当前位置: 首页 > 期刊 > 《毒物学科学杂志》 > 2005年第1期 > 正文
编号:11294682
A Bio-Effect Directed Fractionation Study for Toxicological and Chemical Characterization of Organic Compounds in Bottom Sediment
http://www.100md.com 《毒物学科学杂志》
     Department of Applied Environmental Research (ITM), Stockholm University, SE-106 91 Stockholm, Sweden

    County Administrative Board of Stockholm, Box 22067, SE-104 22 Stockholm, Sweden

    ABSTRACT

    The major aim of this study was to characterize toxic organic compounds in bottom sediments from a PCB polluted bay. To overcome difficulties in pinpointing toxicants in complex environmental samples we applied a bio-effect directed (BED) fractionation approach and investigated the relationships between aromaticity, teratogenicity, and aryl hydrocarbon receptor (AhR) mediated toxicity. Hepatic ethoxyresorufin O-deethylase (EROD) activities and malformations were investigated in rainbow trout (Oncorhynchus mykiss) larvae exposed by injecting sediment extract and fractions (separated by their degree of aromaticity) thereof into newly fertilized eggs. Our results imply that non-additive effects get more pronounced the more complex the exposure. The fraction mainly composed of dicyclic aromatic compounds (DACs), including PCBs, was surprisingly less teratogenic than the fraction mainly composed of polycyclic aromatic compounds (PACs). A major part of the latter potential was isolated in a subfraction mainly composed of three- and four-ring compounds (including alkylated and sulphur-heterocyclic compounds). Though no clear relationship between aromaticity and EROD induction was observed, both the DAC- and the PAC-fractions contributed equally to the EROD induction potential. A major part of the PAC-fraction's induction potential came from a subfraction containing compounds with more than five rings. No clear relationship between teratogenicity and EROD induction was observed, underlining the need for a battery of biomarkers in estimating environmental risk. Two specific malformations not previously described in literature—asymmetric yolk sac and fin edema—could be tracked through the fractionation steps, suggesting that this BED-fractionation strategy is a reliable tool for pinpointing toxic compounds in the environment.

    Key Words: fish; TIE; straight phase HPLC; blue sac disease; pulp mill; EROD.

    INTRODUCTION

    Estimating the environmental risk posed by contaminants in a given location is a difficult challenge. Organisms in the environment are exposed to complex mixtures of compounds, resulting in interactive effects and unknown mechanisms of toxicity, which often complicate the search for etiologic compounds. Furthermore, estimating the toxicity by chemical analyses of environmental samples is difficult, since unknown toxic compounds will not be found.

    Difficulties in pinpointing toxic compounds in complex environmental samples may be circumvented by fractionating samples in several steps by guidance of the isolated fractions' toxicities, leading to a more reliable identification of etiologic compounds than the analysis of a non-fractionated sample. This approach has been applied since the late 1960's (e.g., Das et al., 1969) for identification of toxic and mutagenic compounds (Schuetzle and Lewtas, 1986), and is in many respects similar to toxicity identification evaluations (TIE), reviewed by Burgess (2000). Most studies applying this approach have used invertebrates, algae, or in vitro cell systems (Brack, 2003; Burgess, 2000; Schuetzle and Lewtas, 1986). By use of intact vertebrates—with all its biological complexity—a broader toxicological evaluation may be facilitated.

    Early developmental stages of organisms are generally more sensitive to environmental factors than juvenile and adult stages. Early life-stages of fish are therefore, due to a well-characterized highly developed physiology, frequently used as biological tools for evaluating toxicity. Blue sac disease, named after the invasive yolk sac edema found among salmonid larvae in hatcheries (Wolf, 1969), associates with circulatory defects and can be induced by hypoxia, temperature shock, ammonia exposure, and a number of pollutants (Spitsbergen et al., 1991). Today, blue sac disease also includes other symptoms: edemas of the pericardium, subcutaneous hemorrhages (Brinkworth et al., 2003), as well as craniofacial (Billiard et al., 1999; Spitsbergen et al., 1991) and spinal malformations (Carls et al., 1999; Heintz et al., 1999; Incardona et al., 2004). Symptoms are generally regarded as mediated by the aryl hydrocarbon receptor (AhR) when caused by pollutants, since dicyclic aromatic compounds (DACs) with known AhR affinity—including polychlorinated naphthalenes (PCNs) (Villalobos et al., 2000), biphenyls (PCBs), dibenzo-p-dioxins (PCDDs), and dibenzofurans (PCDFs)—are the most potent, whereas DACs with no AhR affinity do not cause these malformations (Walker and Peterson, 1991). Evidence of the link between the AhR and teratogenicity has been shown in a number of investigations (Brinkworth et al., 2003; Cantrell et al., 1996, 1998; Guiney et al., 1997). Though polycyclic aromatic compounds (PACs) cause similar malformations (Carls et al., 1999; Heintz et al., 1999) the link between teratogenicity and the AhR has been demonstrated only for retene (7-isopropyl-1-methylphenanthrene); symptoms correlated with the AhR-mediated cytochrome P-450 1A (CYP 1A) induction, measured as ethoxyresorufin O-deethylase (EROD) induction (Billiard et al., 1999) or by immunohistochemistry (Brinkworth et al., 2003). Exposure to mixtures of polycyclic aromatic hydrocarbons (PAHs), a sub-class of PACs (Engwall et al., 1994) or benzo[a]pyrene (Wassenberg et al., 2002) provides no strong relationship between CYP 1A induction and malformations, suggesting involvement of alternate mechanisms.

