. Antioxidant efficiency in early life stages of the Antarctic silverfish Pleuragramma antarcticum: responsiveness to pro-oxidant conditions of platelet ice and chemical exposure. Aquat. Toxicol. 75 (1), 43–52. Regoli, F., Nigro, M., Chiantore, M., Winston, G.W., 2002. Seasonal variations of susceptibility to oxidative stress in Adamussium colbecki, a key bioindicator species for the Antarctic marine environment. Sci. Total Environ. 289, 205–211. Regoli, F., Winston, G.W., 1999. Quantification of total oxidant scavenging capacity (TOSC) of antioxidants for peroxynitrite, peroxyl radicals and hydroxyl radicals. Toxicol. Appl. Pharmacol. 156, 96–105. Risso-de Faverney, C., Lafaurie, M., Girare, J.P., Rahmani, R., 2000. The nitroxide stable radical tempo prevents metal-induced inhibition of CYP1A1 expression and induction. Toxicol. Lett. 111, 219–227. Sandvik, M., Beyer, J., Goksøyr, A., Hylland, K., Egaas, E., Skaare, J.U., 1997. Interactions of benzo[a]pyrene 2, 3,3 ,4,4 , 5-hexachlorobiphenyl (PCB156) and cadmium on biomarker responses in flounder (Platichthys flesus L.). Biomarkers 2, 153–160. Scholz, S., Gutzeit, H.O., 2000. 17-Alpha-ethinylestradiol affects reproduction, sexual differentiation and aromatase gene expression of the medaka (Oryzias latipes). Aquat. Toxicol. 50, 363–373. Seubert, J.M., Sinal, C.J., Bend, J.R., 2002. Acute sodium arsenite administration induces pulmonary CYP1A1 mRNA, protein and activity in the rat. J. Biochem. Mol. Toxicol. 16 (2), 84–95.

M. Benedetti et al. / Aquatic Toxicology 85 (2007) 167–175 Sorrentino, C., Roy, N.K., Courtenay, S.C., Wirgin, I., 2005. Co-exposure to metals modulates CYP1A mRNA inducibility in Atlantic tomcod Microgadus tomcod from two populations. Aquat. Toxicol. 75, 238–252. Stohs, S.J., Bagchi, D., 1995. Oxidative mechanisms in the toxicity of metal ions. Free Radic. Biol. Med. 18, 321–336. Towbin, H., Stachelin, P., Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 76, 4350–4354. Tully, D.B., Collins, B.J., Overstreet, J.D., Smith, C.S., Gregg, E.D., Mumtaz, M.M., Chapin, R.E., 2000. Effects of arsenic, cadmium, chromium and

175

lead on gene expression regulated by a battery of 13 different promoters in recombinant HepG2 Cells. Toxicol. Appl. Pharm. 168, 79–90. Vacchi, M., Williams, R., La Mesa, M., 1996. Reproduction in three species of fish from the Ross Sea and Mawson Sea. Antarct. Sci. 8, 185–192. Viarengo, A., Bettela, E., Fabbri, R., Burlando, B., Lafaurie, M., 1997. Heavy metal inhibition of EROD activity in liver microsomes from the bass Dicentrarchus labrax exposed to organic xenobiotics: role of GSH in the reduction of heavy metal effects. Mar. Environ. Res. 44, 1–11. Waisberg, M., Joseph, P., Hale, B., Beyersmann, D., 2003. Molecular and cellular mechanisms of cadmium carcinogenesis. Toxicolology 192, 95–117.