Organohalogen and metabolically-derived

Environment International 33 (2007) 823 – 830 www.elsevier.com/locate/envint

Organohalogen and metabolically-derived contaminants and associations with whole body constituents in Norwegian Arctic glaucous gulls J. Verreault a,⁎, S. Shahmiri b , G.W. Gabrielsen a , R.J. Letcher b a

b

Norwegian Polar Institute, Tromsø, NO-9296, Norway National Wildlife Research Centre, Science and Technology Branch, Environment Canada, Carleton University, Ottawa, Ontario, Canada K1A 0H3 Received 1 February 2007; accepted 26 March 2007 Available online 30 April 2007

Abstract Comprehensive surveys of organohalogen contaminants have been conducted in various tissues and blood of glaucous gulls (Larus hyperboreus), a top scavenger-predator species in Svalbard in the Norwegian Arctic. However, the physico-chemical properties of organohalogens (e.g., type and degree of halogenation and the presence or absence of additional phenyl group substituents) that may influence toxicokinetics, and subsequently tissue-specific accumulation, have yet to be studied in this species. We investigated the concentrations, total body burdens, and compositional patterns of legacy chlorinated compounds (PCBs and chlordanes (CHLs)), metabolically-derived PCBs (methylsulfonyl (MeSO2)and OH-PCBs), brominated flame retardants (polybrominated diphenyl ethers (PBDEs), polybrominated biphenyls (PBBs), totalhexabromocyclododecane (HBCD)), and PBDE metabolites and/or naturally-occurring compounds with similar structures (MeO- and OHPBDEs) in liver, blood and whole body homogenate samples of adult glaucous gulls (n = 19) from Svalbard. Further, we examined the distribution of these organohalogens and metabolites in relation to whole body composition of glaucous gulls, i.e., the total water, protein, lipid and mineral contents in whole homogenate carcasses. The total body burden of organohalogens and metabolites in glaucous gulls ranged between 3.3 and 33.0 mg. Compound class distribution showed that the relative proportions of sum (Σ) OH-PCB and ΣOH-PBDE to the total organohalogen concentrations were significantly highest in blood. Conversely, the ΣCHL and ΣPCB showed generally higher proportions in the lipid-rich liver as well as in whole body homogenates. No significant difference in the compositional patterns of individual congeners/compounds was found among tissues/blood, with the exception of the classes comprised of less polar brominated compounds (PBDEs, PBBs and total-(α)-HBCD). Total proteins isolated from the whole body homogenates of glaucous gulls were significantly associated to the proportions of ΣOH-PCB and ΣPBDE. A non-significant positive association was found between total lipids and the ΣPCB proportions. The present study suggests that both protein association and lipid solubility are important concomitant factors to be considered in the toxicokinetics and fate of contaminants as a function of chemical structure and properties, e.g., chlorination, bromination and the presence of other phenyl substituents such as OH group. An enhanced, selective retention of these organohalogen classes in given tissues/body compartments may thus lead to site-specific toxicological actions and adverse effects in the highly-contaminated Svalbard glaucous gulls. © 2007 Elsevier Ltd. All rights reserved. Keywords: Whole body constituents; Organochlorines; Brominated flame retardants; Metabolites; Lipids; Proteins; Glaucous gull; Arctic

1. Introduction Avian wildlife has frequently been used as indicator of chemical pollution exposure and toxicity for assessing the health of aquatic ecosystems. Specifically, monitoring programs using bird tissue, blood and egg samples have supplied valuable data on spatial and temporal trends of organohalogen ⁎ Corresponding author. Tel.: +47 77 75 05 00; fax: +47 77 75 05 01. E-mail address: [email protected] (J. Verreault). 0160-4120/$ - see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2007.03.013

contaminants in areas of environmental concern. Certain gull species (Laridae) such as the Great Lakes herring gull (Larus argentatus) (Hebert et al., 1999; Mineau et al., 1984), have been identified as key sentinels of freshwater basin contamination in densely populated and industrialized regions. In Norway, the glaucous gull (Larus hyperboreus) also has served as a biomonitor species in the Arctic marine environment (Barrett et al., 1996; Savinova et al., 1995). Monitoring of organohalogens in the glaucous gull has been given high priority in Svalbard (Norwegian Arctic) due to its widespread distribution

