Environ. Sci. Technol. 2003, 37, 832-839
Plasma-Associated Halogenated Phenolic Contaminants in Benthic and Pelagic Fish Species from the Detroit River HONGXIA LI, KEN G. DROUILLARD, ERIN BENNETT, G. DOUGLAS HAFFNER, AND ROBERT J. LETCHER* Great Lakes Institute for Environmental Research, University of Windsor, 401 Sunset Avenue, Windsor, Ontario N9B 3P4, Canada
Halogenated phenolic contaminants (HPCs), polychlorinated biphenyls (PCBs), and other organochlorine (OC) contaminants (DDTs, chlordanes, HCH, and octachlorostyrene (OCS)) were determined (ng/g, wet wt. basis) in the blood plasma of 6 piscivorous/pelagic- and 7 benthicfeeding fish species from the Detroit River. Seven to twelve hydroxylated (OH) PCB congeners with pentachloro- to nonachloro-substitution, pentachlorophenol (PCP) and 4-OHheptachlorostyrene (4-OH-HpCS), as well as a number of other unidentified HPCs, were detected in all the fish species studied. The concentrations of Σ-OH-PCB (0.57-129.5 ng/ g), 4-OH-HpCS (0.02-0.31 ng/g), and PCP (0.05-3.42 ng/ g) in comparison to other major polychlorinated contaminants (Σ-PCB (10.4-909.0 ng/g), Σ-chlordane (0.72-8.68 ng/g), Σ-DDT (2.10-66.57 ng/g), and Σ-HCH (0.04-0.84 ng/g)) demonstrate the importance of HPCs, especially OH-PCBs, as plasma-associated contaminants. OH-PCBs are most likely metabolites of PCBs, and 4-OH-HpCS is a likely metabolite of OCS. Our findings indicate that metabolic biotransformation is a factor in the bioaccumulation, pharmacokinetics, and fate of OCS and highly chlorinated PCB congeners that are generally more recalcitrant in fish. Target tissue exposure to circulating OH-PCBs was variable, and thus potential OH-PCB-mediated toxicological activity and effects (e.g., endocrine-related) and health risks are different among the fish species.
Introduction The Detroit River is a channel connecting Lake Huron and Lake Erie via Lake St. Clair. The sediments and aquatic biota, such as fish, from the Detroit River contain high concentrations of persistent polyhalogenated contaminants including polychlorinated biphenyls (PCBs) and organochlorine (OC) pesticides (e.g., DDTs, chlorobenzenes, and chlordanes) relative to other areas of the Great Lakes (1-5). Octachlorostyrene (OCS) is also a significant contaminant in sediment and aquatic biota from the Huron-Erie corridor due to recent and historical industrial inputs into the St. Clair River (2-5). Dietary uptake of PCBs and OCs has been identified as the most important exposure route for fish from the HuronErie corridor (3, 4). * Corresponding author phone: 519-253-3000 (ext. 3753); fax: 519971-3616; e-mail:
[email protected]. 832
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
Immunochemical and catalytic studies have indicated that cytochrome P4501A (CYP1A) enzyme activity is inducible in fish, whereas CYP2B-like iso-enzyme activity is generally low and noninducible (6, 7). Polycyclic aromatic hydrocarbons (PAHs), which do not generally accumulate in fish, are rapidly metabolized via cytochrome P450 1A (CYP1A) enzymemediated biotransformation to phenolic and epoxide metabolites (8). In contrast, fish have a limited capacity to degrade polyhalogenated contaminants, including PCBs, which tend to be eliminated largely as unchanged compounds to water via gills or to feces (9). However, the role of metabolic biotransformation on the total elimination rate of polyhalogenated contaminants in fish remains unclear (9, 10). Metabolic biotransformation is an important factor in determining the bioaccumulation, biomagnification, pharmacokinetics, and toxicities of organic contaminants in wildlife and humans. Organic contaminants are susceptible to Phase I oxidative metabolism that is mediated by the CYP enzyme system, and dependent on the compound structure and the CYP metabolic capacity of the species (11, 12). In experimental animals and humans, metabolism of non-ortho chlorine-substituted PCBs with vicinal H-atoms at the orthometa carbons is mediated via CYP1A, whereas PCBs with vicinal H-atoms at the meta-para carbons are preferred substrates of CYP2B and CYP3A like iso-enzymes (13). Regardless, depuration kinetics in fish generally indicate slow elimination of PCB congeners and OCS with log Kow (octanolwater partition coefficient) values of >5.5 (3, 9, 10). Hydroxylated (OH) PCBs are halogenated phenolic compounds (HPCs) that have been reported in the blood of humans, birds, and mammals, and are largely derived as metabolites from CYP-mediated Phase I metabolic processes (14). OH-PCBs, by example of the commonly detected congeners 4′-OH-CB-109, 4-OH-CB-146, and 4-OH-CB-187, have structures similar to those of the thyroid hormones thyronine (T3) and thyroxine (T4), and can effectively compete