    In a previous study (Sundberg et al., in press) we investigated EROD induction in rainbow trout (Oncorhynchus mykiss) larvae. Fish eggs were exposed to organic extracts of abiotic matrices from a PCB polluted bay in Sweden using the nanoinjection technique (kerman and Balk, 1995)—an exposure technique that mimics maternal transfer (Walker et al., 1996). Prior to exposure three fractions designated by their main components were isolated from the extract: (1) aliphatic and monocyclic aromatic compounds (MACs), (2) DACs, and (3) PACs. Though PCBs and probably other potent AhR agonists were isolated in the DAC-fraction, the major part of the EROD induction was caused by the PAC-fraction.

    The current study was designed to (1) compare potential toxicity and PAH pollution in sediments from the polluted bay and a reference bay, and (2) to further characterize toxic compounds in the PAC-fraction from the polluted bay by investigating the relationships between (i) aromaticity and teratogenicity, (ii) aromaticity and AhR-mediated toxicity, and (iii) AhR-mediated toxicity and teratogenicity. This is, to the best of our knowledge, the first time these relationships have been investigated using intact vertebrates and a complex environmental sample. In addition to chemical analysis of the sediments, we conducted two independent nanoinjection experiments, investigating EROD induction and malformations in exposed rainbow trout larvae. Exposure solutions for Experiment 1 were extracts of sediments from the two bays; and, for Experiment 2, fractions of the polluted extract obtained by a bio-effect directed (BED) fractionation approach that separated the compounds predominantly according to aromaticity.

    MATERIALS AND METHODS

    Chemicals. Toluene (p.a. grade), n-hexane (LiChrosolv), cyclohexane (p.a. grade), silica gel (60 puriss), anhydrous sodium sulphate (p.a. grade) and copper (p.a. grade, fine powder <63 μm) were obtained from Merck (Darmstadt, Germany). Dimethylformamide (DMF; p.a. grade) was obtained from Riedel-de Han (Seelze, Germany). Water (HiPerSolve) was purchased from BDH Laboratory Supplies (Poole, UK). NADPH (N-0505, 98%), Bovine serum albumin (A-7030), Triolein (T-7140, 99%), sucrose (p.a. grade), ethoxyresorufin (E-3763), resorufin (R-3257) and tris-HCL (p.a. grade) were purchased from Sigma (St. Louis, MO). Agarose gel was bought from Pharmacia (Uppsala, Sweden). Native and deuterium-labelled PAH standards used for chemical analysis were obtained from Cambridge Isotope Laboratories, Inc. (Andover, MA). Benzo[a]pyrene used as positive control in the exposure experiments was bought from Fluka Chemie GmbH (Buchs, Switzerland).

    Sediment sampling. To obtain a large quantity of each fraction (Experiment 2, see below) we used sediment since it is easy to collect and contains the highest concentrations of organic compounds. Surface bottom sediment samples (top 10 cm) were collected from the PCB polluted bay rserumsviken, located southeast of the city of Vstervik and from the reference bay Slingsviken, located in an agricultural area 15 km south of the polluted bay (for positions see Table 1). Ten sediment samples were collected on 15 October 1999 at a water depth of 1–2 m from the inner part of the polluted bay, using a gravity core sampler (Limnos Ltd., Turku, Finland). In the middle part of the reference bay, seven sediment samples were collected on 11 January 2002 at a water depth of 3–4 m using a Kajak-type gravity corer (Blomqvist and Abrahamsson, 1985). Sediment samples were stored for two weeks at –20°C in pre-cleaned polypropylene jars with lids prior to extraction (the sediment sample from the polluted bay was extracted again two years later for Experiment 1).

    Extraction and fractionation of sediment samples—isolation of exposure solutions. The extraction and fractionation procedure is presented as a flow chart in Figure 1. Sediment samples were thawed at room temperature, and, to get a representative sample, samples from the same bay were pooled in a glass beaker and mixed with a stainless steel spoon. Aliquots of wet sediment samples were extracted for 35–50 h in toluene using a Soxhlet apparatus coupled to a Dean-Stark trap for water removal (Lamparski and Nestrick, 1980). To investigate possible biological and chemical interference from the solvents, Experiment 2 included a solvent control that was extracted and fractionated the same manner as the sediment sample from the polluted bay. Dry weights were gravimetrically determined after extraction. In total were 21 g (Experiment 1) and 65 g (Experiment 2) of the polluted sediment and 31 g (Experiment 1) of the reference sediment extracted. Extracts were cleaned on a deactivated (10% H2O) silica gel column (I.D. 1 cm, 10 g of silica gel/g extractable organic matter) with n-hexane (10 ml of n-hexane/g silica gel) as eluent. Elemental sulphur was removed by the addition of a small amount of elemental copper, ultrasonification-bath (4 x 15 min) and overnight incubation at room temperature. Total extracts were obtained after a final clean-up with a silica gel column topped with 1 cm of sodium sulphate and with n-hexane as eluent.

    In Experiment 2 the total extract from the polluted bay was fractionated at room temperature (21–22°C) on a semi-preparative aminopropylsililated silica column (NH2, μBondapak, Waters, Milford, MA, 7.8 x 300 mm, 10 μm) by using an automated HPLC system (Zebühr et al., 1993), with a UV-VIS detector operating at 254 nm and with n-hexane as mobile phase. This well-defined straight phase fractionation technique separates chemicals predominantly according to their number of aromatic rings (Zebühr et al., 1993). But because of the complexity of innumerable compounds in environmental samples, it is not possible to identify all the compounds that might be found in a specific fraction. The retention times for toluene, anthracene, and phenanthrene (flow rate: 3 ml/min) were used to define three fractions, designated by their main components and eluting in the following order: (1) MAC-fraction, (2) DAC-fraction, and (3) PAC-fraction. The latter fraction was collected by back-flushing the column at a flow rate of 5 ml/min (total duration of one cycle: 30 min). Subfractionation of the PAC-fraction was carried out on the same HPLC system. The first nine fractions were collected at equal time intervals (approximately 5 min) at a flow-rate of 2 ml/min using the observed retention times for chrysene and perylene as markers. The tenth fraction was isolated by back-flushing the column at a flow rate of 5 ml/min and collecting the UV-detected peak (total duration of one cycle: 110 min).

