Biochemical responses to sediment-associated contaminants in brown bullhead (Ameriurus nebulosus) from the Niagara River ecosystem

Ecotoxicology 6, 13–34 (1997)

Biochemical responses to sediment-associated contaminants in brown bullhead (Ameriurus nebulosus) from the Niagara River ecosystem NA N C Y A . E U F E M I A 1 , T R AC Y K . C O L L I E R 2 , J O H N E . S T E I N 2 , DAV I D E . WAT S O N 1 and R I CHARD T. DI GI ULI O 1 3 1

Ecotoxicology Laboratory, School of the Environment, Duke University, Durham, NC 27708-0328, USA Environmental Conservation Division, Northwest Fisheries Science Center, National Marine Fisheries Service, 2725 Montlake Boulevard East, Seattle, WA 98112-2097, USA

2

Brown bullhead (Ameriurus nebulosus) were collected from three sites in the Niagara River ecosystem in June and September of 1991, and sediment samples from these sites were obtained in July 1991. The sites were located in the Buffalo River, the Niagara River adjacent to the Love Canal dump site, and in Black Creek, a Canadian tributary of the Niagara River which served as a reference site. Sediment samples from these sites contained measurable concentrations of various polycyclic aromatic hydrocarbons (PAHs) and chlorinated hydrocarbons (CHs). However, the Buffalo River and Love Canal samples were significantly more contaminated than those from Black Creek. Moreover, Buffalo River samples contained greater PAH concentrations than samples from the Love Canal, while the reverse was observed for CHs. Bile and liver of bullhead were used for the following analyses: fluorescent aromatic compounds in bile, a measure of exposure to PAHs, microsomal cytochrome P450 (CYP) and P450IA (CYP1A) contents and ethoxyresorufin-O-deethylase (EROD) activities, total glutathione (TH-GSH) concentrations, concentrations of 8-oxodeoxyguanosine (8-oxo-dG), and concentrations of hydrophobic DNA adducts (as measured by 32 P-postlabelling). Additionally, a laboratory experiment was performed to examine CYP1A-associated responses in bullhead exposed to the model inducer, -naphthoflavone (BNF). Results from the laboratory induction study were generally consistent with those observed in the field study, but the field study results suggested induction of CYP1A in bullhead from the reference site (Black Creek). For both field collections, fish from the Buffalo River displayed the greatest concentrations of fluorescent compounds in bile and hepatic DNA adducts, whilst fish from the Love Canal site displayed the greatest microsomal CYP1A concentrations and EROD activities. TH-GSH concentrations were significantly greater in Buffalo River fish versus Black Creek only for the June sampling. No statistically significant differences in 8oxo-dG concentrations in bullhead hepatic DNA were observed among the sites at either sampling date. The different patterns in biochemical responses observed were consistent with sediment chemistries, and these results suggest that exposure of feral teleosts to different suites of bioavailable contaminants can be associated with expression of a characteristic array of biochemical responses. Keywords: Ameriurus nebulosus; biomarkers; sediment; genotoxicity; oxidative stress; cytochrome P450.

Introduction Environmental contaminants are known to induce measurable biochemical changes in exposed aquatic organisms. Some of these biochemical changes have been used as 3 To whom correspondence should be addressed. 0963–9292

# 1997 Chapman & Hall

14

Eufemia et al.