with the hormone binding sites of thyroid hormone transport proteins such as transthyretin (TTR) in mammals (15). PCP and 4-OH-HpCS can also compete with T4 for TTR (14, 16). Considering metabolic biotransformation of PCBs and other persistent OCs in fish is thought to be slow or negligible, there have been few studies that have attempted to assess biotransformation via the determination and quantification of retained or accumulated metabolites (14). OH-PCBs and other HPCs have received marginal attention as contaminants in fish. OH-PCBs have been identified or detected in the blood of salmon from the Baltic sea (14, 17) and recently in lake trout from the Great Lakes (18). Other HPCs such as OH-polybrominated diphenyl ether (OH-PBDE) and chlorophenol congeners, 4-OH-heptachlorostyrene (4OH-HpCS) and triclosan (a trichlorinated and hydroxylated diphenyl ether) have been identified in the plasma of humans (19, 20), Baltic salmon (Salmo salar) (17), and/or polar bear (Ursus maritimus) (21), although many other HPCs that were detected were not identified. Metabolism of meta-para chlorine-unsubstituted PCBs in fish has also been shown by the presence of persistent and bioaccumulative methyl sulfone PCBs (14) in two species of sculpin, the deepwater sculpin (Myoxocelphalus thompsoni) from Lake Michigan (22) and the four-horned sculpin (Myoxocephalus quadricornis) (23) from the Canadian arctic. Environmentally relevant OH-PCB metabolites and HPCs have been linked to biological and toxicological effects, especially the disruption of endocrine-related processes. OHPCBs, 4-OH-HpCS, and OH-PBDEs have been shown to elicit estrogenic and thyroidogenic effects in vitro and in vivo (14, 10.1021/es026215l CCC: $25.00
2003 American Chemical Society Published on Web 01/24/2003
TABLE 1. Species of Fish Collected from the Detroit River Including the Year and Month of Sampling, Fish Sizes and Weights, Number of Pools, and % Lipid in the Blood Plasma Collecteda common name
species
sampling dateb
total length (cm)
body weight (g)
N
lipid (%)c
black crappie* white bass* largemouth bass†,‡ northern pike*,‡ longnose gar* bowfin‡,¥
Pelagic and Piscivorous Pomoxis nigromaculatus 08/01 27.0-27.0 Morone chrysops 08/01 28.5-29.0 Micropterus salmoides 08/01 30.0-41.0 Esox lucius 08/01 38.5-77.0 Lepisosteus osseus 08/01 55.5-70.0 Amia calva 08/01 51.5-60.0
152-289 288-315 675-1230 292-2800 296-850 1450-1924
1 pool of 2 1 pool of 3 3 pools of 3 1 pool of 4 2 pools of 3 2 pools of 3
3.4 2.7 1.6 (1.6-1.7) 1.3 2.0 ((1.6-2.3) 1.0 (0.8-1.1)
brown bullhead‡ channel catfish¥ freshwater drum* white sucker† common carp†,‡,¥ bigmouth buffalo¥ lake sturgeon (Lake St. Clair)
Benthic Feeding Ameiurus nebulosus 09/01 18.1-34.2 Ictalurus puncyaus 07/01 48.4 Aplodinotus grunniens 05/02 35-40 Catostomus commersoni 08/01 47.0-60.0 Cyprinus carpio 08/01 37.4-71.0 Ictiobus cyprinellus 05/02 36.0-49.0 Acipenser fulvescens 05/01 116.0
250-567 900 Not available 1500-2600 1100-5240 1650-1900 9150
2 pools of 4 1 fish only 1 pool of 3 1 pool of 4 5 pools of 3 1 pool of 3 1 fish only
1.6 (0.8-2.3) 0.8 0.8 1.2 1.4 (1.1-1.7) 0.9 0.6
a The Materials and Methods section contains a description of procedures for blood sampling, plasma isolation, and the pooling of plasma from individuals of the same species. b All individuals were collected in areas in the mid to lower part of the Detroit River (i.e., between Lake St. Clair and Lake Erie) in the vicinity of Grosse Island, i.e., River Canard outlet*, Celeron Island†, Trenton Channel‡ or Turkey Creek outlet¥. c For fish where more than one pool was analyzed, mean and range (in brackets) of % lipid levels in the blood plasma are given.
21, 24). For example, OH-PCBs have been shown to modulate estrogen-dependent and vitellogenin cycles in juvenile rainbow trout (25). There is also an exposure risk to human consumers of fish containing OH-PCBs. A positive correlation was recently reported between elevated OH-PCBs in human blood plasma and the amount of fatty fish consumed by residents from the Baltic Sea area (26). Terrestrial sources of OH-PCBs have also been recently found in plants, which suggests that uptake by herbivore consumers may be possible (27). Given that biota and sediments from the Detroit River contain high PCB and OC concentrations (2-5), fish from this system may possess substantial CYP enzyme activity as a consequence of exposure to CYP enzyme-inducing contaminants. Therefore, a greater metabolic capability may exist for subsequent catalytic generation and exposure to HPC metabolites. In the present study, PCBs, OCs, 4-OH-HpCS, PCP, OH-PCBs, and MeSO2-PCBs were surveyed in the blood plasma of thirteen pelagic- and benthic-feeding fish species from the Detroit River.