    PAC analyses. Phenanthrene-d10, fluoranthene-d10, pyrene-d10, benzo[a]pyrene-d12, and benzo[ghi]perylene-d12 were used as internal standards and added to aliquots of the total extracts from the two bays (Experiment 1) and of subfraction II (Experiment 2). Prior to PAH analysis in Experiment 1, the aliquots were cleaned up using liquid-liquid extraction with cyclohexane and DMF (stman and Colmsj, 1987). Retention times and response factors from a standard mixture containing the 27 analyzed PAHs and internal standards were used for peak identification and quantification of 27 PAHs (Table 2). Analyses were performed on a Fisons GC 8000 (Thermo Electron Corp., Waltham, MA) equipped with a PTE-5 capillary column (30 m x 0.25 mm, 0.25 μm film thickness; Supelco, Bellefonte, PA) coupled to a Fisons MD800 at SIM mode.

    Subfraction II, which caused a major part of the toxic effects (Experiment 2, see below), was further analyzed on the GC/MS system as above. Analyses were performed in full-scan mode (m/z: 50–500) and peaks were identified by: (1) comparing the observed retention times with those of standard PAHs injected on the same day, (2) comparing retention times with retention index (Lee et al., 1979), and (3) comparing mass spectra of unknown peaks with library spectra (National Institute of Standards and Technology [NIST]). Individual concentrations were calculated from the intensities of the molecular ion of the compounds in the sample with those of the standards. When no authentic standard were available, concentrations were calculated from structurally most similar available standards.

    Fish eggs. Two independent fish exposure experiments were performed, both permitted (N 153/00) by the Swedish Animal Welfare Agency. To increase the homogeneity of the biological material, only one family pair was used on each occasion. Eggs and seminal fluid from rainbow trout were collected and shipped to our laboratory from Klarne fiskerifrsksstation (Klarne, Sweden, Experiment 1) and from lvdalslax (lvdalen, Sweden, Experiment 2) on the day of fertilization. Artificial fertilization and water swelling were performed at 6.5°C, whereupon eggs were placed in cup-shaped depressions in 1% agarose gel cast in square petri dishes (Falcon, Becton Dickinson Labware, Franklin Lake, NJ) with maximum 36 eggs/dish (kerman and Balk, 1995).

    Nanoinjection of exposure solutions in rainbow trout eggs. The exposure solutions, their corresponding solvent controls, and benzo[a]pyrene (positive controls) were dissolved in the carrier substance, triolein. Two additional controls were used: carrier controls (eggs exposed solely to triolein) and uninjected controls. Groups and doses of the two experiments are presented in Table 1. The day after fertilization, rainbow trout eggs were exposed to the solutions using the nanoinjection technique (kerman and Balk, 1995; Walker et al., 1996). Briefly, triolein-dissolved exposure solutions were transferred, using a vacuum suction pump, into aluminium silicate capillaries (Sutter Instrument Co., Novato, CA) with sharp elliptical tips. Exposures were performed under a stereo microscope (Leica MZ8, Leica Microscopy and Scientific Instruments Group, Heerburg, Switzerland) with the aid of a one-dimensional hydraulic manipulator (Narishige, Tokyo, Japan) and a Pico-injector (Medical Syst. Corp., Greenvale, NY). After the penetration of the chorion and the vitelline membrane, the solutions were injected into the yolk of the egg. Before injection, each capillary was individually calibrated by adjusting the nitrogen gas pressure and the time of injection so that the diameter of the triolein droplet corresponded to the desired dose (less than 1% [v/v] of the egg volume).

    Fish maintenance. Petri dishes were kept dark in individual flow-though aquariums, containing 2 l of Stockholm municipal drinking water filtered in three consecutive steps (nominal pore size: 50 μm – active carbon – 10 μm) and aerated. The pH ranged from 8.0 to 8.2 and the water temperature from 6.5 to 9.4°C. Temperature was recorded every second or third day along with the removal of dead eggs and larvae. Malformations in newly hatched larvae were investigated under the stereomicroscope. The fish exposure experiments were terminated by severing the spinal column of the larvae when they had consumed two-thirds of their yolk content (expressed in degree-days [°C] post-hatch: 230 to 240 in Experiment 1 and 180 to 210 in Experiment 2). Individual lengths were recorded, livers were dissected, pooled (three larvae correspond to n = 1 for the enzymatic analyses), homogenized with 300 μl 0.25 M sucrose and immediately portioned into ice-kept cryo vials and plunged into liquid nitrogen as previously described (Sundberg et al., in press). The number of dissected livers in the different exposure groups and the control groups ranged from 12 to 15 (n = 4–5).