824

J. Verreault et al. / Environment International 33 (2007) 823–830

and important role as top avian scavenger-predator in the marine food web. Thus far among circumpolar Arctic seabirds, glaucous gulls from Svalbard are reported to accumulate some of the highest concentrations of organohalogen contaminants. These include the metabolically-derived or -associated polychlorinated biphenyls (PCBs) (e.g., methylsulfonyl (MeSO2)and hydroxyl (OH)-PCBs) (Verreault et al., 2005a, 2006), brominated flame retardants (e.g., polybrominated diphenyl ethers (PBDEs), polybrominated biphenyls (PBBs), total-(α)hexabromocyclododecane (total-(α)-HBCD)), PBDE metabolites and/or naturally-occurring compounds with similar structures (MeO- and OH-PBDEs) (Verreault et al., 2005b, 2006) and perfluorinated alkyl substances (Verreault et al., 2005c). Despite comprehensive surveys of legacy organochlorine residues (e.g., PCBs, DDTs and chlordanes (CHLs)) in tissues (liver, brain and subcutaneous fat) and blood/plasma of Svalbard glaucous gulls, no attention has been paid to the physico-chemical properties that can influence toxicokinetics, and by extension tissue-specific accumulation of chemicals. Depending on these properties such as functional groups and molecular structure, contaminant toxicokinetics can be altered due to, e.g., the type of halogenation (e.g., chlorine and bromine substitution), and particularly the presence or absence of additional phenyl group substituents (e.g., MeO and OH groups). Such fundamental characteristics of organohalogens are suggested to have great impact on their affinity to specific body constituents or macromolecules (Hakk and Letcher, 2003; Letcher et al., 2000). Therefore, the characteristics of a compound, including whether it is chlorinated or brominated and MeO-/OH-substituted or unsubstituted, would be expected to selectively influence its partitioning and accumulation within specific body compartments and tissues. In a previous study of glaucous gulls, liver-to-blood concentration ratios (lipid weight basis) were reported for a suite of organochlorines including PCBs, DDTs and CHLs (Henriksen et al., 1998). In this study, the liver-to-blood organochlorine ratios varied between 0.9 and 2.0. In another study of herring gulls from the Great Lakes that also examined PCB, DDT and CHL concentrations, Braune and Norstrom (1989) reported liver-to-whole body homogenate ratios ranging between 0.53 and 1.10. To our knowledge, there has not been an avian study investigating the total body burdens and patterns of chlorinated and brominated contaminants and their metabolites and/or naturally-occurring compounds with similar or analogous physico-chemical characteristics, and their associations to parameters of whole body composition (i.e., total water, protein, lipid and mineral contents). Clearly, a better knowledge on the toxicokinetics and fate of organohalogen residues within major tissues and compartments that are routinely sampled in glaucous gulls, or any other bird species, is needed for achieving consistent biomonitoring data and adequate health risk assessments. In the present study, we investigated the concentrations, total body burdens and compositional patterns of a large suite of persistent organohalogens: PCBs, CHLs, PBDEs, PBBs, total(α)-HBCD, MeSO2-PCBs/-p,p′-DDE, OH-PCBs/-PBDEs and MeO-PBDEs in liver, blood and whole body homogenate samples of adult glaucous gulls from Svalbard. Furthermore,

we examined the relationships between the distribution of these organohalogens and metabolites and the total water, protein, lipid and mineral contents in whole body homogenates. Such approach allowed examining the toxicokinetics implications of some contaminants being less affected by lipid solubility and more influenced by, among others, protein associations. 2. Materials and methods 2.1. Field protocol and sampling The collection of glaucous gull samples was conducted in Ny-Ålesund (79°N, 19°E) in the Svalbard archipelago during early July (2002), which corresponds to the end of the incubation period (28–30 days) in glaucous gulls. Adult (i.e., N5 years of age) glaucous gulls (n = 11 females, 8 males) were captured at the Ny-Ålesund refuse dump using a cannon net triggered by a remote controlled unit. Once captured, the initial body mass of the birds was recorded, as well as various morphometric measurements (i.e., head, bill, wing and tarsus length). Each bird was given a colored identification band placed around the tarsus before they were released into a sheltered outdoor enclosure. For the purpose of a validation study of water flux and body composition in glaucous gulls (Shaffer et al., 2006), equal numbers of males and females were assigned randomly to four treatment groups [group A (no food or water), group B (food only), group C (water only) and group D (both food and water)] to reflect the natural nutritional conditions of randomly-captured glaucous gulls in the field. The fed individuals were given juvenile Arctic cod (Boreogadus saida). Further details on the field protocol can be found in Shaffer et al. (2006). A blood sample (3 mL) was collected from each individual upon experiment completion, that is, three days after capture, and the birds were sacrificed. Blood samples and whole bird carcasses were kept frozen (− 80 °C) until laboratory analyses. A detailed description of blood sampling and sample processing is provided by Verreault et al. (2005a). All field methods employed in this study were approved by the Governor of Svalbard (2002/00483-2 a. 512/2) and the Norwegian National Animal Research Authority (S1030/02).

2.2. Body composition analysis To determine the body composition of glaucous gulls, frozen carcasses were thawed, reweighed, and the liver, heart and kidneys were collected and accurately weighed. Based on published data, we hypothesized that the heart, kidneys and liver had a comparable compositional makeup (i.e., water, lipid, protein and mineral contents) as the rest of the body (Shaffer et al., 2006). The heart and kidneys had no, or negligible adrenal/pericardial fat deposits. Liver was used for chemical analyses (see the following section), whereas the heart and kidneys were saved for other analyses, although these will not be discussed in the present study. The summed weights of the heart and kidneys represented on average 1.7% of the total body weight of glaucous gulls. It is documented that in birds the kidneys and heart are negligible storage tissues for lipophilic substances, and thus do not contribute significantly to the total contaminant body burden, as for example reported in pigeons dosed with PCBs (Borlakoglu et al., 1991). The remainder of the carcass was homogenized in a food grinder with (n = 9) or without (n = 10) feathers. These two distinct carcass processing methods were utilized based on the knowledge that they yield different body composition essentially by lipid/protein pool dilution, and thus would enhance the inter-individual variations in proportions of whole body constituents using a very small sample size, as in the present study (n = 19). Five pre-weighed aliquots (mean ± 1 standard error (SE): 22.5 ± 2.8 g) of whole homogenate samples were dried overnight in an oven at 55 °C until their mass remained constant. The total body water (TBW) content in the whole homogenate was calculated based on the bird's final body mass (FBM) recorded after euthanasia and the proportion of water evaporated from the whole homogenate (i.e., mean of the five replicates). The extractable lipid content in dried whole homogenate aliquots was determined through Soxhlet extraction using 300 mL of diethyl ether and 95% ethanol (1:3 volume ratio), and the total body fat (TBF) was calculated for the FBM. An aliquot of the fat-free and water-free whole homogenate tissue (i.e., containing proteins, carbohydrates and minerals) that


825

was left from the fat extraction was burned overnight in a muffle furnace at 500 °C. The total ash content (TAC) obtained from this tissue aliquot, which was composed largely of minerals (i.e., mainly calcium and phosphorus), was calculated for the FBM. The total ash-free lean dry mass (TAFLDM) representing the content of primarily proteins (i.e., typically 91–94% of TAFLDM) and carbohydrates in the whole homogenate, was calculated from the following equation: TAFLDM = TBM − (TBW + TBF + TAC).