biomarkers of exposure to contaminants (Dunn et al., 1987; Bayne et al., 1988; Stein et al., 1992). In these and other studies, biochemical responses complemented chemical data for sediments by addressing issues such as patchy distributions of sedimentassociated contaminants, bioavailability, important sublethal effects such as DNA damage, and interactive effects of complex contaminant assemblages in aquatic ecosystems. The utility of this approach is perhaps most apparent for compounds expected to exert impacts through chronic, sublethal exposures, as opposed to acute effects. In this regard, assessing the effects of sediment-associated compounds related to mutagenesis and carcinogenesis may greatly benefit from the incorporation of appropriate biochemical responses. The selection of appropriate biomarkers is complicated by the interactive nature of the biochemical responses of fish to xenobiotic exposure and the chemical complexity and non-uniform distribution of sediment contamination. Recent studies have found that a variety of chemical and biochemical indices were generally responsive to differences in contaminant exposure among sites in Puget Sound, WA (Stein et al., 1992) and sites along the North American Atlantic coast (Collier et al., 1993; Wirgin et al., 1994). Similar studies in Langesundfjord, Norway (Bayne et al., 1988) also found strong correlations between sediment contamination and biochemical markers in flounder (Platichthys flesus) and mussels (Mytilus edulis). The work done in both Puget Sound and Norway compared biomarkers in sites with similar sediment contaminant compositions, primarily PAHs and chlorinated hydrocarbons through a gradient of contaminant concentrations while the studies of Wirgin et al., (1994) included sites contaminated with PAHs and CHs as well as a site receiving pulp-mill effluent. These studies all provide evidence that examining a group of biochemical indices enhances the ability to detect exposure and sublethal effects in fish populations inhabiting environments contaminated with anthropogenic chemicals. Here, we present data from measurements of monooxygenase-associated variables, fluorescent xenobiotic compounds in bile, indices of genotoxicity, and hepatic glutathione, in fish taken from three sites in the Niagara River ecosystem with different suites of dominant sediment associated contaminants. Two sites were contaminated by PAHs and chlorinated hydrocarbons, but with marked differences in relative amounts of these two classes of contaminants. Previous histopathology studies have indicated high rates of hepatic neoplasia in resident brown bullhead (Ameriurus nebulosus) from both contaminated sites (Black et al., 1980; Black, 1983). Thus, the Niagara River ecosystem afforded an opportunity to test the utility of several biochemical responses in fish from two sites with different mixtures of sedimentassociated genotoxic compounds. Two indices of DNA damage were employed in the present study: 32 P-postlabelling of hydrophobic-DNA adducts and 8-oxo-deoxyguanosine (8-oxo-dG). Using the 32 Ppostlabelling assay, Maccubbin et al. (1990) found that hepatic DNA adducts were associated with sediment PAH concentrations and fluorescent aromatic compounds (FACs) in the bile of brown bullhead from the Buffalo River, and Collier et al. (1993) similarly found that hepatic DNA adducts in oyster toadfish (Opsanus tau) from the Elizabeth River, Virginia were significantly related to concentrations of biliary FACs and sediment PAHs. The measurement of another type of DNA lesion, 8-oxo-dG was employed as an index of oxidative DNA damage. 8-oxo-dG is a nucleoside thought to be produced from the hydroxylation of 29-deoxyguanosine by hydroxyl radical. Whereas

Biochemical responses in Niagara River bullhead

15

the 32 P-postlabelling assay is now a well-accepted method for examining DNA damage, measurement of 8OH-dG is a relatively new procedure for which few published studies with aquatic animals are currently available (Malins et al., 1990; Nishimoto et al., 1991; Kelly et al., 1992). In addition to biological markers of genotoxicity, indices of induction of the cytochrome P450 (CYP) monooxygenase system have been effectively utilized as biomarkers of exposure to PAHs and certain planar chlorinated hydrocarbons. Induction of the CYP1A isozyme, a member of the P450 super gene family, has been identified in several teleost species, including scup (Stenotomus chrysops), winter flounder (Pseudopleuronectes americanus) (Kloepper-Sams et al., 1987), rainbow trout (Onchorynchus mykiss) (Williams and Buhler, 1984) and English sole (Pleuronectes vetulus) (Varanasi et al., 1986), and appears to be homologous among all species studied. To the best of our knowledge, elucidation of CYP isozymes has not been carried out for brown bullhead. Due to the lack of background information on monooxygenase function, a laboratory study measuring CYP content and activity in brown bullhead exposed to the model inducer -napthoflavone (BNF) was also conducted. Catalytic activity assays and total CYP and CYP1A quantitation were conducted in all laboratory and field studies to compare the various measures of monooxygenase induction and to determine if a PAH inducible isozyme similar to the one found in other fish species exists in brown bullhead. The tripeptide reduced glutathione (GSH) plays important detoxification functions in vertebrates, both as a phase II conjugant of electrophiles including halogenated hydrocarbons and epoxide metabolites of PAHs, and as an antioxidant (Larsson et al., 1983). Regarding the former function, the conjugation of GSH to electrophiles is catalyzed by the glutathione-S-transferases (GST). A principal antioxidant function of GSH is its use as a co-factor for glutathione peroxidase (GPx), which catalyzes the reduction of peroxides associated with oxidative stress to their corresponding alcohol (e.g. H2 O2 to H2 O). In this process, GSH is oxidized to glutathione disulfide (GSSG). Thus, compounds that are either suitable GST substrates (either as parent material or metabolite) or enhance the production of reactive oxygen species may perturb the glutathione status of target tissues (Brigelius et al., 1981; Gallagher et al., 1992). Relatively little work has been done with glutathione as an environmental marker, although its utility appears promising (Stein et al., 1992; Di Giulio et al., 1993; Nishimoto et al., 1995). Materials and methods Chemicals 7-ethoxyresorufin (resorufin ethyl ether) was obtained from Molecular Probes Inc. (Junction City, OR). All other chemicals were of the highest grade commercially available and were obtained from Sigma Chemical Co. (St Louis, MO). The 8-oxo-dG standard was a generous gift from Dr R.A. Floyd, Oklahoma Medical Research Foundation, Molecular Toxicology Research Group (Oklahoma City, OK). Laboratory monooxygenase induction study Juvenile brown bullhead (Ameriurus nebulosus) were obtained from Zett’s Fish Hatchery (Drifting, PA) and held under static conditions in 100 litre tanks in 23 8C water for a one