    Measurement of EROD activity. Cytochrome P-450 1A activity, measured as EROD activity, was analyzed based on the method described by Prough et al. (1978) in a Jasco FP-777 spectrofluorometer (Japan Spectroscopic Co., Ltd., Tokyo, Japan). Measurements (emission intensities at = 583 with excitation = 530) were performed at room temperature (21–22°C) in glass cells with a final volume of 2 ml as previously described (Sundberg et al., in press). In short, ethoxyresorufin and resorufin were dissolved in 0.1 M tris-HCl buffer, pH = 7.8 and were kept dark at 4°C. A fresh solution of NADPH in distilled water was made each day of measurement and kept on ice. One liver homogenate at a time was rapidly thawed in running tap water (30°C) using a Vortex mixer to prevent any temperature rise in the sample. The thawed homogenate was put on ice until measurement, which was performed within a few minutes. Background intensities, in cells containing 10 to 30 μl liver homogenate and 2.5 μM (final concentration) ethoxyresorufin, were recorded for each sample before adding NADPH to a final concentration of 50 μM to start the enzymatic reaction. All samples were measured in duplicate, and the linear relationship between protein content and enzyme activity was regularly tested. Protein was quantified in liver homogenates according to Lowry et al. (1951) and the enzyme activities of the samples, expressed as pmol resorufin x mg protein–1 x min–1, were calculated from the measured values of the standard buffer containing resorufin.

    Statistics. Statistical differences in malformations and mortalities between exposure and control groups were determined using Chi-square analysis. Parametric tests (ANOVA, Tukey HSD) were used to investigate statistical differences in larval length. Statistical differences in EROD activities were investigated using non-parametric tests (Kruskall-Wallis, Dunn's Multiple Comparison). As -level for statistically significant differences between groups a p-value of less than 0.05 was used.

    RESULTS

    PAC Concentration

    PAH concentrations in sediment samples from the reference and the polluted bay are expressed as μg/kg dry sediment (Table 2). Samples from the polluted bay were found to have 2 to 24 times higher PAH concentrations as samples from the reference bay, with the exception of perylene, which was twice as high in the reference bay than in the polluted bay. The sum of PAH concentration was approximately four times higher in the polluted bay than in the reference bay.

    Table 3 shows the concentrations of PACs in subfraction II. Half of fluoranthene and of the three methylpyrenes, one-fourth of pyrene and of benzo[a]fluorene and 4% of benzo[b]fluorene were isolated in this subfraction. In addition, high concentrations of alkylated (C1 to C4) phenanthrene, fluoranthene, and three alkylation degrees (C0 to C2) of the thiaarene benzonaphtothiophene were found in this subfraction, totalling 3.1 mg PAC/kg sediment. The identified compounds represent 85% of the total amount in the sample, if assuming similar response factors for the unidentified compounds.

    Mortalities and Malformations in Larvae

    To investigate malformations among newly hatched larvae at similar developmental stages, and, to minimize a possible bias due to regenerating malformations, they were recorded within two days. Hence, due to large sample sizes only the larvae exposed to the highest or intermediate doses were examined. Malformations in newly hatched larvae (Figs. 2A–2E) included hemorrhages in the yolk sac, head, trunk and tail region; edemas in the yolk sac, pericardium and fins; and asymmetry of the medio-lateral axis of the yolk sac (asymmetric yolk sac). The three control groups—carrier control, uninjected control, and solvent control—did not differ statistically in larval length; exposure groups were therefore compared with carrier controls for the statistical analyses. Larval length at dissection ranged from 23.0 to 27.0 mm in Experiment 1 and 22.8 to 28.3 mm in Experiment 2. Statistical differences for this endpoint were only observed in Experiment 2. Larvae exposed to the highest doses of subfraction I and IV were significantly shorter than controls (p < 0.05, Tukey HSD). Larvae exposed to the lowest and the intermediate dose of the polluted bay's total extract, the intermediate doses of subfraction III, VI and VIII were significantly longer than controls (p < 0.05, Tukey HSD).

    The pie charts (Figs. 3A–3C) illustrate the percentages of normal and malformed larvae, and of mortalities in the egg and larval stage at the time of recording. In some groups larvae died after malformations were recorded. Hence, cumulative larval mortalities throughout the entire animal experiment were used for statistical analyses. The bar charts (Figs. 3A–3C) show percentages of larvae suffering from edemas, hemorrhages, or asymmetric yolk sac.

    Experiment 1, when larvae were exposed to total extracts from the two bays, is presented in Figure 3A. Less than 7% of the carrier control larvae were malformed and less than 10% died during the egg stage. Dose-response relationships were evident for percentages of malformed larvae exposed to the total extracts from the two bays. The highest dose (74 g sediment/kg) of the total extract from the reference bay caused a significant increase of edemas, while the polluted bay's total extract caused a higher percentage of malformed larvae, suffering mainly from edemas at the intermediate dose (15 g sediment/kg egg). Fin edemas were found in larvae exposed to the highest dose (2900 μg/kg egg) of benzo[a]pyrene (3.6%) and the highest doses of the total extracts from the reference bay (3.2%) and the polluted bay (17%). Asymmetric yolk sac were observed among larvae exposed to the highest doses of benzo[a]pyrene and the total extract from the polluted bay.

    Mortalities and malformations among larvae exposed to the highest doses of the total extract from the polluted bay and its fractions are presented in Figures 3B and 3C (Experiment 2). Carrier control larvae died during both at the egg (10%) and at the larval (5%) stage and approximately 15% of the larvae were malformed. A dose of 5000 μg benzo[a]pyrene/kg egg caused significantly elevated percentages of malformed larvae and mortalities during the egg and larval stages. All variables except egg stage mortality were significantly elevated among larvae exposed to the total extract (64 g sediment/kg egg). Larvae exposed to the MAC-fraction (67 g sediment/kg egg) appeared to be less affected than carrier control larvae in terms of malformations. The DAC-fraction (68 g sediment/kg egg) caused significantly higher percentages of malformed larvae and of larvae suffering from hemorrhages. All surviving larvae exposed to the PAC-fraction (69 g sediment/kg egg) were malformed, suffering from one or several types of malformations, resulting in that the percentage of malformed larvae were significantly higher than control. But since the larvae were in such bad condition that we didn't record how many of the larvae suffered from hemorrhages, edemas, and asymmetric yolk sac (indicated in Fig. 3B as b), statistics for these three endpoints were not calculated. Asymmetric yolk sac was caused by benzo[a]pyrene, the total extract and the PAC-fraction. Fin edemas were caused by the total extract (12%) and the PAC-fraction.

