Alkoxyresorufin-O-deethylase Activities and Polychlorinated Biphenyl

Aug 18, 2005 - Nico W. van den Brink , Dennis R. Lammertsma , Wim J. Dimmers , and Marie Claire Boerwinkel. Environmental Science & Technology 2011 ...
0 downloads 0 Views 136KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 7337-7343

Alkoxyresorufin-O-deethylase Activities and Polychlorinated Biphenyl Patterns in Shrews as Biomarkers in Environmental Risk Assessments: Sensitivity and Specificity NICO W. VAN DEN BRINK* AND ALBERTUS (BART) T. C. BOSVELD† Alterra, Wageningen UR, Box 47, NL6700AA, Wageningen, The Netherlands

Alkoxyresorufin-O-deethylase (AROD) biomarkers are useful indicators of the exposure of organisms to dioxinlike compounds. In the current study, an in vivo validation of the use of such biomarkers in shrews was conducted. Furthermore, the use of changes in polychlorinated biphenyl (PCB) patterns as an animal-friendly alternative to AROD biomarkers was evaluated. Two experiments and a field study were conducted in which dose-response relations were established between levels of Σ-PCBs in shrews on one hand and their AROD activities and changes in PCB patterns on the other. We demonstrate that the changes in PCB patterns are as sensitive as the classic AROD biomarkers. The experiments also showed a substrate-specific induction of AROD biomarkers and a related PCB congenerspecific metabolism. This implies that congener-specific analysis of PCBs can reveal activities of specific AROD biomarkers. Gender-specific induction of AROD activities in shrews was shown in the field study, whereas the relationship between exposure and changes in PCB patterns did not differ between genders. It is concluded that (i) AROD biomarkers are useful biomarkers to assess exposure of shrews to specific organochlorines and that (ii) changes in PCB patterns can be used as an animalfriendly alternative to these AROD biomarkers.

Introduction The current widespread contamination of soil, water, and air by human activities may adversely affect biota and even humans. For a sound management of the environment, it is essential to know the potential and actual risks that such contaminants may pose to ecological endpoints (1). When the risks of soil contamination to wildlife in temperate climates are addressed, food chains based upon earthworms are considered to be important routes of accumulation (2). In The Netherlands, for instance, badgers (Meles meles) and little owls (Athene noctua) rely on earthworms as a major prey item in their diet (3, 4), and these species are expected to be at risk at contaminated sites. However, for ethical and practical reasons, the inclusion of such endangered species in monitoring programs to assess environmental risks is * Corresponding author phone: +31-317-477872; fax: +31-317424988; e-mail: [email protected]. † Current address: SETAC Europe, Av. de la Toison d’Or 67, B-1060 Brussels, Belgium. 10.1021/es0504688 CCC: $30.25 Published on Web 08/18/2005

 2005 American Chemical Society

problematic. Therefore, shrews, which feed on earthworms (5, 6), have been studied as alternative indicators of endangered predators on earthworms (7-9). In general, ecotoxicological studies on shrews have focused on the assessment of body burdens of contaminants, while only a few have focused on biomarkers (8, 10). Biomarkers have been thoroughly validated in environmental risk assessments (ERAs) in aquatic ecosystems, are generally applied in fish, and have proven to provide essential information on the actual risks of contaminants (11). Application of such biomarkers in shrews would generate similar essential information on actual risks of exposure (12). A recent study addressed the induction of alkoxyresorufinO-deethylase (AROD) biomarkers in an American shrew species (10). In that study, it was confirmed that AROD activities can be induced in shrew species, but little information was delivered on effect levels, i.e., the sensitivity of the species to exposure to contaminants. This sensitivity is of great importance in an ERA, and the current study was designed to validate the applicability of AROD biomarkers in European shrews as indicators of exposure to dioxin-like compounds, focused on both sensitivity and specificity. In addition to the more classical AROD biomarker approach, we have previously proposed an alternative procedure based upon a detailed analysis of polychlorinated biphenyl (PCB) patterns in organisms (13). The use of changes in PCB patterns as an alternative to AROD biomarkers is based upon the hypothesis that specific cytochrome P450 (CYP) iso-enzymes (associated with specific AROD activities) metabolize specific PCB congeners, resulting in altered PCB patterns in samples of affected organisms (14-16). The use of such PCB pattern changes would have several advantages over the application of AROD biomarkers. First, a single PCB analysis would provide information on both exposure, i.e., PCB levels in tissues, and effects, i.e., changes in PCB patterns. Second, PCB patterns can be analyzed in nondestructively obtained material (3), in contrast to AROD biomarkers, which are generally analyzed in liver material. Third, PCB patterns can be analyzed in material stored frozen, which allows for retrospective studies (13). Although correlative studies have related changes in PCB patterns to exposure to contaminants (3, 14, 17, 18), this has not been validated experimentally. Therefore, before changes in PCB patterns can be applied as biomarkers in ERAs, they have to be validated in vivo, and causal dose-effect relationships between exposure and changes in PCB patterns have to be established. As far as the authors are aware, no such experimental in vivo validation of the relation between AROD biomarker responses and PCB metabolism in wildlife has been published. The design of the current study allowed for such a validation, so the analyses were used to meet this objective as well. To address the two main objectives of the study, i.e., (i) in vivo validation of the use of AROD biomarkers in shrews and (ii) in vivo validation of the use of changes in PCB patterns as alternatives to the AROD biomarkers, two experimental exposure studies and one field study were conducted. The two laboratory exposure studies focused on (i) sensitivity to establish experimental dose-response relations between exposure to PCBs on one hand and AROD activities and changes in PCB patterns on the other and (ii) specificity to assess the compound-specific induction of AROD biomarkers and PCB metabolism in shrews. A field study was designed to validate the experimentally derived dose-effect relationships with actual field data collected from environmentally exposed specimens. VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7337