BDE28 ISs that yielded variable results. Nonetheless, the OH-PCBs/-PBDEs, which were determined using an IS method, were automatically recoverycorrected. Blank samples showed negligible background contamination for all analyte classes. The duplicate extractions and injections demonstrated on average 10% and 5%, respectively, analytical variation of selected compound concentrations. The method limits of quantification (MLOQ) were set as 10 times the signal-to-noise ratio (S/N).

2.3. Chemical analysis

2.4. Statistical treatment

The extractable lipid content in blood samples was determined by the sulfophospho-vanillin reaction using pure olive oil as a calibrant (Frings et al., 1972). The analytical procedures employed for the determination of phenolic and less polar chlorinated and brominated compounds in glaucous gull samples were based on methods by Verreault et al. (2005a,b) with modifications described here. Briefly, blood samples (3 g) were spiked with internal standards (ISs), which included a 13C12–PCB mixture (six congeners for PCBs and CHLs), 3MeSO2-2-CH3-2′,3′,4′,5,5′-pentaCB (for MeSO2-PCBs/-p,p′-DDE), BDE30 and 71 (for PBDEs, MeO-PBDEs, total-(α)-HBCD and PBBs), a 13C12–OHPCB mixture (four congeners) and 2′-OH-BDE28, followed by acidification with 1 mL of 6 M HCl and addition of 3 mL of 2-propanol. The denatured blood was extracted three times with 6 mL of methyl-tert-butyl-ether (MtBE)/nhexane (50:50 volume ratio). Sub-samples of whole homogenate (with [5 g] or without [2.5 g] feathers) and liver (1.5 g) samples were homogenized with anhydrous Na2SO4 (8:1 weight ratio), transferred to an extraction column, spiked with the ISs and extracted with 200 mL of dichloromethane (DCM)/nhexane (50:50 volume ratio). Whole homogenate and liver extracts were concentrated and adjusted to a volume of 10 mL, and the lipid content was determined gravimetrically from a 10% (by volume) aliquot. To the denatured blood, whole homogenate and liver extracts, 6 mL of 1% KCl (weight/volume) was added, and the organic phase was partitioned with 6 mL of 1 M KOH in ultrapure ethanol and distilled water (50:50 volume ratio). Two phases were obtained: an aqueous phase containing the deprotonated phenolics and an organic phase containing the less polar compounds and MeSO2-PCBs/-p,p′-DDE. The aqueous phase was acidified with H2SO4 and the phenolics were back-extracted with 6 mL of MtBE/n-hexane (50:50 volume ratio), then dried over Na2SO4 and derivatized to their MeO-analogues via a methylation reaction using diazomethane. The MeO-compound derivatives were cleaned up on an acidified silica gel column (5.0 g, 22% H2SO4) (grade 62, 60– 200 mesh, 150 Å) (Aldrich Chemicals, Milwaukee, WI, USA) by elution with 50 mL of DCM/n-hexane (15:85 volume ratio), in preparation for gas chromatography–mass spectrometry (GC–MS) quantification. The organic phase from the partitioning step was concentrated and fractionated with a Florisil® column (8.0 g, 1.2% H2O deactivated) (magnesium silicate, F100– 500, 60–100 mesh) (Fisher Scientific, Ottawa, ON, Canada). Two fractions were collected. The first fraction (F1), containing the neutral chlorinated and brominated analytes, was eluted with 75 mL of DCM/n-hexane (50:50 volume ratio). The second fraction (F2), containing the MeSO2-PCBs/-p,p′-DDE, was eluted with 80 mL of methanol/DCM (7:93 volume ratio) and concentrated to approximately 1 mL. The F2 was further cleaned up on a basic alumina column (3.0 g, 2.3% H2O deactivated) (Brockman activity grade 1, 60–325 mesh) (Fisher Scientific) with 50 mL of DCM/n-hexane (50:50 volume ratio) to remove possible co-eluting biogenic artifacts. The F1 and F2 were concentrated for quantification using GC–MS (Agilent 6890; Agilent Technologies, Palo Alto, CA, USA). GC separation was performed using a fused silica DB-5 capillary column (30 m, 0.25 mm i.d., 0.25 μm film thickness) (J&W Scientific, Folsom, CA, USA). GC–MS (electron capture negative ionization (ECNI)) was used for all analytes except the PCBs and CHLs, which were quantified using GC–MS (electron impact (EI)). GC–MS (ECNI) utilized methane as a buffer gas. GC–MS (ECNI) and GC–MS (EI) analyte determinations were accomplished in the selected ion-monitoring (SIM) mode. Quality assurance and quality control procedures included standard reference material (polar bear plasma #01-2004) from the National Wildlife Research Centre (Ottawa, Canada), method blanks, duplicate extractions and injections of standard compounds and cleaned-up glaucous gull sample extracts for each block of six samples to monitor for quantitative reproducibility and instrument sensitivity. The variation of recoveries based on the added ISs showed great consistency, with the exception of the 13C12–OH-PCB and 2′-OH-

The organohalogen pattern differences between glaucous gull blood, liver and whole body homogenate samples were investigated using principal component (PC) analysis on the correlation matrix. This was done by extracting PCs from the relative proportions (wet weight basis) of ten organohalogen sums (Σ) composed of closely-related compounds (i.e., ΣPCB, ΣCHL, ΣMeSO2PCB, 3-MeSO2-p,p′-DDE, ΣOH-PCB, ΣOH-PBDE, ΣMeO-PBDE, ΣPBDE, BB101 and total-(α)-HBCD) or individual compounds and congeners within each class to the total organohalogen concentrations. Individual organohalogen congeners/compounds were included in sums only if they were detected in 60% or more of the samples in a given matrix. For these compounds, the samples with concentrations below the MLOQs were assigned a randomly-generated value between zero and the compound-specific MLOQ using a common software random function. The values obtained respected the criterion of normal distribution, and thus did not introduce biased variance in the dataset. A complete list of the congeners/compounds composing the organohalogen sums is provided in Table 1. Relative concentration proportions were arcsinetransformed prior to the PC analysis. The PCs with eigenvalues above one were considered to account for a significant contribution to the total variance according to the latent root criterion (Hair et al., 1998). Analytes with correlation coefficients in the PC analysis (i.e., PC loadings) greater than ±0.65 on any PCs were considered significant. The differences between tissues/blood in organohalogen concentrations (log10-transformed) and patterns (i.e., arcsine-transformed proportions to sums) were investigated using the analysis of variance (ANOVA), followed by the Fisher post-hoc test. Correlations between two variables were expressed using the Pearson coefficient r. The statistical package utilized was Statistica® (StatSoft, Tulsa, OK, U.S.A.) and α was set at 0.05.