    Malformations and mortalities among larvae exposed to the highest doses (68–69 g sediment/kg egg) of the ten subfractions of the PAC-fraction (Experiment 2) are presented in Figure 3C. Exposure to the subfractions caused no significant increase of larval or egg stage mortality compared with controls. Percentages of malformed larvae and larvae suffering from hemorrhages and edemas were significantly elevated in two exposure groups: subfraction II and IX. Asymmetric yolk sac was observed among larvae exposed to subfraction II.

    EROD Induction in Larval Liver Homogenates

    In the two experiments, no statistical differences were found between EROD activities in carrier controls, uninjected controls, and solvent controls. EROD inductions in the exposure groups were therefore calculated as % of carrier control. Specific activities of the carrier controls, expressed in pmol x min–1 x mg protein–1, were 14 ± 5 (95% confidence limit) in Experiment 1 and 28 ± 9 (95% confidence limit) in Experiment 2.

    A dose-response relationship for EROD induction was evident for benzo[a]pyrene exposure in both experiments (Figs. 4A–4B). In Experiment 1 (Fig. 4A), significant EROD induction was caused by all three doses of the total extract from the polluted bay, whereas only the highest dose of the total extract from the reference bay caused significant EROD induction (p < 0.05, Dunn's multiple comparisons). To reach the same induction level caused by the lowest dose of the total extract from the polluted bay a 25-fold higher dose of the total extract from the reference bay was required.

    In Experiment 2, the highest dose of benzo[a]pyrene (5000 μg/kg egg) caused EROD induction of similar magnitude as the highest doses of the total extract and the DAC- and PAC-fractions (Fig. 4B). The induction potential of the total extract from the polluted bay was similar in both experiments. The corresponding MAC-fractions caused no EROD induction; the only effect observed was a comparatively large variance. The sum of EROD activity at all three dose-levels of the MAC-, DAC-, and PAC-fraction was, after background subtraction, higher than that of the total extract. The lowest doses of the DAC- and PAC-fractions were as potent as the total extract, whereas the intermediate and highest doses of the DAC-fraction caused the highest EROD induction (Fig. 4B). The differences between these exposure groups were not statistically significant. Among the ten subfractions (Fig. 4C), significant EROD inductions were observed in larvae exposed to subfractions II, III, IV, and X. Dose-response relationships were, however, evident for all ten subfractions. The sum of EROD activity, after background subtraction, caused by the highest doses of the subfractions was equal to the highest dose of the PAC-fraction. There were no indications that the dose-response curves reached plateau levels in the two experiments.

    DISCUSSION

    Relative Potential Toxicity and PAC Pollution

    Since the DAC-fraction, which contains PCBs and, if present, other potent EROD inducers such as PCDD/Fs (Sundberg et al., in press), contributes to half the EROD induction and to a minor part of teratogenicity (present investigation), our discussion is mainly focused on compounds in the PAC-fraction. Though the experiments were terminated when one-third of the yolk remained, it is important to consider that PACs are metabolized more rapidly than the DACs mentioned above (Whyte et al., 2000), resulting in a decreased exposure of PACs compared with DACs. This metabolism, however, may lead to a production of more toxic metabolites than their parent PACs (Whyte et al., 2000). Another important consideration is that the chemical compositions of the sediment extracts are not identical to what are maternally transferred in feral fish, but since sediment can be regarded as rather stationary we can compare the potential toxicity between the bays. Although inorganic compounds are known to cause adverse effects in organisms we focus on organic compounds; consequently, we used sediment since it contain highest concentrations of organic compounds.

    Sediment PAH concentrations in the polluted bay were 4 times higher than in the reference bay and 10 times higher than in the remote Baltic Sea (Pettersen et al., 1999). Based on mortality and malformations, the potential toxicity was more than 5 times higher in the polluted bay than in the reference bay and based on EROD induction 25 times higher. Even though the PAH pollution correlated with toxicity the difference in PAH concentrations at the two locations appears to be too low for explaining the difference in potential toxicity. The highest dose of the total extract from the polluted bay, corresponding to a dose of 39 μg benzo[a]pyrene/kg egg, was in itself, insufficient for significant EROD induction. Other potent EROD inducers, e.g., indeno[1,2,3-cd]pyrene and benzo[k]fluoranthene, were present in the sediment at similarly low concentrations, suggesting that the analyzed PAHs contributed to a minor degree of the observed EROD induction. Furthermore, in Experiment 2 we observed high concentrations of alkylated and heterocyclic compounds that might have contributed to the toxicity. The origins of the compounds in the PAC-fraction needs further investigation, though the former pulp mill industry (Axelman and Broman, 1999) is the most likely candidate since no other source of these pollutants has been identified.

    Indications of Non-Additive Interactions

    The sum of EROD activity, after background subtraction, caused by the MAC-, DAC, and PAC-fraction was higher than that of the total extract, indicating non-additive interactions from compounds in these fractions. The EROD induction in the subfractions, on the other hand, was additive, implying more pronounced interactive effects, the more complex the exposure. For malformations non-additive interactions from compounds in the MAC- and DAC-fraction appear to exist. The PAC-fraction caused more malformed larvae and than the total extract. To further investigate the apparent non-additive effects the three fractions should be recombined.