Methods and Materials Laboratory Exposure Study 1: Sensitivity. This study assessed dose-response relationships between PCB exposure and its effects on hepatic AROD activities and tissue congener patterns (relative depletion of metabolizable congeners) and established effect levels. One male and two female white-toothed shrews (Crocidura russula) were captured in April and May 2000 in a public garden in the town of Bennekom, The Netherlands, using Longworth live traps. They were kept separately in plastic boxes (60 × 40 × 30 cm3) on wood pulp bedding, were daily fed freshwater and cat food (Felix cat food) ad libitum, and had free access to a treadmill. The daylight regime was set at 10 h of light and 14 h of darkness. After 5 weeks, the male shrew was placed with each female in turn and was allowed to mate. This resulted in a litter for each female, after which the young were used for further experiments. After a month, the young animals were placed in separate plastic boxes, similar to those used for the adults. After another 2 months, 9 subadult animals (5 ?, 4 /) were placed under experimental conditions from 16 September 2000 to 8 January 2001. Our use of the F1 generation was not intended to study genetically determined responses to PCB exposure but was merely adopted to minimize the need to trap free ranging animals. It also minimized the possible biological variation in metabolic capacity due to genetic and past environmental differences, which might increase the power of the experiment. The animals were exposed to PCBs through the diet. PCBs were dissolved in corn oil and mixed in with the food. The nominal concentrations of total PCBs in the food were (with the gender of the shrew in brackets): 10-5(/), 10-4(/), 10-3(?), 10-2(/), 0.1(?), 0.3(/), 1(?), 3(?), and 10(?) µg/g wet weight. The animals were all feeding approximately 10 g per day, although at higher concentrations a slight decrease in daily food intake was detected during the first two weeks. Total daily dietary PCB uptake ranged from 0.0001 to 100 µg/day. After the experiment had been terminated, the animals were sacrificed by decapitation. Samples of the liver were collected and frozen in liquid nitrogen, after which the samples were stored at -80 °C prior to further AROD analysis. Subcutaneous fat was collected and stored at -20 °C prior to PCB analysis. Laboratory Exposure Study 2: Specificity. This study addressed inducer-specific induction of AROD biomarkers in combination with the congener-specific metabolism of a PCB mixture. Common shrews (Sorex araneus) were captured in Longworth live traps from 18 to 27 June 2001 at the Sinderhoeve, a noncontaminated site near Renkum, The Netherlands. A total of 20 shrews (10 ? and 10 /) were used in the experiment. All animals were subadults and were kept and fed in a similar condition as the animals in experiment 1. Prior to the experiment, the animals were acclimatized for a minimum of 3 weeks. It was decided to use free ranging animals instead of animals from an F1 generation, because it was not feasible to catch adult shrews in June, so no young could be bred. Due to restrictions following the occurrence of foot-andmouth disease in The Netherlands in early 2001, it was not possible to go out into the field and catch animals earlier in the season. The animals were exposed to relatively low concentrations of PCBs in combination with phenobarbital (PB) and/or β-naphthoflavone (NF) via the diet. PCBs, NF, and PB were dissolved in corn oil and mixed in with the cat food. The total PCB concentration was approximately 100 pg/g wet weight. Experiment 1 had shown that AROD activities are not induced at this exposure level. Nominal concentrations of NF and PB in the food were set at 50 µg/g wet weight. The animals were all feeding approximately 10 g of cat food per day, although 7338

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005

this varied in time. No animals were known to feed significantly less than others. Twenty shrews were distributed over four treatments groups (n ) 5 per group). One group received food that was spiked with PCBs only (2/, 3?). The other groups received food spiked with PCBs and NF (3/, 2?), PCBs and PB (2/, 3?), or a combination of PCBs, NF, and PB (3/, 2?). Hence, all animals received PCBs at low levels, but some were cotreated with NF, PB, or a combination of these compounds, which allowed for analysis of variance (ANOVA) of the data. Total exposure time was 30 days. Field Study. The field study examined the effects of environmental exposures on AROD activity and PCB congener patterns. In the field study, 26 common shrews (Sorex araneus, 17 /, 9 ?) and earthworms were collected at the Volgermeer area, The Netherlands. The Volgermeer is a former refuse dumping site with known contamination by polyhalogenated aromatic hydrocarbons (PHAHs), approximately 20 km north of Amsterdam, where chemical waste was dumped between approximately 1950 and 1985. Shrews were captured between 12 and 17 July 2001, using Longworth live traps. Earthworms were collected by hand in a 20 by 20 cm2 square in which the top 10-cm soil layer was checked. All worms were “starved” for 2 days to eliminate ingested soil and frozen prior to analysis of heavy metals and organochlorines. PCBs were analyzed in worms and individual shrews in homogenates of their carcasses (except liver and brain). Biochemical and Statistical Analyses. PCB Analyses. Approximately 200-600 mg sample was Soxhlett extracted with hexane (6 h). The extract was cleaned over aluminum oxide, eluted with hexane, to eliminate lipids. After the cleanup, individual PCB congeners were analyzed by gas chromatography with mass selective detection (GC-MSD) in single-ion mode. Of each congener, two specific masses were monitored during a run: trichlorinated congeners m/z 256/ 258, tetrachlorinated m/z 290/292, pentachlorinated m/z 326/ 328, hexachlorinated m/z 366/368, heptachlorinated m/z 394/ 396, octachlorinated m/z 428/430, nonachlorinated m/z 462/ 464, and decachlorinated m/z 496/498. Standards of individual PCB congeners were used as references. Lipids were detected gravimetrically. Concentrations were normalized on lipid content. Detection limits were approximately 0.5-1 ng/g lipid wt, depending on the congener. The following PCB congeners were analyzed and are reported: 28/31, 49, 52, 61, 66, 95, 101, 105, 118, 138, 149, 153, 170, 180, 183, and 187. According to their substitution patterns, these congeners are classified as follows: OM congeners with two adjacent nonsubstituted carbons at an ortho-meta site (congeners 28/31, 49, 61, 66, 105, and 118) and PM congeners with two adjacent nonsubstituted carbons at a meta-para site (congeners 28/31, 49, 52, 61, 95, and 101). Some congeners have both OM and PM characteristics and are thus included in both groups. The metabolic fraction (MFPCB) was defined for all metabolizable PCB congeners, without further classification (all congeners in either the OM or the PM group). AROD Analyses. AROD activity was assessed as the specific transformation of alkoxyresorufin (A-RR) into resorufin (RR) by P450 iso-enzymes. Ethoxyresorufin-O-deethylase (EROD), methoxyresorufin-O-demethylase (MROD), pentoxyresorufin-O-depenthylase (PROD), and benzyloxyresorufin-O-debenzylase (BROD) were measured in the hepatic microsomal fraction. The microsomal fraction was isolated from liver material, which was kept at -80 °C. The livers were homogenized in phosphate buffer and centrifuged, and the pellets, containing the microsomes, were suspended in phosphate buffer with 20% (v/v) glycerol. AROD activities were determined by incubating microsomal protein in the dark at 37 °C. The NADPH-dependent formation of resorufin was measured in a multiwell plate