3. Results and discussion 3.1. Whole body composition Water made up nearly two-thirds of the whole body homogenate mass, followed by proteins (∼ 1/5), minerals (∼ 1/10) and lipids (∼ 1/ 20), which was in good agreement with the published data reviewed elsewhere (Clawson et al., 1991). No significant sex and treatmentrelated effect was found on the whole body composition of present glaucous gulls. Nonetheless, carcasses processed with or without feathers yielded distinct body composition profiles (Fig. 1). Whole body homogenate samples for which the feathers were left on the carcass contained 5% more proteins. This enhanced contribution of proteinic components in the feathered carcasses was a result of the large amount of keratins (N 80% of plumage dry weight: Murphy and King, 1982; Murphy et al., 1990). Furthermore, lipids comprised a proportionally greater amount of the defeathered carcass weight. The total extractable lipid contents in the homogenized carcasses obtained from two different procedures (see the sections Body composition analysis and Chemical analysis) were nearly identical (r2 = 0.97; p b 0.001). Moreover, lipid contents extracted during chemical analysis, i.e., gravimetrically or via the sulfo-phospho-vanillin reactions, varied within a 10-fold range among the blood, liver and whole homogenate samples (Table 1). 3.2. Tissue-specific contaminant distribution Concentrations (wet weight basis) of a selection of chlorinated and brominated chemicals of legacy and more recent environmental

826


Table 1 Mean (± 1 standard error (SE)) and range of extractable lipids and sum (Σ) or individual concentrations (ng/g wet weight) of selected chlorinated and brominated contaminants and metabolically-derived analogues in blood, liver and whole body homogenate samples (with or without feathers) collected from adult glaucous gulls in Svalbard Blood (n = 19)

Lipid (%)

a

ΣCHL b ΣPCB c ΣOH-PCB d ΣMeSO2-PCB e 3-MeSO2-p,p′-DDE Total-(α)-HBCD BB101 ΣPBDE f ΣMeO-PBDE g ΣOH-PBDE h

Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range Mean ± SE Range

0.55 ± 0.04 0.28–0.95 104 ± 29.8 16.4–502 1853 ± 722 327–12,619 52.5 ± 18.3 4.36–265 2.02 ± 0.69 0.45–12.3 0.36 ± 0.22 0.03–3.84 3.29 ± 0.87 0.51–11.2 0.28 ± 0.13 N.D.–2.14 51.5 ± 15.9 8.97–244 2.78 ± 0.80 0.39–13.4 3.54 ± 0.97 N.D.–13.2

Liver (n = 19)

4.19 ± 0.16 3.13–6.30 1399 ± 356 169–5606 20,114 ± 6036 3115–106,406 28.5 ± 14.0 2.49–276 24.7 ± 6.45 3.82–117 1.28 ± 0.71 0.03–13.7 75.6 ± 5.88 5.88–292 2.04 ± 0.68 0.11–9.80 522 ± 154 62.9–2657 32.2 ± 12.4 4.37–233 3.57 ± 2.83 N.D.–54.4

Whole body homogenate With feathers (n = 9)

Without feathers (n = 10)

1.87 ± 0.42 0.80–4.04 355 ± 60.4 117–659 4733 ± 1035 1363–10,712 9.35 ± 2.92 1.59–25.7 3.58 ± 0.89 0.91–9.42 0.53 ± 0.32 0.13–3.09 117 ± 22.2 52.6–270 0.65 ± 0.29 0.10–2.77 178 ± 35.5 72.1–380 9.23 ± 1.86 3.64–20.6 0.87 ± 0.45 N.D.–3.59

4.61 ± 0.72 1.67–7.74 701 ± 173 256–2034 8319 ± 1538 4291–17,919 4.84 ± 1.21 2.48–13.9 6.97 ± 1.56 3.53–18.6 0.34 ± 0.08 0.10–0.71 91.0 ± 16.0 38.4–194 1.50 ± 0.64 0.15–5.19 202 ± 20.1 143–321 19.4 ± 4.28 5.43–40.3 0.27 ± 0.07 N.D.–0.57