Materials and Methods Sample Collection. Table 1 lists the species studied and summarizes sampling and biological details such as body length and weight ranges of the fish collected, and average plasma lipid contents of the pooled samples. Fish were collected in the mid to lower portions of the Detroit River in the vicinity of Grosse Island, i.e., River Canard outlet, Celeron Island, Trenton Channel, or Turkey Creek. The fish were collected mainly from August to September 2001, except Freshwater Drum and Bigmouth Buffalo, which were captured in May 2002. Fish were captured by gill net, trap net, or electrofishing techniques. Fish blood was collected by cardiac puncture using a 20-gauge needle and placed into 10-mL vacutubes (sodium heparin as the anticoagulant). The blood samples were immediately separated into plasma and red blood cells by centrifugation (3000 rpm for 5 min) and stored at -20 °C until analysis. Chemical Analysis. The lipid contents of plasma were determined using a colorimetric method based on the procedure described by Frings et al. (28), using pure olive oil as the calibration standard. Procedures used for analysis of PCBs, OCs, HPCs, and MeSO2-PCBs in plasma, with minor modifications, have been described elsewhere (14, 20, 22, 29, 30). The plasma from 3 or 4 individual fish per species
and collection site (Table 1) required pooling to permit adequate detection and quantification of polyhalogenated compounds, especially HPCs. Briefly, 4 g of pooled plasma was spiked with surrogate internal standards, followed by acidification with HCl (6 M, 1 mL) and addition of 2-propanol (3 mL). The denatured plasma was extracted three times with an equal volume of methyl-tert-butyl ether (MtBE)/ hexane (1:1). The organic extracts were partitioned with potassium hydroxide (1 M in 50% ethanol). The alkaline phase was acidified and the HPCs were re-extracted in hexane/ MtBE (1:1), dried over sodium sulfate, and derivatized to their methoxy-analogues via a methylation reaction using diazomethane. The methyoxy-HPCs were purified on a silica/ sulfuric acid (22%) column (5 g) and eluted with 50 mL of methylene chloride (DCM)/hexane (1:1). Samples were reduced to a final volume of 100 µL for analysis by gas chromatography with micro-electron capture detection (GCµECD) and confirmed by GC-mass spectrometric (MS) analysis in the electron capture, negative ion (ECNI) mode. The hexane phase from the partitioning with alkali, containing neutral compounds such as PCB, was cleaned up on a Florisil column (1.2% w/w H2O deactivated, 8 g). Fraction 1 from Florisil (38 mL of n-hexane) contained PCB congeners and DDE. Fraction 2 (34 mL of 15% DCM/hexane) contained OCs (such as DDT-related components, chlordane, HCH, and OCS). Fraction 3 (54 mL of 50% DCM/hexane) contained heptachlor epoxide and dieldrin. Fraction 4 (80 mL of 7% MeOH/DCM) was the aryl sulfone fraction (e.g., MeSO2PCBs). Each fraction was concentrated by rotoevaporation and blown down with nitrogen to a final volume of 100 µL for GC-µECD and GC-MS(ECNI) analysis. An external standard quantification approach was used for PCBs based on peak area of the ECD response. ΣPCB is the sum of concentrations of the 40 individual or coeluting congeners (in order of elution on a DB-5 GC column: PCB 31, 28, 52, 49, 44, 42, 64, 74, 70, 66/95, 60, 101, 99, 97, 110, 151, 149, 118, 146, 153, 105, 141, 179, 138, 158, 129/178, 187, 183, 128, 174, 177, 171/202/156, 200, 172, 180, 170/190, 201, 203, 195, 194, and 206). PCBs are numbered according to the corrected IUPAC numbering scheme as described by Guitart et al. (31). ΣDDTs is the sum of p,p′-DDT, p,p′-DDD, and p,p′-DDE. ΣHCH is the sum of the R, β, and γ HCH isomers. ΣChlordane is the sum of oxychlordane, trans-chlordane, cis-chlordane, trans-nonachlor, cis-nonachlor, and heptachlor epoxide. An internal standard approach was used for OH-PCB (i.e., VOL. 37, NO. 5, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
833
834
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
a Concentrations are in ng/g (wet weight basis). For fish where more than one pool was analyzed (Table 1), listed are the mean and the range (in brackets) of concentrations, and ΣOH-PCB/ ΣPCB, 4-OH-HpCS/OCS, ΣPCB/CB-153, and ΣOH-PCB/CB-153 concentrations ratios. b Structure confirmed by GC-MS (ECNI) in full scan mode or in selected ion monitoring mode (SIM). See Materials and Methods section for the identity of the SIM ions used.
909.04 66.57 8.68 0.08 0.01 112.50 117.84 0.16 0.04 0.13 3.11 8.14 1.05 10.41 (9.56-11.25) 2.07 (1.93-2.21) 0.72 (0.60-0.83) 0.04 (0.03-0.05) 0.02 (