FIGURE 1. (A) Relationship between ∑-PCB and EROD activity in white-toothed shrews. In shrews fed the three lowest doses, some less chlorinated PCB congeners occurred in concentrations near or below detection limits. Since it is detectable concentrations that matter when analyzing PCB patterns, these samples were omitted from the analyses of the relationships between AROD activities and specific PCB congeners. Samples marked b were not included in analyses of PCB patterns, whereas samples marked + were included in these analyses. (B) Relationship between EROD activity (natural-logarithm-transformed) and MFPCB in white-toothed shrews (n ) 6; experiment 1). For details on relationships, see Table 1.

TABLE 1. Experiment 1: Output of Logistic Regression Analyses of the Relationships between ∑-PCB AROD Activities and Linear Regressions between MFPCB and AROD Activities (Natural-Logarithm-Transformed) in White-Toothed Shrews (n ) 9, Experiment 1)a Σ-PCB

Σ-PCB

Σ-PCB

MFPCB

variable

significance of regression

EC50 (µg/g) (95% confidence interval)

maximum response (pmol/(min/mg))

significance of regression

EROD MROD PROD BROD

*** *** *** ***

29.1 (26.9-31.6) 24.3 (21.1-28.0) 36.3 (32.5-40.5) 36.7 (30.9-43.5)

2422 818 40 38

** ** * *

a Logistic model used in regression for Σ-PCB: AROD ) a - c/(1 + exp(-b(ln([PCB])-m)), with a ) lower asymptote, c ) upper asymptote, b ) maximum slope of curve, and m ) ln(EC50). Maximum response: c - a. Significance of regressions: ***, p < 0.001; **, 0.001 < p < 0.01; *, 0.01 < p < 0.05.

reader at excitation/emission of 530/590 nm and was linearly time- and protein-dependent. AROD activities were normalized by microsomal protein. The protein contents of the microsome suspensions were determined using fluorescamine as the detecting agent (19). Bovine serum albumin (BSA) was used as a standard. Statistical Analyses. Linear regression analyses were performed with the Genstat 5.3 program, using the leastsum-of-squares method and logistic regressions by maximum likelihood (20). Prior to regression analyses, Σ-PCB concentrations were transformed into their natural logarithm to approximate a normal distribution and homoscedasticity of the distribution. The relative occurrences of the different PCB classes were normalized to Σ-PCB and not transformed prior to statistical analysis. AROD activities were only naturallogarithm-transformed if they were used as independent variables in a statistical analysis. Analyses of variance (ANOVAs) were used to assess the effect of PB and NF on the induction of the different AROD activities (21).

Results Experiment 1: Sensitivity. Dose-Response Relations between Σ-PCB and AROD Activities. The experiment detected significant dose-response relations between Σ-PCB levels and AROD activities (Figure 1a, Table 1). The AROD activities were highly correlated with each other. Pairwise comparison of the different responses showed correlation coefficients all above 0.99. The sensitivities of the different AROD responses were similar, indicated by similar EC50 values of approximately 24-37 µ/g PCBs (Table 1). The maximum activities, however, differed between the AROD responses: EROD had the highest

maximum activity, while MROD had an intermediate response and PROD and BROD showed relatively small responses (Table 1). Dose-Response Relations between AROD Activities and MFPCB. Because the AROD responses were highly correlated in experiment 1, it was not possible to relate responses of specific AROD activities to the relative concentrations of specific classes of PCB congeners (i.e., OM or PM congeners). Hence, only comparisons with the sum of all metabolizable congeners (i.e., MFPCB) were made in this experiment. Relations between AROD responses (natural-logarithmtransformed) and the relative concentrations of MFPCB were significant (Figure 1b, Table 1). The relations were significant and negative (Table 1), indicating lower levels of MFPCB after induction of AROD activities. Experiment 2: Specificity. Specific AROD Induction. Specific AROD induction was related to exposure to a specific inductor (Figure 2, Table 2). EROD, MROD, and PROD activities were induced by NF, but EROD and MROD activities were inhibited to some extent in PB-exposed animals (Table 2). The significant interaction between NF and PB exposures in the case of EROD and MROD implies that EROD and MROD are induced by NF in the animals exposed to both NF and PB, but this induction is less than that in the animals exposed to NF but not to PB (Table 2). BROD activity was significantly induced by exposure to PB but not by NF (Table 2). These different responses indicate that EROD, MROD, and PROD were correlated with each other (significance of linear regressions between natural-logarithm-transformed activities p < 0.001), while the response of BROD was not related to any of the other AROD activities (significance of linear VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7339