N.D.: Not detected. a Total lipid content extracted during chemical analysis. b ΣCHL: heptachlor epoxide, trans-nonachlor, cis-nonachlor, trans-chlordane, cis-chlordane and oxychlordane. c ΣPCB: CB31/28, 49, 52, 56/60, 64/41, 66, 74, 85, 87, 95, 99, 101/90, 105, 110, 118, 128, 138, 146, 149, 153, 156, 158, 170/190, 171, 172, 174, 177, 178, 179, 180, 183, 187, 194, 195, 196/203, 200, 201, 202 and 206. d ΣOH-PCB: 4-OH-CB97, 4-OH-CB107/4′-OH-CB108, 4′-OH-CB120, 4-OH-CB134, 3′-OH-CB138, 4-OH-CB146, 4′-OH-CB159, 4-OH-CB162, 4-OHCB163, 4′-OH-CB172, 4′-OH-CB177, 4-OH-CB178, 3′-OH-CB180, 3′-OH-CB183, 3′-OH-CB184, 4-OH-CB187, 4-OH-CB193, 4′-OH-CB198, 4′-OH-CB199, 4′-OH-CB200, 4′-OH-CB201, 4′-OH-CB202, 4,4′-diOH-CB202, 3′-OH-CB203 and 4′-OH-CB208. e ΣMeSO2-PCB: 3-MeSO2-CB49, 4-MeSO2-CB49, 4-MeSO2-CB64, 3-MeSO2-CB70, 4-MeSO2-CB70, 4-MeSO2-CB87, 3-MeSO2-CB101, 4-MeSO2-CB101, 4MeSO2-CB110, 4-MeSO2-CB132, 3-MeSO2-CB149 and 4-MeSO2-CB174. f ΣPBDE: BDE28, 47, 66, 85, 99, 100, 138, 153, 154/BB153 (co-eluting), 183, 190 and 209. g ΣMeO-PBDE: 2′-MeO-BDE28, 4-MeO-BDE42, 3-MeO-BDE47, 6-MeO-BDE47, 4′-MeO-BDE49 and 6′-MeO-BDE49. h ΣOH-PBDE: 3-OH-BDE47, 5-OH-BDE47, 6-OH-BDE47, 4′-OH-BDE49 and 6′-OH-BDE49.

concern, as well as their metabolic products and naturally-occurring analogues were determined in blood, liver and whole homogenate samples (Table 1). Organohalogen class and individual congener/ compound concentrations differed largely among blood, liver and whole homogenate samples on some occasions by several orders of magnitude. The results of the legacy organochlorines were corroborated by those reported in previous surveys for paired blood–liver and liver–whole body samples of Svalbard glaucous gulls (Henriksen et al., 1998) and Great Lakes herring gulls (Braune and Norstrom, 1989), respectively. Present liver-to-blood and liver-to-whole body concentration ratios for congeners/compounds of selected major PCB congeners and CHL compounds were within a range comparable to that reported in those studies (data not shown). No significant sex and treatment-related effect was found on the concentrations and contaminant class proportions between blood, liver and whole body homogenate samples of glaucous gulls. This non-significant difference in contaminant burdens (and contaminant class proportions) between blood and tissue samples collected from male and female glaucous gulls post-egg-laying was in partial disagreement with previous investigations in this species (e.g., Henriksen et al., 1998; Verreault et al., 2005a,b). This could be the result of a low sample size. Nevertheless, it could not be completely disregarded that since the

individuals were collected at the end of the incubation period (i.e., early July), females may have partially compensated for the loss of body resources associated with egg formation and laying. Therefore, females

Fig. 1. Mean percentage (+ 1 standard error) distribution of whole body constituents in whole Svalbard glaucous gull carcasses processed with or without feathers. Significant (p ≤ 0.05) differences between mean percentages are marked with an asterisk (⁎).


827

Fig. 2. Proportions of ten major chlorinated and brominated contaminant classes or individual compounds to the total organohalogen concentrations plotted using the two first principal components (PCs), PC 1 and PC 2. Mean (±1 standard error) factor scores (right biplot) are showed for blood, liver and whole body homogenate samples (with or without feathers) of Svalbard glaucous gulls. The percent variability explained by PC 1 and PC 2 is provided.

may have reached their steady-states in contaminant levels (Bustnes et al., 2003). A detailed discussion on the resource mobilization and maternal transfer of brominated and chlorinated contaminants and metabolites in Svalbard glaucous gulls can be found in Verreault et al. (2006). The mean (±1 SE) total burden of organohalogens and metabolites in present adult glaucous gull males and females (mean body mass: 1.49 ± 0.05 kg), calculated from the combined organohalogen concentrations in whole blood aliquots, liver and body homogenates, was 13.3 ± 2.0 mg (range: 3.3–33.0 mg). Based on the wide annual distribution and current population size of Svalbard glaucous gulls, this is indicative of the relative importance of breeding glaucous gulls as biological vectors of organohalogen contaminants within the Norwegian Arctic marine environment (Evenset et al., 2007). By comparison, in another Arctic apex marine predator, the polar bear (Ursus maritimus) from East Greenland, where adult males and females had a mean body mass of 449 ± 194 kg, for the identical suite of chlorinated and brominated compounds and metabolites the total body burden was roughly 3-fold greater (mg/kg body mass basis) relative to the present Svalbard glaucous gulls (Gebbink et al., submitted for publication). This glaucous gull/polar bear comparison should be interpreted with caution as total contaminant body burdens are likely considerably variable as a function of seasonally-dependent energetic status. A discussion on the environmental significance and toxicological implications of present organohalogen classes in Svalbard glaucous gulls can be found in Verreault et al. (2005a,b, 2006).

The contaminant patterns among glaucous gull samples were compared by examining the structure in the relationships between the proportions of the ten major analyte classes and individual compounds to the total organohalogen concentration using the first two PCs, PC 1 and PC 2, which made up 49% of the total variance (Fig. 2). PC 3, which also had an eigenvalue above one, accounted for less than 14% of the total variance and did not provide noteworthy additional results. Blood, relative to the tissues, was distinguished by significantly higher (p ≤ 0.001) proportions of ΣOH-PCB and ΣOH-PBDE. A higher relative proportion in blood was also observed for the p,p′-DDE metabolite 3-MeSO2-p,p′-DDE, although it did not comply with the criterion of significance (p ≤ 0.06). Comparatively, the ΣCHL fraction in the total organohalogen concentrations measured in blood was lower, albeit not significantly (p ≤ 0.06), compared to liver and whole homogenate samples. Liver showed a significantly (p ≤ 0.005) more pronounced accumulation for ΣPCB relative to blood and the feathered whole homogenate. The contaminant patterns among glaucous gull samples were also investigated within each of the organohalogen classes by examining the proportions of individual congeners/compounds making up those classes to their summed concentrations using PC 1 and PC 2. No significant difference in the distribution of individual congeners/ compounds among the analyte classes (PCBs, CHLs, MeSO2-PCBs/-p, p′-DDE, OH-PCBs/-PBDEs and MeO-PBDEs) was found, except for the classes comprised of less polar brominated compounds (PBDEs, PBBs and total-(α)-HBCD) (Fig. 3). PC 1 and PC 2 made up 60% of the

Fig. 3. Proportions of ten individual major brominated compounds to the total brominated compound concentrations plotted using the two first principal components (PCs), PC 1 and PC 2. Mean (± 1 standard error) factor scores (right biplot) are showed for blood, liver and whole body homogenate samples (with or without feathers) of Svalbard glaucous gulls. The percent variability explained by PC 1 and PC 2 is provided.