    Relationship between Aromatic Structure and Teratogenic Effects

    The finding that the PAC-fraction exhibited more teratogenic potential than the MAC- and DAC-fraction agree with previous studies using life-stages of fish. Black et al. (1983) demonstrated that mortalities and malformations in rainbow trout larvae increases with degree of aromaticity when exposing them to one- to three-ring aromatic compounds. Incardona et al. (2004) showed that zebrafish (Danio rerio) larvae suffers from cardiac dysfunction when exposed to dibenzothiophene and phenanthrene (three-ring compounds) and peripheral vascular defects, anemia and neuronal cell death when exposed to pyrene (four-ring compound) whereas naphthalene (two-ring compound) caused no teratogenic effects.

    None of the PAC-subfractions caused fin edema or increased mortality, implying that the doses of the subfractions were too low. Hence, higher doses are needed to investigate whether these endpoints are attributed to compounds of certain aromatic structure. Skeletal disorders, due to retarded growth among TCDD-exposed salmonids (Spitsbergen et al., 1991), might also get more pronounced at higher doses. Though we observed reduced length in some groups no dose-response relationship was found. The small differences in degree-days (°C) between exposure groups have apparently no effect on this variable since no significant differences were observed between carrier controls.

    The two subfractions that caused a significant increase of malformed larvae are, based on the fractionation technique (Zebühr et al., 1993), mainly composed of three- and four-rings (subfraction II) and four- and five-rings (subfraction IX). Our BED-fractionation approach as a reliable tool for pinpointing toxic chemicals is illustrated by the isolation of chemicals causing fin edemas and asymmetric yolk sac through the fractionation steps. Fin edemas could be traced to the PAC-fraction and asymmetric yolk sac to subfraction II via the PAC-fraction. Asymmetric yolk sac, to the best of our knowledge not previously described in the literature, is most likely not unique for the polluted bay since it was also caused by benzo[a]pyrene. Certain molecular properties, however, appear to be essential for asymmetric yolk sac development, e.g., containing three to five rings. In addition, the qualitative analysis of subfraction II, covering an estimated 85% of the total amount of compounds in this fraction, revealed high concentrations of PACs with three and four rings (including a thiaarene, with sulphur incorporated in the ring structure), many substituted with alkyl groups containing up to four carbons. Substitution and incorporation of elements are demonstrated to contribute to adverse effects. Alkylated PACs are more toxic to early life-stages of fish than non-alkylated homologues (Barron et al., 2004; Rice et al., 1977). Eastmond et al. (1984) showed that thiaarenes are more lethal to Daphnia magna than non-sulphur homologues. Grimmer et al. (1987) fractionated flue gas condensate and found that the major part of carcinogenity to mice was isolated in a fraction containing thiaarenes and other PACs with more than three rings. Although these two studies used other species and investigated endpoints that likely act through different mechanisms than the present study, it is interesting to note the similarities with our findings. Still, to determine which compounds cause these toxic effects, exposure to a synthetic PAC-fraction is needed.

    Relationship between Aromatic Structure and EROD Induction

    Whereas EROD induction potencies of several DACs have been studied, few PAHs and other PACs have been evaluated, though they are known EROD inducers with varying degrees of potency (Whyte et al., 2000). In our previous investigation (Sundberg et al., in press), we found that the PAC-fraction had higher EROD induction potential than the DAC-fraction; in this study, the DAC- and PAC-fractions had similar potentials for corresponding doses and samples. The PAC-fraction's higher potential, however, was more pronounced outside the bay. Furthermore, the samples were collected differently (present investigation: top 10 cm, previous investigation: top 5 cm), suggesting more potent compounds in the DAC-fraction at deeper sediment layers. Nevertheless, compounds in the PAC-fraction contributed to a significant part of the EROD induction potential, which also has been demonstrated in other locations (Brack, 2003), indicating a need for greater emphasis on these compounds.

    The subfractions, with the exception of subfraction X, had similar EROD induction potential, establishing no clear relationship between AhR-mediated toxicity and aromaticity. Subfraction X, had highest EROD inducing capacity, and contains compounds with higher aromaticity than perylene (more than five rings), e.g., indeno[1,2,3-cd]pyrene, benzo[ghi]perylene, and coronene, suggesting a major contribution to the EROD induction of these compounds. Basu et al. (2001) also found no clear relationship between aromaticity and EROD induction potential in rainbow trout exposed to PAHs dissolved in water, observing the following order of potential: (1) benzo[k]fluoranthene (four-ring), (2) benzo[b]fluoranthene (four-ring), (3) benzo[b]fluorene (three-ring), and (4) dibenzo[a,h]anthracene (five-ring). Though in vitro tests using hepatic cell lines from rainbow trout (Bols et al., 1999) support the findings of Basu et al. (2001), no other study has investigated differences in EROD induction between various PACs in salmonid species (Whyte et al., 2000). Our results suggest that the EROD induction was caused by a number of structurally different PACs and considering the rather specific affinity of the AhR (Safe, 1990), it is surprising that malformations were better than EROD induction at pinpointing the most toxic subfraction. As discussed above, there are other factors besides aromaticity that are important for biological activity, such as structure and substitution. In addition, recent investigations have shown that the AhR binding may not be as specific as formerly believed (reviewed in Denison and Heath-Pagliuso, 1998). Chemical analysis of only known pollutants is therefore insufficient for estimating the potential toxicity in highly complex environmental samples.

    Relationship between EROD Induction and Teratogenicity

    Recent studies have shown that teratogenic effects caused by a number of PAHs in early life-stages of fish are manifested through an AhR-independent pathway (Barron et al., 2004). The poor relationship between EROD induction and malformations in larvae exposed to the DAC-fraction and subfraction X support those studies. The mechanism behind the asymmetric yolk sac is not known, but we can conclude that it is not an AhR-mediated response per se, since it was caused by neither the DAC-fraction nor subfraction X—both potent EROD inducers. While the mechanism for fin edemas, also described for the first time, is probably similar to other edemas, e.g., circulatory defects (Incardona et al., 2004), the asymmetric yolk sac might occur at an earlier developmental stage.