FIGURE 2. Relationships (A) between relative concentrations of OM congeners and EROD and (B) between PM congeners and BROD. Significance of relationships between OM congeners and EROD, MROD, and PROD: p < 0.001, with BROD p > 0.05. Significance of relationships between PM congeners and BROD: p < 0.001, with EROD, MROD, and PROD p > 0.05 (n ) 20, experiment 2).

TABLE 2. Experiment 2: ANOVA Analysis of Σ-PCB Concentrations and Induction of Specific AROD Responses in Common Shrews Treated with Phenobarbital (PB) and/or β-Naphthoflavone (NF) (n ) 5)a NF PB NF/PB

PCB

EROD

MROD

PROD

BROD

n.s. n.s. n.s.

*** (+) *** (-) **

*** (+) *** (-) **

*** (+) n.s. n.s.

n.s. *** (+) n.s.

a Significance: ***, p < 0.001; **, 0.001 < p < 0.01; n.s., p > 0.05. Induction of specific activity (+); inhibition of specific activity (-).

TABLE 3. Field Study: Geometric Means of the Concentrations of PCBs (µg/g Lipid Wt) in Fat from Male and Female Common Shrews and Hepatic Activities of the Various AROD Biomarkersa all shrews (n ) 26) / shrews (n ) 17) ? shrews (n ) 9) male vs female

Σ-PCB

EROD

MROD

PROD

BROD

2.9 2.6 3.4 p > 0.1

47 34 83 p < 0.1

11 7 26 p < 0.1

2.4 1.9 3.8 p < 0.1

4.6 3.6 7.0 p > 0.1

a Male vs female significance of differences between genders assessed by t-test, natural-logarithm transformation prior to statistical test.

regressions between natural-logarithm-transformed activities p > 0.05). Relationships between AROD and Specific PCB Congeners. Relative concentrations of the different types of PCB congeners were negatively related to (natural-logarithmtransformed) AROD activities (Figure 2). OM congeners were significantly negatively related to EROD, MROD, and PROD activities (Figure 2a). The relative concentrations of PM congeners were negatively related to BROD activity (Figure 2b). Field Experiment. PCBs and AROD Biomarkers. Table 3 lists the PCB concentrations and the different AROD biomarker activities (geometric means). The PCB concentrations were highly variable, ranging from 0.25 to 13.2 µg/g, and the same was true for the AROD biomarkers (EROD, 6-294 pmol/ (min/mg); MROD, 0.1-147 pmol/(min/mg); PROD, 1-20 pmol/(min/mg); BROD, 0.1-23 pmol/(min/mg)). The PCB concentrations were not significantly different between genders, and the same was found for BROD activities. The 7340

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005

FIGURE 3. Relationship between EROD activity and Σ-PCB concentrations in male and female shrews. See Table 5 for details on regressions.

TABLE 4. Field Study: Significance of Linear Relations between Σ-PCB Concentrations and MFPCB in Male and Female Shrews and Different AROD Responsesa variable

Σ-PCB

Σ-PCB

MFPCB

MFPCB

gender

? shrews (n ) 9) ** ** * *

/ shrews (n)17) n.s. n.s. n.s. n.s.

? shrews (n ) 9) ** ** * * ***

/ shrews (n ) 17) ns ns ns. ns **

EROD MROD PROD BROD Σ-PCB

aAll data natural-logarithm-transformed prior to analysis. Significance: ***, p < 0.001; **, 0.001 < p < 0.01; *, 0.01 < p < 0.05; n.s., p > 0.05.

EROD, MROD, and PROD activities were slightly higher in female shrews than those in males, although these differences were only significant at the 0.1 level. The activities of the different AROD biomarkers were all highly correlated with each other (significance of linear regressions between natural-logarithm-transformed activities p < 0.001). Only the activities of the AROD biomarkers in the female shrews were significantly related to the PCB concentrations (Figure 3, Table 4). Figure 3 shows the relation between EROD activity and PCB concentrations by way of example. It is evident from this figure that the relationship between PCB concentrations and EROD activity was significant for female shrews but not for male shrews. This is similar for the other AROD activities (Table 4). Metabolizable PCB Congeners. Significant relationships between MFPCB and AROD biomarkers were found only in female shrews (Table 4). Figure 4a illustrates the relationship between EROD activity and MFPCB, and it is evident that only the female shrews show a significant relationship. When Σ-PCB was applied as an explanatory variable, the relationship between the MFPCB and Σ-PCB became significant for both male and female shrews, and this relationship did not differ significantly between the genders (Table 4, Figure 4b).