828


total variance, whereas the contribution from PC 3 (eigenvalue = 1.09) was 11%. Total-(α)-HBCD had significantly (p ≤0.001) higher proportions in whole homogenate samples with and without feathers compared to blood and liver. In contrast, proportions of BDE47 and 99 were significantly (p ≤ 0.02) greater in blood and liver comparatively to the two whole homogenate types. Based on the present suite of chlorinated and brominated compounds (MeSO2- and MeO-/OHsubstituted or unsubstituted), it appears that the proportions of different compound classes, and on some occasions individual congeners and compounds, making up the total organohalogen content are not conserved among blood and tissues of glaucous gulls. Differences in compound class composition were also detected between whole homogenate carcasses including or excluding feathers (see the following section). Such disparity in organohalogen composition among blood and tissues is likely to be explained by tissue-specific characteristics (e.g., metabolic capacity in terms of enzyme profile, activity and/or induction) as well as protective mechanisms such as macromolecule binding (e.g., hormone carrier proteins). One possible way to explore this is to define the linkages between organohalogen concentrations and whole body composition of glaucous gulls. 3.3. Associations to whole body constituents Compound patterns in whole glaucous gull carcasses (homogenized with or without plumage) with known body composition were investigated using a similar PC approach as described in the previous sections. The two first PCs accounted for 51% of the total variance, while the contribution from PC 3 was 15%. Even though PC 2 and PC 3 had a nearly comparable percentage contribution to the total variance, their graphical representation as a function of PC 1 yielded corroborative results. As shown in Fig. 4, the difference in total lipid and protein contents between the feathered and non-feathered body homogenates led to a notable variation in the distribution of the monitored organohalogens. Total proteins isolated from the lean-dry whole carcasses were significantly associated to the proportions (i.e., to the summed organohalogen concentrations) of ΣOH-PCB and ΣPBDE. Hence, contents of proteins in whole carcasses correlated positively with proportions of ΣOH-PCB (r2 = 0.36; p = 0.008) and ΣPBDE concentrations (r2 = 0.41; p = 0.004). A consistent, but non-significant tendency was found for the total-(α)-HBCD and ΣOH-PBDE proportions. In contrast, the total contents of lipids did not show any definitive association with any of the organohalogen classes. However, the spatial organization of the variables (Fig. 4) pointed to

a tendency for a weak association between total lipids and the ΣPCB proportions. Proportions of individual congeners/compounds in PC biplots did not reveal significant associations with proteins or lipids for each of the organohalogen classes quantified in whole body homogenate samples. This may perhaps be explained by a low sample size and important variations among glaucous gull individuals. There are several factors, and thus plausible explanations as to the observed lipid and protein associations to the distribution of the organohalogens determined in glaucous gulls. Current understanding of PCB biotransformation and detoxification mechanisms suggests that the OH-PCBs retained in vertebrates mainly are derived from enzymemediated processes (e.g., cytochrome P450 (CYP) monooxygenases), whereas the OH-PBDEs may be formed via metabolism of major PBDEs and/or accumulated as naturally-occurring compounds (e.g., via formation in algae and sponges) (Hakk and Letcher, 2003; Letcher et al., 2000; Malmberg et al., 2004, 2005; Marsh et al., 2006). The occurrence of OH-PCBs/-PBDEs has been confirmed in various tissues and body compartments of several environmentally and experimentally PCB and PBDE-exposed animals (Gebbink et al., submitted for publication; Hakk and Letcher, 2003; Letcher et al., 2000; Verreault et al., 2005b). The present state of knowledge on the toxicokinetics and accumulation of OH-PCBs/-PBDEs in birds and mammals also suggests that the concentrations and tissue distribution of these phenolic metabolites, and/or natural products, are strongly influenced by mechanisms such as protein binding. One mechanism leading to OH-PCB/-PBDE retention that has received most scientific attention operates through the competitive binding of OH-PCBs/-PBDEs with natural ligands for binding sites. This also protects the OH group from enzyme-mediated conjugative processes such as with the glucuronic acid. Important structural similarities and binding affinities have been reported in in vivo and in vitro studies between certain OH-PCB congeners and the thyroxine (T4) carrier protein, transthyretin (TTR) (Legler et al., 2002; Letcher et al., 2000; Purkey et al., 2004). Studies on OH-PBDEs also have revealed occasionally high TTR binding potencies for the OH-PBDE congeners (Legler et al., 2002; Legler and Brouwer, 2003; Malmberg, 2004; Meerts et al., 2000). For example, Malmberg (2004) showed that the para-OH-substituted congeners 3OH-BDE47, 4-OH-BDE42 and 4′-OH-BDE49, which are all identified in glaucous gulls (Table 1; Verreault et al., 2005b), had binding affinities to human TTR that were four times stronger than T4. Despite the relatively high, although compound-specific lipophilic properties of PCBs, MeSO2-PCBs/-p,p′-DDE, CHLs, PBDEs (and

Fig. 4. Proportions of ten major chlorinated and brominated contaminant classes or individual compounds to the total organohalogen concentrations, as well as proportions of total lipid and protein contents to the whole body composition, plotted using the two first principal components (PCs), PC 1 and PC 2. Mean (±1 standard error) factor scores (right biplot) are showed for whole body homogenates processed with or without feathers in Svalbard glaucous gulls. The percent variability explained by PC 1 and PC 2 is provided.