16

Eufemia et al.

week period prior to use. Bullhead were fed Westco (Charlotte, NC) floating catfish food, to satiation, every other day. Water was changed every four days to keep ammonia concentrations ,1 ppm. Fifteen bullhead were injected intraperitoneally (i.p.) with a low dose of BNF (10 mg BNF per kg body weight, 13 mg BNF suspended in 1 ml corn oil), seventeen bullhead were injected i.p. with a high dose of BNF (100 mg BNF per kg body weight, 13 mg BNF dissolved in 1 ml corn oil), five bullhead, serving as carrier controls, were injected i.p. with an equivalent amount of corn oil. Three untreated bullhead were used as absolute controls. At 48 hr, 72 hr, and 7 days (d) after exposure, 5–6 fish per treatment were sacrificed by cervical dislocation. Livers were excised and microsome preparation was carried out on fresh liver tissue. Microsomal samples were assayed for total CYP content, CYP1A levels and EROD activity by the methods described in the following sections. Because controls were not sacrificed at each time point (controls only harvested at 7 days), no statistical comparisons among treatments were made. Sediment chemistry Sediment samples were collected in July, 1991, by Buffalo State College personnel, at the sites of feral fish collection. Samples were mixed with a mortar mixer to insure uniformity and were analysed for selected 2–5 ring PAH’s (Table 1) and CHs (Table 2) according to Sloan et al. (1993). Field collection Adult brown bullhead (72–647 g) were collected from three sites: (1) the Buffalo River (Buffalo, NY) approximately 2 km upstream from the mouth of the river, (2) from the 102nd Street bay area of the Niagara River near the Love Canal–102nd Street dump site (Wheatfield, NY) and (3) in Black Creek (Ontario, Canada) 3–5 km upstream from the river mouth (Fig. 1). All collections were made using a 17-foot Boston whaler equipped with an electroshocking apparatus and hoop nets. The sampling sites were selected in accordance with ongoing monitoring studies by the US Fish and Wildlife Service (coordinated by John Hickey, Cortland, NY). Fish collections were performed in June and September, 1991. In June, fish were kept alive on ice for up to 4 hr during transport to the Buffalo State College Great Lakes Laboratory, where they were sacrificed by cervical dislocation. In September, fish were sacrificed immediately. During both sampling periods, weight, length and external abnormalities were recorded. Livers were excised and washed in ice-cold 0.15 M KCl (pH 7.4), cut into pieces approximately 250 mg each, placed in vials and immersed in liquid nitrogen. Bile was withdrawn from the gall bladder and frozen in liquid nitrogen. All samples were transported in liquid nitrogen to the Duke University Ecotoxicology Laboratory where they were stored at 270 8C until use. Tissue preparation One piece of liver (250 mg) from each fish was thawed, rinsed with 0.15 M KCl buffer, weighed and homogenized (at a total tissue volume of 20% w/v in 0.25 M sucrose per 0.05 M Tris HCl, pH 7.4) and microsomes were prepared according to the method of Erikkson et al. (1978) and stored at 270 8C. A 10 l aliquiot was removed from each microsome sample and diluted to a concentration of 10 g per ml in 50 mM sodium

Biochemical responses in Niagara River bullhead ÿ1

Table 1. Concentrations (ng g Niagara River ecosystem

17

wet weight) of aromatic hydrocarbons in sediment samples from the Site

Compound naphthalene 2-methylnaphthalene 1-methylnaphthalene biphenyl 2,6-dimethlnaphthalene acenaphthylene acenaphthene 2,3,5-trimethylnaphthalene fluorene phenanthrene anthracene 1-methylphenanthrene fluoranthene pyrene benz[a]anthracene chrysene benzo[b]fluroanthene benzo[k]fluoranthene-f benzo[e]pyrene benzo[a]pyrene perylene indeno[1,2,3-cd]pyrene dibenz[a,h]anthracene benzo[ghi]perylene

Black Creek ,4 ,7 ,7 ,5 ,6 ,4 ,6 ,6 ,4 12 ,2 ,3 39 30 12 22 25 17 16 13 170 14 ,1 17

Buffalo River 81 35 31 ,18 ,24 ,14 37 ,23 68 440 180 82 890 740 450 540 430 160 190 350 160 270 64 250