Discussion Sensitivity. Experiment 1, addressing the dose-response relationships between exposure to PCBs, EROD induction, and changes in MFPCB found significant relationships between PCB and AROD and between AROD and MFPCB (Figure 1 and Table 1). These results support the hypotheses that (i) AROD

FIGURE 4. Relations (A) between EROD activity and MFPCB in male and female shrews and (B) between Σ-PCB and MFPCB. Linear regressions performed on natural-logarithm-transformed data. For statistical output, see Table 5. biomarkers in shrews are related to exposure to PCBs in a dose-responsive manner, which was also shown for another shrew species (10), and that (ii) PCB patterns change with AROD activities and PCB exposure in a dose-dependent manner. However, a major aspect yet to be resolved is the relative sensitivity of the different methods. The EC50 of the relationship between EROD and Σ-PCB was approximately 29 µg/g lipid wt, and we found a similar value for MROD, while the EC50 values for PROD and BROD were slightly higher (Table 1). The EC50 of the relationship between the relative concentrations of MFPCB and Σ-PCB was approximately 16 µg/g lipid wt. This suggests that changes in PCB patterns and induction of AROD activities in shrews occur at similar exposure levels, which implies that the application of pattern analysis in shrews is as sensitive as the use of AROD biomarkers. For the short-tailed shrew (Blarina brevicauda), significant induction of EROD, MROD, and BROD was detected at PCB levels above 20-25 µg/g lipid wt (10), which suggests similar sensitivity with the shrew species of the current study. In mink (Mustela vison), significant induction of EROD has been found between 1 and 10 µg/g lipid wt (Clophen A50), while PROD was significantly induced between 10 and 50 µg/g lipid wt (22). This suggests that the white-toothed shrew is slightly less sensitive than mink with regards to EROD induction and similar with regards to PROD induction. However, the mink results were based upon longterm exposure (18 months), in which the lowest experimental dose of a PCB mixture was approximately 100 µg/day, i.e., the highest dose of the current study. Hence, it appears that the shrews in the current study showed comparable sensitivities relative to the exposure dose in the food, for both EROD and PROD. Findings of another experimental feeding study of mink EROD activities allow an EC50 of approximately 60 µg/g w/w in fat tissue to be deduced (23). Assuming 80% lipids in fat, this would imply an EC50 of 75 µg/g lipid wt, which is slightly higher than the EC50 of EROD in the current study. It appears that, as judged by AROD activities, shrews show a similar sensitivity to PCB exposure as mink. Furthermore, as far as we are aware, this was the first experimental in vivo validation to demonstrate that PCB metabolism in shrews has a dose-response relationship with the activity of CYP enzymes, as measured by AROD biomarkers. Specificity. Induction of Specific AROD Activities. The common shrews in experiment 2 showed clear compoundspecific induction of various AROD biomarkers (Table 2). Exposure to NF resulted in induction of EROD and MROD, which is indicative of CYP1A activity, but also in elevated PROD activity, which is generally indicative of CYP2B1 (Table 2). In studies using rats (Rattus norvegicus), NF induced

CYP1A activity, while exposure to PB resulted in induction of CYP2B (24). For the rat, it has been established that induction of CYP1A1 is related to EROD, CYP1A2 to MROD, and CYP2B to PROD and BROD, although for the last two of these this was not exclusive (24, 25). However, species-specific deviations of the substrate affinity of CYPs have been reported for other species (26). In the short-tailed shrew, exposure to NF (intraperitoneal) resulted in induction of EROD, MROD, and BROD, but exposure to PB (intraperitoneal) did not result in any AROD induction (10). In cotton rats (Sigmodon hispidus), PB exposure led to induction of CYP2B and CYP3A (27). Similar observations have been made in other species of the Cricetidae family (26). Since the shrew species in the current study belong to the Soricidae family (Insectivora), taxonomic differences may be responsible for the differences in substrate affinities of the different forms of CYPs. Nevertheless, it is clear that substrate-specific induction of P450 iso-enzymes occurs in shrews and that information on induction patterns can be used to obtain information on the type of contaminants that specimens may be exposed to. Structure-Specific Metabolism of PCB Congeners. It has been reported, based on quantitative structure-activity relationships (QSARs), that the metabolism of PCB congeners with para-meta vicinal hydrogen atoms is associated with the induction of P450 iso-enzymes by phenobarbital (PB induction of CYP2B), while the metabolism of congeners with ortho-meta vicinal hydrogen atoms is related to P450 isoenzymes induced by 3-methylcholanthrene (3-MC induction of CYP1A) (28). It is especially the chlorination of the ortho site that is of importance (29). In rats, PB-induced in vitro metabolism of brominated biphenyls included congeners with vicinal hydrogen atoms at meta-para sites, while the metabolism induced by 3-MC was restricted to congeners with vicinal hydrogen atoms at ortho-meta sites (30). NF has a similar induction pattern as 3-MC, and it was expected that the relative concentrations of OM congeners would be closely related to AROD activities induced by NF, while the relative concentrations of PM congeners would mostly be related to PB-induced activities. The metabolic patterns found in the current study are in agreement with these expectations (Figure 2). The relative concentrations of all OM congeners were negatively related to the EROD, MROD, and PROD activities, all induced by NF, while the relative concentrations of the PM congeners showed a negative relationship with BROD, which was PB-induced. This experiment illustrates that in shrews different types of CYP enzyme are involved in the in vivo metabolism of specific PCB congeners. PCB congeners with vicinal hydrogen atoms at ortho-meta sites (OM congeners) are metabolized following induction by NF, while congeners with vicinal hydrogen atoms at meta-para sites are metabolized as a result of induction by PB. This implies that a specific analysis of PCB patterns focusing on OM and PM congeners provides detailed information on the induction patterns of CYP iso-enzymes, similar to the AROD biomarkers. Field Validation. Σ-PCB levels in the current study were in the range of 0.3-13.2 µg/g lipid wt. An earlier study found PCB concentrations ranging from 4.6 to 16 µg/g lipid wt in shrews in floodplains of the river Waal in The Netherlands (31). In a reference area in the eastern part of The Netherlands, PCB concentrations of 0.2-0.8 µg/g lipid wt were detected in shrew liver, while levels of 0.8-6.2 µg/g lipid wt were found in specimens from floodplains (3). The levels found in the shrews from the Volgermeer site in the current study (Table 3) were in the same range as those in the shrews from floodplains. Since floodplains in The Netherlands are considered to be moderately or heavily polluted, the levels of Σ-PCB found in the current study are indicative of a more or less polluted situation, as expected from the history of the Volgermeer site. The AROD responses in the field study (Table VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7341