MeO-PBDEs) and total-(α)-HBCD, these contaminant classes showed distinct tissue-specific retention profiles and body constituent associations. This is highly indicative that some of these contaminant classes may be affected concomitantly by lipid solubility and protein associations, as well as specific biochemical mechanisms (e.g., oxidative and conjugative processes). Unexpectedly, ΣCHL concentrations were associated with neither lipids nor proteins (Fig. 4). As ΣCHL concentrations were composed largely of the cis- and trans-chlordane metabolite oxychlordane (range: 58–83%), this lack of relationship with the body constituents, particularly the bulk lipids, may be due to biotransformation-dependent processes. In fact, it has been shown in Larids that oxychlordane formation and retention are species-specific as a result of the inherent differences in biotransformation capacities among species (Fisk et al., 2001). With respect to the MeSO2-PCBs/-p, p′-DDE, these metabolites are reported to possess generally lower octanol–water partition coefficients (KOW) weighed against their precursor PCBs, and exhibit higher binding affinities to proteins (Letcher et al., 2000). In line with these findings, albumin, an important T4 binding protein in birds (McNabb, 2000), as well as fatty acidbinding protein (FABP) and major bile and urine protein interactions have been demonstrated in rats for several congeners of PBDEs (Hakk et al., 2002, 2006; Mörck et al., 2003). Furthermore, a study by Hamers et al. (2006) recently reported the T4-TTR (human) competing potencies for selected PBDEs and structural isomers of HBCD (α, β and γ). In this study, non-hydroxylated PBDE congeners and HBCD isomers had low T4-competing affinities. To our knowledge, no work has been carried out in avian species with respect to affinities between environmentallyrelevant brominated flame retardants and proteins.

4. Conclusions In line with the results of experimental designs, the present study of glaucous gulls supports generally enhanced protein association for selected major PCB and PBDE compounds with phenyl group substituents (i.e., OH) that modify physicochemical characteristic such as polarity, but also PBDEs and total-(α)-HBCD. However, substances typically shown as having lipophilic properties such as PCBs, CHLs and to some extent PBDEs and MeSO2-PCBs/-p,p-DDE, did not show clear associations with the total lipid pool isolated from homogenized whole glaucous gull carcasses. The present study suggests that tissue-specific protein and lipid contents play an important role, together with numerous other biochemical processes, in the toxicokinetics and fate of chlorinated and brominated contaminants and metabolic and natural products. An enhanced, selective retention of certain organohalogen classes in given tissues/body compartments may thus lead to site-specific toxicological actions and adverse effects in the highly-contaminated Svalbard glaucous gulls. Further studies are warranted to investigate the mechanisms of macromolecule binding interactions and organohalogen accumulation in avian species. Acknowledgements This project received funding from the Norwegian Polar Institute and the Norwegian Research Council (to J.V.) as well as supplemental funding from the Wildlife Toxicology and Disease Program, NWRC (to R.J.L.). We wish to thank Prof. Scott A. Shaffer for assistance with the design of the study and

829

fieldwork. We also would like to thank Dr. Ross Norstrom and anonymous reviewers for their constructive comments on an early draft of the manuscript. References Barrett RT, Skaare JU, Gabrielsen GW. Recent changes in levels of persistent organochlorines and mercury in eggs of seabirds from the Barents Sea. Environ Pollut 1996;92:13–8. Borlakoglu JT, Wilkins JP, Dils RR. Distribution and elimination in vivo of polychlorinated biphenyl (PCB) isomers and congeners in the pigeon. Xenobiotica 1991;21:433–45. Braune BM, Norstrom RJ. Dynamics of organochlorine compounds in herring gulls: III, tissue distribution and bioaccumulation in Lake Ontario gulls. Environ Toxicol Chem 1989;8:957–68. Bustnes JO, Bakken V, Skaare JU, Erikstad KE. Age and accumulation of persistent organochlorines: a study of Arctic-breeding glaucous gulls (Larus hyperboreus). Environ Toxicol Chem 2003;22:2173–9. Clawson AJ, Garlich JD, Coffey MT, Pond WG. Nutritional, physiological, genetic, sex, and age effects on fat-free dry matter composition of the body in avian, fish, and mammalian species: a review. J Anim Sci 1991;69:3617–44. Evenset A, Carroll J, Christensen GN, Kallenborn R, Gregor D, Gabrielsen GW. Seabird guano is an efficient conveyer of persistent organic pollutants (POPs) to arctic lake ecosystems. Environ Sci Technol 2007;41:1173–9. Frings CS, Fendley TW, Dunn RT, Queen CA. Improved determination of total serum lipids by the sulfo-phospho-vanillin reaction. Clin Chem 1972;18:673–4. Fisk AT, Moisey J, Hobson KA, Karnovsky NJ, Norstrom RJ. Chlordane components and metabolites in seven species of Arctic seabirds from the Northwater Polynya: relationships with stable isotopes of nitrogen and enantiomeric fractions of chiral components. Environ Pollut 2001;113: 225–38. Gebbink WA, Sonne C, Dietz R, Kirkegaard M, Born EW, Muir DCG, Letcher RJ. Tissue-specific composition and body burdens of brominated and chlorinated contaminant classes in polar bears (Ursus maritimus) from East Greenland. Environ Sci Technol. March submitted for publication. Hair JF, Anderson RE, Tatham RL, Black WC. Multivariate data analysis. 5th ed. New-Jersey, USA: Prentice Hall; 1998. p. 87–138. Hakk H, Letcher RJ. Metabolism in the toxicokinetics and fate of brominated flame retardants — a review. Environ Int 2003;29:801–28. Hakk H, Larsen G, Bergman Å, Orn U. Binding of brominated diphenyl ethers to male rat carrier proteins. Xenobiotica 2002;32:1079–1091. Hakk H, Huwe J, Low M, Rutherford D, Larsen G. Tissue disposition, excretion and metabolism of 2,2′,4,4′,6-pentabromodiphenyl ether (BDE-100) in male Sprague–Dawley rats. Xenobiotica 2006;36:79–94. Hamers T, Kamstra JH, Sonneveld E, Murk AJ, Kester MH, Andersson PL, et al. In vitro profiling of the endocrine-disrupting potency of brominated flame retardants. Toxicol Sci 2006;92:157–73. Hebert CE, Norstrom RJ, Weseloh DVC. A quarter century of environmental surveillance: the Canadian Wildlife Service's Great Lakes Herring Gull Monitoring Program. Environ Rev 1999;7:147–66. Henriksen EO, Gabrielsen GW, Skaare JU. Validation of the use of blood samples to assess tissue concentrations of organochlorines in glaucous gulls. Chemosphere 1998;37:2627–43. Legler J, Brouwer A. Are brominated flame retardants endocrine disruptors? Environ Int 2003;29:879–85. Legler J, Cenijn PH, Malmberg T, Bergman Å, Brouwer A. Determination of the endocrine disrupting potency of hydroxylated PCBs and flame retardants with in vitro bioassays. Organohalog Compd 2002;56:53–6. Letcher RJ, Klasson-Wehler E, Bergman Å. Methyl sulfone and hydroxylated metabolites of polychlorinated biphenyls. In: Paasivita J, editor. The handbook of environmental chemistry — new types of persistent halogenated compounds, vol. 3. Heidelberg, Germany: Springer-Verlag; 2000. p. 315–59. part K. Malmberg T. Identification and characterization of hydroxylated PCB and PBDE metabolites in blood: congener specific synthesis and analysis. Ph.D. thesis, Stockholm University. Stockholm, Sweden; 2004.