Love Canal 44 30 18 16 11 4 24 ,5 34 240 73 31 470 360 230 250 270 190 210 230 130 200 58 210

bicarbonate. The diluted microsome samples were shipped on dry ice to the Northwest Fisheries Center (Seattle, WA) for measurement of CYP1A isozyme content, described below. Microsomal protein concentrations were determined using the bicinchoninic acid protein assay (Sigma procedure No. TPRO-562). Monooxygenase assays The reduced carbon monoxide binding spectra of CYP were measured by the method of Matsubara et al. (1976). After a baseline spectrum was recorded, the sample cuvette was reduced with a few mg of sodium dithionite (Na2 SO4 ), Carbon monoxide gas was bubbled through both cuvettes and the difference spectra was recorded. Each cuvette contained approximately 1 mg microsomal protein per mL. Total CYP content was calculated by subtracting the absorbance at 490 nm from the absorbance at 450 nm and dividing by an extinction coefficient of 106 cmÿ1 mMÿ1 . Microsomal samples were diluted in 50 mM sodium bicarbonate for measurement of CYP1A content by the semi-quantitative enzyme-linked (ELISA) assay, as described by Goksøyr (1991). Rabbit anti-cod P450c was used as the primary antibody. Ethoxyresorufin O-deethylase (EROD) activity was measured by a modification of the

18

Eufemia et al. ÿ1

Table 2. Concentrations (ng g the Niagara River ecosystem

wet weight) of chlorinated hydrocarbons in sediment samples from Site

Compound

Black Creek

hexachlorobenzene lindane (gamma-BHC) heptachlor aldrin heptachlorepoxide alpha-chlordane trans-nonachlor dieldrin mirex o.p9-DDE p,p9-DDE o,p9-DDD p,p9-DDD o,p9-DDT p,p9-DDT trichlorobiphenyls tetrachlorobiphenyls pentachlorobiphenyls hexachlorobiphenyls heptachlorobiphenyls octachlorobiphenyls nonachlorobiphenyls decachlorobiphenyl

,0.1 0.3 ,0.1 ,0.1 ,0.1 ,0.1 ,0.1 ,0.1 ,0.1 ,0.2 1 0.5 0.6 ,0.2 ,0.2 2 6 3 8 2 0.4 0.2 0.5

Buffalo River 0.9 ,0.6 ,0.7 ,0.7 ,0.6 ,0.6 0.6 ,0.6 ,0.7 ,1 4 3 3 1 ,0.9 11 84 35 87 28 12 0.6 1

Love Canal 37 24 4 11 2 ,0.5 8 1 130 2 3 5 12 9 4 360 320 220 120 91 40 3 6

method of Burke and Mayer (1974). The reaction mixture consisted of 2.0 ml Tris-HCl buffer, 25 l microsomes and 5 l of 400 M ethoxyresrufin (in methanol) in a 3 ml cuvette. After a 2 min pre-incubation, 7.5 l of 50 mM NADPH was added to the reaction mixture and the change in fluorescence units per min was recorded for 3 min on a Perkin-Elmer LS-3 fluoresence spectrophotometer (Norwalk, CT) at a wavelength pair of 530/586 nm. The rate obtained was converted to pmols resorufin formed per mg protein per minute by using the equation for the linear regression line produced from resorufin standards; an extinction coefficient of 73 cmÿ1 mMÿ1 was used. Analysis of bile Samples (n = 5 per month) of bile (6 l aliquots) were analysed for fluorescent aromatic compounds according to the HPLC method of Krahn et al. (1984) with slight modification. Separations were performed on a Perkin-Elmer Series 400 liquid chromatograph equipped with a 3 cm C18 guard column and a reversed phase C18 analytical column. An Upchurch Scientific (Oak Harbor, WA) A-318 precolumn filter with replaceable frit was placed in-line and used with all samples and standards. The mobile phase flow rate was 1.0 mL per min. A linear gradient of 100% aqueous (5 l acetic acid per litre of water) to 100% methanol was performed in 10 min, followed by

Biochemical responses in Niagara River bullhead

19

LAKE ONTARIO

N LOWER RIVER

RIVER FLOW LOVE CANAL 102ND ST. DUMP NIAGARA FALLS

GRAND ISLAND

BLACK CREEK

Toronto LAKE ONTARIO

BUFFALO RIVER

Buffalo LAKE ERIE

LAKE ERIE STUDY AREA

Fig. 1. Sampling sites for brown bullhead and sediments in the Niagara River system.