    ACKNOWLEDGMENTS

    The municipality of Vstervik, Sweden funded this study. We thank Tracy Collier (NOAA, Seattle, WA) for providing articles in press; Peter Hodson (Queens University, Kingston, ON, Canada) for fruitful input regarding the asymmetric yolk sac and fin edema; Karin Jnson and Christer Hermansson for assistance during sampling of bottom sediment; Bodil Widell and Kerstin Grunder (Stockholm University, Sweden) for assistance during analytical work; Marsha Hanson (Stockholm University, Sweden) for revising the language.

    REFERENCES

    kerman, G., and Balk, L. (1995). A reliable and improved methodology to expose fish in the early embryonic stage. Mar. Environ. Res. 39, 155–158.

    Axelman, J., and Broman, D. (1999). Inventories and fluxes of polychlorinated biphenyls from a contaminated industrial site. Environ. Toxicol. Chem. 18, 1871–1881.

    Barron, M. G., Carls, M. G., Heintz, R., and Rice, S. D. (2004). Evaluation of fish early life-stage toxicity models of chronic embryonic exposures to complex polycyclic aromatic hydrocarbon mixtures. Toxicol. Sci. 78, 60–67.

    Basu, N., Billiard, S., Fragoso, N., Omoike, A., Tabash, S., Brown, S., and Hodson, P. (2001). Ethoxyresorufin-O-deethylase induction in trout exposed to mixtures of polycyclic aromatic hydrocarbons. Environ. Toxicol. Chem. 20, 1244–1251.

    Billiard, S. M., Querbach, K., and Hodson, P. V. (1999). Toxicity of retene to early life stages of two freshwater fish species. Environ. Toxicol. Chem. 18, 2070–2077.

    Black, J. A., Birge, W. J., Westerman, A. G., and Francis, P. C. (1983). Comparative aquatic toxicology of aromatic hydrocarbons. Fundam. Appl. Toxicol. 3, 353–358.

    Blomqvist, S., and Abrahamsson, B. (1985). An improved Kajak-type gravity core sampler for soft bottom sediments. Schweiz. Z. Hydrol. 47, 81–84.

    Bols, N. C., Schirmer, K., Joyce, E. M., Dixon, D. G., Greenberg, B. M., and Whyte, J. J. (1999). Ability of polycyclic aromatic hydrocarbons to induce 7-ethoxyresorufin-O-deethylase activity in a trout liver cell line. Ecotoxicol. Environ. Saf. 44, 118–128.

    Brack, W. (2003). Effect-directed analysis: A promising tool for the identification of organic toxicants in complex mixtures Anal. Bioanal. Chem. 377, 397–407.

    Brinkworth, L., Hodson, P., Tabash, S., and Lee, P. (2003). CYP1A induction and blue sac disease in early developmental stages of rainbow trout (Oncorhynchus mykiss) exposed to retene. J. Toxicol. Environ. Health 66A, 627–646.

    Burgess, R. M. (2000). Characterizing and identifying toxicants in marine waters: A review of marine toxicity identification evaluations (TIEs). Int. J. Environ. Pollut. 13, 2–33.

    Cantrell, S. M., Joy-Schlezinger, J., Stegeman, J. J., Tillitt, D. E., and Hannink, M. (1998). Correlation of 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced apoptotic cell death in the embryonic vasculature with embryotoxicity. Toxicol. Appl. Pharmacol. 148, 24–34.

    Cantrell, S. M., Lutz, L. H., Tillitt, D. E., and Hannink, M. (1996). Embryotoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): The embryonic vasculature is a physiological target for TCDD-induced DNA damage and apoptotic cell death in medaka (Orizias latipes). Toxicol. Appl. Pharmacol. 141, 23–34.

    Carls, M. G., Rice, S. D., and Hose, J. E. (1999). Sensitivity of fish embryos to weathered crude oil: Part I. Low-level exposure during incubation causes malformations, genetic damage, and mortality in larval Pacific herring (Clupea pallasi). Environ. Toxicol. Chem. 18, 481–493.

    Das, B. S., Reid, S. G., Betts, J. L., and Patrick, K. (1969). Tetrachloro-o-benzoquinone as a component in bleached kraft chlorination effluent toxic to young salmon. J. Fish. Res. Bd. Can. 26, 3055–3067.

    Denison, M. S., and Heath-Pagliuso, S. (1998). The Ah receptor: A regulator of the biochemical and toxicological actions of structurally diverse chemicals. Bull. Environ. Contam. Toxicol. 61, 557–568.

    Eastmond, D. A., Booth, G. M., and Lee, M. L. (1984). Toxicity, accumulation, and elimination of polycyclic aromatic sulfur heterocycles in Daphnia magna. Arch. Environ. Contam. Toxicol. 13, 105–111.

    Engwall, M., Brunstrm, B., Brewer, A., and Norrgren, L. (1994). Cytochrome P450IA induction by a coplanar PCB, a PAH mixture, and PCB-contaminated sediment extracts following microinjection of rainbow trout sac-fry. Aquat. Toxicol. 30, 311–324.

    Grimmer, G., Brune, H., Deutsch-Wenzel, R., Dettbarn, G., and Misfeld, J. (1987). Contribution of polycyclic aromatic hydrocarbons and polar polycyclic aromatic compounds to the carcinogenic impact of flue gas condensate from coal-fired residential furnaces evaluated by implantation into the rat lung. J. Natl. Cancer Inst. 78, 935–942.