830


Malmberg T, Hoogstraate J, Bergman Å, Klasson-Wehler E. Pharmacokinetics of two major hydroxylated polychlorinated biphenyl metabolites with specific retention in rat blood. Xenobiotica 2004;34:581–9. Malmberg T, Athanasiadou M, Marsh G, Brandt I, Bergman Å. Identification of hydroxylated polybrominated diphenyl ether metabolites in blood plasma from polybrominated diphenyl ether exposed rats. Environ Sci Technol 2005;39:5342–8. Marsh G, Athanasiadou M, Athanassiadis I, Sandholm A. Identification of hydroxylated metabolites in 2,2′,4,4′-tetrabromodiphenyl ether exposed rats. Chemosphere 2006;63:690–7. McNabb FMA. Thyroids. In: Whittow GC, editor. Sturkie's avian physiology. 5th ed. London: Academic Press; 2000. p. 461–471. Meerts IA, van Zanden JJ, Luijks EA, van Leeuwen-Bol I, Marsh G, Jakobsson E, et al. Potent competitive interactions of some brominated flame retardants and related compounds with human transthyretin in vitro. Toxicol Sci 2000;56:95-104. Mineau P, Fox GA, Norstrom RJ, Weseloh DV, Hallett DJ, Ellenton JA. Using the herring gull to monitor levels and effects of organochlorine contamination in the Canadian Great Lakes. In: Nriagu JO, Simmons MS, editors. Toxic contaminants in the Great Lakes. New York, USA: John Wiley & Sons; 1984. p. 426–52. Mörck A, Hakk H, Orn U, Klasson-Wehler E. Decabromodiphenyl ether in the rat: absorption, distribution, metabolism, and excretion. Drug Metab Dispos 2003;31:900–7. Murphy ME, King JR. Amino acid composition of the plumage of the whitecrowed sparrow. Condor 1982;84:435–8. Murphy ME, King JR, Taruscio TG, Geupel GR. Amino acid composition of feather barbs and rachises in three species of Pygoscelid penguins: nutritional implications. Condor 1990;92:913–21.

Purkey HE, Palaninathan SK, Kent KC, Smith C, Safe SH, Sacchettini JC, et al. Hydroxylated polychlorinated biphenyls selectively bind transthyretin in blood and inhibit amyloidogenesis: rationalizing rodent PCB toxicity. Chem Biol 2004;11:1719–28. Savinova TN, Polder A, Gabrielsen GW, Skaare JU. Chlorinated hydrocarbons in seabirds from the Barents Sea area. Sci Total Environ 1995;160/ 161:497–504. Shaffer SA, Gabrielsen GW, Verreault J, Costa DP. Validation of water flux and body composition in glaucous gulls (Larus hyperboreus). Physiol Biochem Zool 2006;79:836–45. Verreault J, Letcher RJ, Muir DCG, Chu S, Gebbink WA, Gabrielsen GW. New organochlorine contaminants and metabolites in plasma and eggs of glaucous gulls (Larus hyperboreus) from the Norwegian Arctic. Environ Toxicol Chem 2005a;24:2486–99. Verreault J, Gabrielsen GW, Chu S, Muir DCG, Andersen M, Hamaed A, et al. Flame retardants and methoxylated and hydroxylated polybrominated diphenyl ethers in two Norwegian Arctic top-predators: glaucous gulls and polar bears. Environ Sci Technol 2005b;39:6021–8. Verreault J, Houde M, Gabrielsen GW, Berger U, Haukås M, Letcher RJ, et al. Perfluorinated alkyl substances in plasma, liver, brain and eggs of glaucous gulls (Larus hyperboreus) from the Norwegian Arctic. Environ Sci Technol 2005c;39:7439–45. Verreault J, Agudo Villa R, Gabrielsen GW, Skaare JU, Letcher RJ. Maternal transfer of organohalogen contaminants and metabolites to eggs of Arcticbreeding glaucous gulls. Environ Pollut 2006;144:1053–60.