TABLE 5. Field Study: Significance and Slopes of Linear Relations between Σ-PCB Concentrations and MFPCB in Shrewsa

ln(Σ-PCB) vs ln(MFPCB) ln(Σ-PCB) vs ln(PCB118) ln(Σ-PCB) vs ln(MFPCB) minus PCB28/31 and PCB61 aAll

significance

slope

standard error of slope

lower 95%

upper 95%

*** ** ***

-0.78 -0.56 -0.69

0.13 0.18 0.14

-1.05 -0.92 -0.97

-0.51 -0.20 -0.41

data natural-logarithm-transformed prior to analysis. Significance: ***, p < 0.001; **, 0.001 < p < 0.01.

3) were much lower than the levels found in experiment 1 (Table 1). It has been reported that nonexposed captive shrews showed elevated levels of AROD activities when compared to those of wild-caught animals (10). Furthermore, differences in induction between animals fed environmentally contaminated food and those fed spiked food have been found for mink. A study of mink fed on environmentally contaminated fish found average induced EROD activities of 260 pmol/(mg/min) (23), while mink fed on spiked food showed average induced EROD activities of 3000 pmol/(mg/ min) (22). Relationship between AROD Biomarkers and MFPCB in the Field Study. The activities of the various AROD biomarkers in the shrews exposed in the field were strongly correlated. This hampered an analysis of AROD-specific metabolism of PCB congeners, based upon structural characteristics of the different PCB congeners, as validated in experiment 2. Hence, this section of the discussion focuses on the relation between EROD and MFPCB, but it should be noted that the other AROD biomarkers show similar reactions. The relation between Σ-PCB and EROD in the shrews exposed in the field was just significant for females but not for males (Figure 3). Gender-specific induction of cytochrome P450 iso-enzymes has been described before, for instance, in rats (32, 33). However, Figure 3 illustrates that induction of EROD occurred in some specimens of both male and female shrews. It appears that other factors than PCBs influence the induction of EROD in male shrews or that PCB levels in male shrews show large short-term variations. Such factors may be obscuring the relation between Σ-PCB and EROD activity. Furthermore, it is also possible that EROD induction is inhibited in certain individuals, as has been shown in frogs from the same location at the Volgermeer site (34). MFPCB showed a significant relationship with Σ-PCB, which was independent of gender (Figure 4b). This would suggest that variations in Σ-PCB levels are not the main cause of the lack of significance of the relationship between Σ-PCB and EROD in male shrews. However, the significant relationship between Σ-PCB and MFPCB should be interpreted with caution. This relationship may be an artifact of the methods, since in some samples the concentrations of some metabolizable PCB congeners were lower than the detection limits. In such cases, a concentration of 50% of the detection limit was assigned to the specific congener. If this were the case for the majority of the congeners, then the relative concentration of MFPCB would become completely dependent on Σ-PCB and would show a similar relationship as that shown in Figure 4b. To analyze this, we looked at the relation between the relative concentrations of a single metabolizable PCB congener that was found above the detection limit in all samples, PCB118, and Σ-PCB. Table 5 shows the output for both MFPCB and PCB118. Both the relations between Σ-PCB and MFPCB and the relative concentrations of PCB118 were highly significant. The slopes of the two regression lines were not significantly different. (The 95% interval of each slope includes the average of the other.) If all metabolizable congeners had occurred below the detection limits, then the slope of the regression between ln(Σ-PCB) and ln(MFPCB) would be -1. The slope of MFPCB is closer to -1 than the 7342

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 18, 2005

slope of PCB118, indicating that the relationship between MFPCB and Σ-PCB may to some extent be affected by some congeners with concentrations below detection limits. The 95% interval of the slope ranges from -1.05 to -0.51 and thus includes -1. This would suggest that there may be an impact of congeners at the detection limits. If congeners 28/31 (which were below detection limits in 25% of the samples) and congener 61 (which was below detection limits in 95% of the samples) are excluded from the MFPCB, then the slope of the regression between Σ-PCB and MFPCB becomes -0.69 (Table 5), with a 95% confidence interval ranging from -0.97 to -0.41, i.e., not including -1. This implies that the effect of congeners occurring below detection limits is eliminated. This illustrates that, when using relative concentrations in MFPCB, one should always be aware of PCB congeners that occur at levels below detection limits. The field study described here illustrates that the MFPCB signal appears to be a better indicator of long-term exposure than EROD induction. MFPCB integrates the AROD activity over time, so it does not show short-term fluctuations in AROD activities, the likely cause of the lack of significance in the relation between Σ-PCB and AROD in male shrews. This may be an advantage in studies focused on long-term exposure but perhaps less appropriate in studies examining short-term fluctuations. The above argument allows us to conclude that MFPCB can be used as a signal for AROD activity in shrews. Similar to the AROD biomarker, MFPCB may be used as an early warning signal, in a qualitative way. Since the dose-response curve in Figure 1 is steep, it appears less appropriate to use either AROD biomarkers or MFPCB in a strict quantitative way in an ERA. Nevertheless, congener-specific analysis of PCBs is possible in samples that can be obtained nondestructively, for instance, blood and uropygial oil (3, 18, 35). It is therefore concluded that MFPCB analysis can be used as a nondestructive alternative to AROD biomarkers in field studies.