20 min of 100% methanol, a 3 min linear gradient return to aqueous conditions and 10 min re-equilibration prior to the next injection. The eluant was monitored with a Perkin Elmer fluorescence spectrophotometer at wavelength pairs for benzo[a]pyrene (BaP – 380/430 nm excitation/emission) or phenanthrene (256/380 nm excitation/ emission) (Krahn et al., 1984). Chromatograms were recorded and the peaks integrated and converted to PAH equivalents on the EZChrom Chromatography Data System, version 4.0 (Scientific Software, Inc. San Ramon, CA).

20

Eufemia et al.

Analysis of DNA modifications Portions of whole liver from 5–7 randomly selected samples were analysed for hydrophobic DNA adducts using the nuclease P1 version of the 32 P-postlabelling method. Hepatic DNA was isolated by the method of Reddy and Randerath (1987), and the 32 Ppostlabelling assay was conducted essentially according to Gupta and Randerath (1988). Salmon sperm DNA from Atlantic salmon (Salmo salar) was used as a negative control in each set of analyses. Briefly, DNA (10 g) was enzymatically hydrolyzed using micrococcal endonuclease and spleen phosphodiesterase. Nuclease P1 was used to degrade normal nucleotides, thereby enriching the mixture in adducted 39-monophosphates (Reddy and Randerath, 1986). These samples were then postlabelled using [ª32 P]ATP (specific activity 2400 Ci/mmol) synthesized according to Gupta and Randerath (1988). The 32 P-labelled adducts were chromatographed on polyethyleneimine-cellulose thinlayer chromatography (TLC) sheets prepared in the laboratory (Gupta and Randerath, 1988). The solvent systems used in the multidirectional chromatography were as follows: D1 – 1.0 M sodium phosphate, pH 6.0; D2 was omitted; D3 – 7.65 M urea and 4.32 M lithium formate, pH 3.5; and D4 – 7.65 M urea, 1.44 M lithium chloride and 0.45 M Tris, pH 8.0. Elution of the chromatograms in D5 was not done. The 32 Plabelled DNA-adducts on the chromatograms were located and quantitated using storage phosphor imaging technology (Reichert et al., 1992). Total nucleotides were determined by one-dimensional TLC of 59-labelled nucleotides using 0.24 M ammonium sulfate in 8 mM sodium phosphate, pH 7.4, as the solvent, followed by quantitation of the deoxyguanosine 39,59-bisphosphate spot which was assumed to represent 21% of the total nucleotides. Another portion of liver (approximately 500 mg) was used to measure concentrations of 8-oxo-dG in DNA. Hepatic DNA was isolated by the method reported by Gupta (1984). A 400 g aliquot from each isolate was incubated with 20 g Nuclease P1 for 45 min at 37 8C; 1.3 units of alkaline phosphatase were then added, and the incubation continued for an additional 1.5 hr. Levels of 8-oxo-dG were determined by a reversed phase HPLC method, adapted from Park et al. (1989). The mobile phase consisted of 5% methanol in 50 mM potassium phosphate at pH 5.0. Separation of the nucleosides was achieved using a 5 C18 reversed-phase HPLC column (250 3 4.6 mm) containing 5 M particles. Detection of 29dG was achieved using a Perkin-Elmer LC-95 UV/Vis Spectrophotometer set at 260 nm. 8-oxo-dG was detected using an EG & G Princeton Applied Research (Princeton, NJ) Model 400 electrochemical detector equipped with a glassy carbon electrode, operated at an applied voltage of 600 mV and adjusted for a full-scale output of 2 nA. Analysis of total glutathione Total glutathione concentrations (TH-GSH) were measured according to a method adapted from Griffith (1980). Briefly, pieces of whole liver were homogenized (total tissue volume of 20% w/v of 5% 5-sulfosalicylic acid) and homogenates were centrifuged for 5 min at 5000 g. A 100 l portion of the supernatant was combined with 900 l deionized water and kept on ice. In a 1 ml cuvette, 750 l 0.24 mM NADPH in 125 mM Na-phosphate/6.3 mM Na-EDTA, 100 l 6 mM DTNB and 100 l diluted supernatant were allowed to equilibrate for 1–2 min. The reaction was initiated by the addition of