    Guiney, P. D., Smolowitz, R. M., Peterson, R. E., and Stegeman, J. J. (1997). Correlation of 2,3,7,8-tetrachlorodibenzo-p-dioxin induction of cytochrome P4501A in vascular endothelium with toxicity in early life stages of lake trout. Toxicol. Appl. Pharmacol. 143, 256–273.

    Heintz, R. A., Short, J. W., and Rice, S. D. (1999). Sensitivity of fish embryos to weathered crude oil: Part II. Increased mortality of pink salmon (Oncorhynchus gorbuscha) embryos incubating downstream from weathered Exxon Valdez crude oil. Environ. Toxicol. Chem. 18, 494–503.

    Incardona, J. P., Collier, T. C., and Scholz, N. L. (2004). Defects in cardiac function precede morphological abnormalities in fish embryos exposed to polycyclic aromatic hydrocarbons. Toxicol. Appl. Pharmacol. 196, 191–205.

    Lamparski, L. L., and Nestrick, T. J. (1980). Determination of tetra, hexa, hepta, and octachlorodibenzo- para-dioxin isomers in particulate samples at parts per trillion levels. Anal. Chem. 52, 2045–2054.

    Lee, M. L., Vassilaros, D. L., and White, C. M. (1979). Retention indices for programmed-temperature capillary-column gas chromatography of polycyclic aromatic hydrocarbons. Anal. Chem. 51, 768–773.

    Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951). Protein measurement with Folin phenol reagent. J. Biol. Chem. 193, 265–275.

    stman, C. E., and Colmsj, A. L. (1987). Backflush HPLC for the isolation of polycyclic aromatic compounds. A comparative study. Chromatographia 23, 903–908.

    Pettersen, H., Axelman, J., and Broman, D. (1999). The relative contribution of spatial-, sampling- and analytical variation to the PAH and PCB concentrations in Baltic Sea sediments. Chemosphere 38, 1025–1034.

    Prough, R. A., Burke, M. D., and Mayer, R. T. (1978). Direct fluorometric methods for measuring mixed-function oxidase activity. In Methods in Enzymology (S. Fleicher and L. Packer, Eds.), vol. L II, part C, pp. 372–377. Academic Press, New York.

    Rice, S. D., Short, J. W., and Karinen, J. F. (1977). Comparative oil toxicity and comparative animal sensitivity. In Fate and Effects of Petroleum Hydrocarbons in Marine Organisms and Ecosystems (D. A. Wolfe, Ed.), pp. 78–94. Pergamon, Elmsford, NY.

    Safe, S. (1990). Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and related compounds: environmental and mechanistic considerations which support the development of toxic equivalency factors (TEFs). Crit. Rev. Toxicol. 21, 51–88.

    Schuetzle, D., and Lewtas, J. (1986) Bioassay-directed chemical analysis in environmental research. Anal. Chem. 58, 1060A–1075A.

    Spitsbergen, J. M., Walker, M. K., Olson, J. R., and Peterson, R. E. (1991). Pathologic alterations in early life stages of lake trout, Salvelinus namaycush, exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin as fertilized eggs. Aquat. Toxicol. 19, 41–71.

    Sundberg, H., Tjrnlund, U., kerman, G., Grunder, K., Ishaq, R., Hammar, T., Broman, D., and Balk, L. (in press). The distribution and relative toxic potential of organic chemicals in a PCB contaminated bay. Mar. Pollut. Bull.

    Villalobos, S. A., Papoulias, D. M., Meadows, J., Blankenship, A. L., Pastva, S. D., Kannan, K., Hinton, D. E., Tillitt, D. E., and Giesy, J. P. (2000). Toxic responses of medaka, d-rR strain, to polychlorinated naphthalene mixtures after embryonic exposure by in ovo nanoinjection: a partial life-cycle assessment. Environ. Toxicol. Chem. 19, 432–440.

    Walker, M. K., and Peterson, R. E. (1991). Potencies of polychlorinated dibenzo-p-dioxin, dibenzofuran, and biphenyl congeners, relative to 2,3,7,8-tetrachlorodibenzo-p-dioxin, for producing early life stage mortality in rainbow trout (Oncorhynchus mykiss). Aquat. Toxicol. 21, 219–238.

    Walker, M. K., Zabel, E. W., kerman, G., Balk, L., Wright, P., and Tillitt, D. E. (1996). Chapter Four: Fish egg injection as an alternative exposure route for early life stage toxicity studies. Description of two unique methods. In Techniques in Aquatic Toxicology (G. K. Ostrander, Ed.), pp. 41–72. CRC Press, Boca Raton, FL.

    Wassenberg, D. M., Swails, E. E., and Di Giulio, R. T. (2002). Effects of single and combined exposures to benzo(a)pyrene and 3,3'4,4'5-pentachlorobiphenyl on EROD activity and development in Fundulus heteroclitus. Mar. Environ. Res. 54, 279–283.

    Whyte, J. J., Jung, R. E., Schmitt, C. J., and Tillitt, D. E. (2000). Ethoxyresorufin-O-deethylase (EROD) activity in fish as a biomarker of chemical exposure. Crit. Rev. Toxicol. 30, 347–570.

    Wolf, K. (1969). Blue-sac disease of fish. Fish Disease Leaflet 15, U.S. Fish and Wildlife Service, Kearneysville, WV.

    Zebühr, Y., Nf, C., Bandh, C., Broman, D., Ishaq, R., and Pettersen, H. (1993). An automated HPLC separation method with 2 coupled columns for the analysis of PCDD/Fs, PCBs and PACs. Chemosphere 27, 1211–1219.(Henrik Sundberg, Rasha Is)