Acknowledgments Paul de Bie, Dennis Lammertsma, and Jos Bodt skillfully conducted the field work. Steven Crum and Marleen Lee performed the chemical and biochemical analyses. This study was funded by a grant from The Netherlands Organization for Health Research and Development (Grant No. PAD9802) and by the Dutch Ministry of Agriculture, Nature and Food Quality (DWK-program 384). The experiments and field study were conducted in compliance with Dutch legal requirements.

Literature Cited (1) Calow, P. Environmental risk assessment and management: The whats, whys and hows? In Handbook of Environmental Risk Assessment and Management; Calow, P., Ed.; Blackwell Science: Oxford, U.K., 1998. (2) Morgan, A. J.; Morgan, J. E.; Turner, M.; Winters, C.; Yarwood, A. Metal relationships of earthworms. In Ecotoxicology of Metals in Invertebrates; Dallinger, R., Rainbow, P. S., Eds.; Lewis Publishers: Boca Raton, FL, 1993.

(3) Van den Brink, N. W.; Groen, N. M.; De Jonge, J.; Bosveld, A. T. C. Ecotoxicological suitability of floodplain habitats in The Netherlands for the little owl (Athene noctua vidalli). Environ. Pollut. 2003, 122 (1), 127-134. (4) Goszczynski, J.; Jedrzejewska, B.; Jedrzejewski, W. Diet composition of badgers (Meles meles) in a pristine forest and rural habitats of Poland compared to other European populations. J. Zool. 2000, 250 (4), 495-505. (5) Butterfield, J.; Coulson, J. C.; Wanless, S. Studies on the distribution, food, breeding biology and relative abundance of the pygme and sorex shrews (Sorex minutus and S. araneus) in upland areas of northern England. J. Zool. 1981, 195, 169-180. (6) Saarikko, J. Foraging behavior of shrews. Ann. Bot. Fenn. 1989, 26, 441-423. (7) Ma, W. C.; Denneman, W.; Faber, J. Hazardous exposure of ground-living small mammals to cadmium and lead in contaminated terrestrial ecosystems. Arch. Environ. Contam. Toxicol. 1991, 20 (2), 266-270. (8) Hamers, T.; Smit, L. A. M.; Bosveld, A. T. C.; van den Berg, J. H. J.; Koeman, J. H.; van Schooten, F. J.; Murk, A. J. Lack of a distinct gradient in biomarker responses in small mammals collected at different distances from a highway. Arch. Environ. Contam. Toxicol. 2002, 43 (3), 345-355. (9) Reinecke, A. J.; Reinecke, S. A.; Musilbono, D. E.; Chapman, A. The transfer of lead (Pb) from earthworms to shrews (Myosorex varius). Arch. Environ. Contam. Toxicol. 2000, 39 (3), 392-397. (10) Russell, J. S.; Halbrook, R. S.; Woolf, A.; French, J. B.; Melancon, M. J. Blarina brevicauda as biological monitor of polychlorinated biphenyls: Evaluation of hepatic cytochrome P450 induction. Environ. Toxicol. Chem. 2004, 23 (8), 1867-1873. (11) Van der Oost, R.; Beyer, J.; Vermeulen, N. P. E. Fish bioaccumulation and biomarkers in environmental risk assessment: A review. Environ. Toxicol. Pharmacol. 2003, 13 (2), 57-149. (12) Blankenship, A. L.; Giesy, J. P. Use of biomarkers of exposure and vertebrate tissue residues in the hazard characterization of PCBs at contaminated sitessApplication to birds and mammals. Environ. Anal. Contam. Sites 2002, 153-180. (13) Van den Brink, N. W.; De Ruiter Dijkman, E. M.; Broekhuizen, S.; Reijnders, P. J. H.; Bosveld, A. T. C. Polychlorinated biphenyls pattern analysis: Potential nondestructive biomarker in vertebrates for exposure to cytochrome P450-inducing organochlorines. Environ. Toxicol. Chem. 2000, 19 (3), 575-581. (14) Boon, J. P.; Van der Meer, J.; Allchin, C. R.; Law, R. J.; Klungsøyr, J.; Leonards, P. E. G.; Spliid, H.; Storr-Hansen, E.; Mckenzie, C.; Wells, D. E. Concentration-dependent changes of PCB patterns in fish-eating mammals: structural evidence for induction of cytochrome P450. Arch. Environ. Contam. Toxicol. 1997, 33, 298-311. (15) Shimada, T.; Sawabe, Y. Activation of 3,4,3′,4′-tetrachlorobiphenyl to protein-bound metabolites by rat liver microsomal cytochrom P448 containing monooxygenase system. Toxicol. Appl. Pharmacol. 1983, 70, 486-943. (16) Murk, A. J.; Morse, D.; Boon, J. P.; Brouwer, A. In vitro metabolism of 3,3′,4,4′-tetrachlorobiphenyl in relation to ethoxyresorufinO-deethylase activity in liver microsomes of some wildlife species and rat. Eur. J. Pharmacol., Environ. Toxicol. Pharmacol. Sect. 1994, 3 (2-3), 253-261. (17) Heidmann, W. A.; Staats de Yanes, G.; Bu ¨ the, A.; Ru ¨ ssel-Sinn, H. Correlation between concentration and composition of PCB mixtures in seals (Phoca vitulina). Chem.sEur. J. 1992, 24, 11111117. (18) Van den Brink, N. W.; Bosveld, A. T. C. PCB concentrations and metabolism patterns in common terns (Sterna hirundo) from different breeding colonies in The Netherlands. Mar. Pollut. Bull. 2001, 42 (4), 280-285. (19) Kennedy, S. W.; Lorenzen, A. A fluorescence-based protein assay for use with a microplate reader. Anal. Biochem. 1993, 214, 346-348. (20) Lane, P. W. Regression analysis. In Genstat 5, Release 3: Reference Manual; Genstat 5 Committee, Ed.; Oxford Science Publications: Oxford, U.K., 1993.