Biochemical responses in Niagara River bullhead

21

50 l of 20 units per ml glutathione reductase and the change in absorbance was recorded at 412 nm on a Shimadzu (Kyoto, Japan) UV 260 spectrophotometer. A GSH standard curve was determined with each set of samples analysed. Statistical analysis Data were analysed using the SAS statistical analysis package for personal computers, version 6.03 (SAS Institute, 1986). Data were first tested for normality using the ShapiroWilkes Test ( p , 0.05). Since data generally appeared non-normally distributed, nonparametric statistics were used for further analyses. Relationships among collection site and biochemical responses were studied using the NPAR1WAY procedure, a nonparametric one-way analysis of variance based on rank with multiple comparison (SAS Institute, 1986). Comparisons among the three sites included intrasite, sex and season comparisons, using the Mann-Whitney U test. The Spearman Rank Correlation Coefficient was determined to examine correlations among various indices. All statistical analyses were performed at a significance level of p , 0.05. Results Sediment chemical analysis Analysis of sediments revealed high concentrations of PAHs at Buffalo River and Love Canal, high concentrations of PCBs at Love Canal and lower concentrations of PAHs and PCBs at Black Creek (Tables 1 and 2). The sums of measured PAH concentrations (wet weight) were approximately 5400 ng per g at Buffalo River, 3300 ng per g at Love Canal and 390 ng per g at Black Creek. The sums of measured chlorinated hydrocarbon concentrations (wet weight), which were dominated by PCBs at all sites, were approximately 1400 ng per g at Love Canal, 270 ng per g at Buffalo River aned 25 ng per g at Black Creek. In addition, sediments from Love Canal showed elevated concentrations of several other chlorinated pesticides, in particular, hexachlorobenzene, lindane, aldrin and mirex when compared to the concentrations in sediments from either Buffalo River or Black Creek. Seasonal and sex differences No significant differences for any assay were observed between sexes at the three study sites except EROD activities measured in June. Since this difference was an apparent exception, sexes were pooled for all further analyses. Seasonal differences between samples collected in June and September followed no apparent trend, but were significant for some assays. Therefore, sampling periods were kept separate for statistical analyses. Monooxygenase assays – laboratory time course study The laboratory time course exposure (Table 3) provided a basis of comparison for the field data. Maximum median EROD activity for bullhead injected with 100 mg BNF per kg body weight was 26 pmol per min per mg protein and 55 pmol per min per mg protein for fish injected with 10 mg BNF per kg body weight. Median total CYP concentrations over the 7 day time course for bullhead exposed to a high dose of BNF were 0.19–0.54 nmol per mg protein, and for low dose BNF exposure were 0.13– 0.27 nmol per mg protein. A similar trend was seen in CYP1A content for brown bullhead exposed to a high dose of BNF (median range over time course 580–650 mOD

22

Eufemia et al.

Table 3. Monooxygenase measures for laboratory time course study, brown bullhead exposed to 10 mg kgÿ1 and 100 mg kgÿ1 BNF (median and range)

Exposure group Untreated control

Corn oil control

10 mg/kg 48 hours

96 hours

168 hours

100 mg/kg 48 hours

96 hours

168 hours

Cytochrome P450 (nmol mgÿ1 protein)

Cytochrome P450IA (mOD units/mg somal protein)

EROD activity (pmol per mg protein per min)

0.3 3

N.D. 3

N.D. 3

n=1 0.04 0.02–0.08 n=3

n=3 0 0–582 n=3

n=3 N.D. 3

0.35 0.32–0.39 n=3 0.2 0.04–0.5 n=3 0.55 0.51–0.58 n=3

580 456–943 n=5 330 0–462 n=5 500 169–777 n=5

54 34–128 n=5 12 0–28 n=5 18 10–30 n=5

0.2 3

860 0–2126 n=4 580 502–863 n=4 650 487–1163 n=4

4.7 0–8 n=4 22 12–22 n=4 26 0–34 n=4

n=1 0.2 0.2–0.3 n=3 0.1 0.05–0.2 n=3

n=3

N.D. not detectable 3 no maximum and minimum values reported due to small sample size or all values below detection limits.

per mg microsomal protein) and a low dose of BNF (median range over time course 330–580 mOD per mg microsomal protein). Monooxygenase assays – field study Median, maximum and minimum values for all hepatic monooxygenase analyses performed on field-derived fish are reported in Table 4. Median hepatic EROD activities (pmol/mg protein/min) were significantly higher at Love Canal than at Buffalo River and Black Creek for June and September samples (3–6 fold and 3–5 fold, respectively), but not significantly different between Buffalo River and Black Creek for either sampling period. The EROD activities for Love Canal (median = 24.6 pmol per mg protein per min for June and 25.5 pmol per mg protein per min for September) were similar while Buffalo River values (median = 7.7 pmol per mg protein per min for June and 9.2 pmol per mg protein per min for September) were slightly lower than values for bullhead exposed to both high (100 mg BNF per kg body weight) and low (10 mg BNF per kg

Biochemical responses in Niagara River bullhead

23

Table 4. Hepatic monoxygenase measures in brown bullhead from the Niagara River ecosystem (median and range)

Location June Buffalo River

Love Canal

Black Creek

September Buffalo River

Love Canal

Black Creek

Cytochrome P450 (nmol mgÿ1 protein)

Cytochrome P450IA (mOD units per mg somal protein)

EROD (pmol per mg protein per min)

NADPH-cytochrome P450 reductase3 (nmol per mg minÿ1 )

0.3 a,b 0.1–0.8 n = 10 0.4 a 0.2–0.8 n = 10 0.2 b 0.12–0.8 n=9

250 a,b 145–373 n = 17 290 a 168–752 n = 27 190 b 86–329 n = 13

7.8 b 1.8–32.5 n = 30 25 a 2.5–249 n = 30 3.8 b 0–21 n = 19

29 6.1–42 n = 30 31 12.5–56 n = 30 31 0–61 n = 19

0.5 a,b 0.1–0.9 n = 13 0.6 a 0.1–0.8 n = 13 0.32 b 0.09–0.72 n = 13

110 a 38–864 n = 14 240 a 36–435 n = 13 85 b 0–208.2 n = 11

9.2 b 0–30 n = 15 25 a 4.5–85 n = 14 4.2 b 0–9.81 n = 14

N.D.

N.D.

N.D.

a,b,c = medians with the same letter are not significantly different (p , 0.05) 3 NADPH-cytochrome P450 reductase activity measured only in June samples. N.D. not determined.

body weight) doses of BNF. EROD activity in brown bullhead from the reference site, Black Creek (median = 3.76 pmol per min per mg for June and 4.19 pmol per min per mg for September), was higher than activity in unexposed and corn oil treated control bullhead (median activity , 1 pmol per min per mg protein for both groups). Median total CYP concentrations were significantly greater at Love Canal (0.43 and 0.63 nmol per mg protein) than at Black Creek (0.19 and 0.33 nmol per mg protein) for June and September samples, respectively; Buffalo River P450 concentrations (0.27 and 0.47 nmol per mg protein) fell between concentrations for the other two sites for both sampling periods. Median values for all sites were in the same range as total CYP content for bullhead exposed to high or low doses of BNF. Despite sampling and storage precautions, spectral analysis of CYP suggested that some sample degradation had occurred, as evidenced by the peaks at 420 nm observed in most CO-difference spectra of reduced bullhead microsomes from all three sites. Median CYP1A concentrations, as measured by an ELISA, followed a similar trend to EROD activity and total CYP concentrations, with June and September Love Canal median values (290 and 240 mOD per mg microsomal protein, respectively) being significantly higher than those for Black Creek (190 and 85 mOD per mg microsomal

24

Eufemia et al.

protein, respectively). Buffalo River concentrations (250 and 110 mOD per mg microsomal protein for June and September, respectively) were also significantly higher then Black Creek values for June samples, but not significantly higher in September samples. Median CYP1A values were lower in field collected brown bullhead than for high and low dose BNF-exposed bullhead. Fluorescent aromatic compounds in bile The chromatograms of bile from brown bullhead caught in contaminated sites showed a complex mixture of fluorescent aromatic compounds (FACs) in bile at wavelength pairs for phenanthrene and benzo(a)pyrene (BaP). Biliary FACs measured at phenanthrene wavelengths were significantly greater (10–40 fold) in Buffalo River bullhead than Love Canal or Black Creek bullhead for June and September collected samples (Table 5). Phenanthrene equivalents at Love Canal were significantly greater than Black Creek in June but not in September collected fish. Levels of biliary FACs of BaP equivalents were significantly higher at Buffalo River (9–12 fold) than Love Canal or Black Creek in June and September, but no differences existed between median biliary BaP FACs for Love Canal and Black Creek. Median phenanthrene equivalents (ng gÿ1 ) were significantly higher (6–20 fold) than median BaP equivalents (ng gÿ1 ) at all three sites. Hepatic DNA adducts Median, maximum and minimum hepatic DNA adduct concentrations (as detected by 32 P-postlabelling) are reported in Table 5 and representative chromatograms from the three sites are displayed in Fig. 2. Median DNA adduct levels for the June sampling were signficantly higher at Buffalo River (94 nmol adducts per mol nucleotides), than levels in fish from Love Canal (61 nmol adducts per mol nucleotides) and Black Creek (20 nmol adducts per mol nucleotides) and levels in fish from Love Canal were also significantly greater than in fish from Black Creek. A similar trend was seen in September samples. Median DNA adduct levels for September samples were significantly higher at Buffalo River (64 nmol per ml nucleotides) than Love Canal (11.5 nmol adducts per mol nucleotides) or Black Creek (4.5 nmol adducts per mol nucleotides). No significant differences were detected between Love Canal and Black Creek for either sampling period. No statistically significant differences in 8-oxo-dG concentrations in brown bullhead hepatic DNA were observed between any of the sites at either sampling date (Table 5). However, the rank order of median values of these sites is the same in June and in September, with the reference site having the lowest values at each sampling date. Total hepatic glutathione Median, maximum and minimum TH-GSH values are reported in Table 5. Brown bullhead collected in June had median TH-GSH concentrations that were significantly higher in Buffalo River (1.57 umol gÿ1 tissue) than Black Creek (1.14 umol gÿ1 tissue), with Love Canal (1.27 umol gÿ1 tissue) having intermediate levels, not significantly different from the others. No significant differences were observed among bullhead from all three sites in September.