(21) Payne, R. W. Design and analysis of experiments. In Genstat 5, Release 3: Reference Manual; Genstat 5 Committee, Ed.; Oxford Science Publications: Oxford, U.K., 1993. (22) Brunstrom, B.; Lund, B. O.; Bergman, A.; Asplund, L.; Athanassiadis, I.; Athanasiadou, M.; Jensen, S.; Orberg, J. Reproductive toxicity in mink (Mustela vison) chronically exposed to environmentally relevant polychlorinated biphenyl concentrations. Environ. Toxicol. Chem. 2001, 20 (10), 2318-2327. (23) Halbrook, R. S.; Aulerich, R. J.; Bursian, S. J.; Lewis, L. Ecological risk assessment in a large river-reservoir: 8. Experimental study of the effects of polychlorinated biphenyls on reproductive success in mink. Environ. Toxicol. Chem. 1999, 18 (4), 649-654. (24) Nims Raymond, W.; Lubet Ronald, A. Induction of cytochrome P-450 in the Norway rat, Rattus norvegicus, following exposure to potential environmental contaminants. J. Toxicol. Environ. Health 1995, 46 (3), 271-292. (25) Nerurkar, P. V.; Park, S. S.; Thomas, P. E.; Nims, R. W.; Lubet, R. A. Methoxyresorufin and benzyloxyresorufin: Substrates preferentially metabolized by cytochromes P4501A2 and 2B, respectively, in the rat and mouse. Biochem. Pharmacol. 1993, 46 (5), 933-943. (26) Qualls, C. W.; Lubet, R. A.; Lochmiller, R. L.; Elangbam, C. S.; Lish, J. W.; Nims, R. W. Cytochrome P450 induction in feral Cricetid rodents: a review of field and laboratory investigations. Comp. Biochem. Physiol., Part C: Pharmacol., Toxicol. Endocrinol. 1998, 121 (1-3), 55-63. (27) Jones, C. R.; Lubet, R. A.; Henneman, J. R.; Nims, R. W. Doseresponse relationships for cytochrome P450 induction by phenobarbital in the cotton rat (Sigmodon hispidus). Comp. Biochem. Physiol., Part C: Pharmacol., Toxicol. Endocrinol. 1998, 121 (1-3), 197-203. (28) Lewis, D. F. V.; Lake, B. G.; Parke, D. V. Molecular orbitalgenerated QSARs in a homologous series of alkoxyresorufins and studies of their interactive docking with P450s. Xenobiotica 1995, 25 (12), 1355-1369. (29) Kaminsky, L. S.; Kennedy, M. W.; Adams, S. M.; Guengerich, F. P. Metabolism of dichlorobiphenyls by highly purified isozymes of rat liver cytochrome P-450. Biochemistry 1981, 20, 73797384. (30) Mills, R. A.; Millis, C. D.; Dannan, G. A.; Guengerich, F. P.; Aust, S. D. Studies on the structure-activity relationships for the metabolism of polybrominated biphenyls by rat liver microsomes. Toxicol. Appl. Pharmacol. 1985, 78 (1), 96-104. (31) Hendriks, A. J.; Ma, W. C.; Brouns, J. J.; De, R. D. E. M.; Gast, R. Modelling and monitoring organochlorine and heavy metal accumulation in soils, earthworms and shrews in rhine-delta floodplains. Arch. Environ. Contam. Toxicol. 1995, 29 (1), 115127. (32) Blaich, G.; Goettlicher, M.; Cikryt, P.; Metzler, M. Induction of P-450 isoenzyme activities in Syrian golden hamster liver compared to rat liver as probed by the rate of 7-alkoxyresorufinO-dealkylation. Chem. Biol. Interact. 1988, 67 (1-2), 129-138. (33) Mahnke, A.; Strotkamp, D.; Roos, P. H.; Hanstein, W. G.; Chabot, G. G.; Nef, P. Expression and inducibility of cytochrome P450 3A9 (CYP3A9) and other members of the CYP3A subfamily in rat liver. Arch. Biochem. Biophys. 1997, 337 (1), 62-68. (34) Van den Brink, N. W.; Lee-de Groot, M. B. E.; De Bie, P. A. F.; Bosveld, A. T. C. Enzyme markers in frogs (Rana spp.) for monitoring risk of aquatic pollution. Aquat. Ecosyst. Health Manage. 2003, 6 (4), 441-448. (35) Van den Brink, N. W.; Van Franeker, J. A.; De Ruiter-Dijkman, E. M. Fluctuating concentrations of organochlorine pollutants during a breeding season in two Antarctic seabirds: Adelie penguin and southern fulmar. Environ. Toxicol. Chem. 1998, 17 (4), 702-709.

Received for review March 9, 2005. Revised manuscript received June 6, 2005. Accepted July 18, 2005. ES0504688

VOL. 39, NO. 18, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7343