1.1 0.5–1.8 n = 13 1.1 0.4–1.7 n = 11 1 0.6–2 n = 13

1.6 a 1.2–1.9 n = 12 1.3 a,b 0.8–2.6 n = 12 1.1 b 1–1.5 n=6 8.2 a 4.5–26 n = 12 11 a 6.8–45 n=9 7.2 a 0.5–19 n=9

5.6 a 3.1–20 n = 25 5.7 a 3.2–29 n = 27 3.8 a 1.5–11 n = 15

8OH-dG [8OH-dG per 29dG (e25)]

65 a 37–250 n=7 11 b 5.1–21 n=6 4.5 c 2–8 n=6

94 a 78–214 n=5 61 b 22–87 n=7 20 c 15–31 n=6

DNA adducts (nmol adducts per mol DNA)

a,b,c = medians with the same letter are not significantly different between sites (p , 0.05)

Black Creek

Love Canal

September Buffalo River

Black Creek

Love Canal

June Buffalo River

Location 1ÿ

TH-GSH (umol g tissue)

1500 a 1262–8019 n=5 110 b 8–755 n=5 130 b 52–440 n=5

6400 a 2208–10855 n=5 340 b 198–791 n=5 150 c 70–184 n=5

Phenanthrene equivalents (ng PHE eq g 1ÿ

bile)

190 a 110–2127 n=5 23 b 8–69 n=5 20 b 10–546 n=5

340 a 283–1591 n=5 31 b 19–90 n=5 19 b 10–240 n=5

Benzo(a)pyrene equivalents (ng BaP eq g bile)

Table 5. Hepatic total glutathione, 8OH-dG, DNA adducts and fluorescent aromatic metabolites in brown bullhead collected from the Niagara River ecosystem (median and range)

1ÿ

26

Eufemia et al.

Fig. 2. Autoradiograms of chromatograms of 32 P-postlabelled hepatic DNA digests from brown bullhead from (A) Buffalo River, (B) Love Canal, (C) Black Creek, and (D) 7R, 8S, 9S, 10R-(N2 deoxyguanosyl-39-phosphate)-7,8,9,10-tetrahydrobenzo[a]pyrene (BaPDE-dG-39p) standard. The autoradiogram for the BaPDE-dG-39p standard is shown to indicate the chromatographic characteristics of a PAH-DNA adduct.

Discussion Regional differences in sediment contaminant profiles exist throughout the Niagara River system (NRTC, 1984). Sediments from Buffalo River contain high concentrations of 2–5 ring PAHs related to petroleum and combustion industries. Sediments from the Love Canal–102nd Street bay sampling site exhibited high concentrations of PCBs, dioxins and

Biochemical responses in Niagara River bullhead

27

furans and pesticides, as well as PAHs. Black Creek was originally selected as a reference site because it was thought to be a relatively uncontaminated creek; however sediment analysis showed it to have PAH concentrations higher than ideal for a reference site. The differences among the sampling sites in terms of predominant sediment contaminants provided the opportunity to compare the biochemical responses of brown bullhead to different mixtures of xenobiotic compounds in a natural setting. The observed effects appear consistent with sediment chemistry data from the sampling sites. Chlorinated hydrocarbons such as TCDD and PCBs are potent CYP1A inducers (Stegeman and Lech, 1991) and fish from the Love Canal, a site with high concentrations of these compounds, exhibited marked CYP1A induction. In contrast, Buffalo River is contaminated primarily with PAHs, which are metabolized by teleosts to reactive metabolites that covalently bind to DNA (Varanasi et al., 1989b). Brown bullhead from this site had higher levels of DNA adducts than bullhead from the other two sites. Laboratory studies comparing metabolic responses from exposure to PAHs and certain chlorinated hydrocarbons (e.g. coplanar PCBs or TCDD) support the results of the Niagara River field study (Addison et al., 1978; Vodicnik et al., 1981; Pesonen et al., 1992). In these laboratory studies, PCBs and TCDD produced either higher levels or more persistent induction of the CYP1A system than did BNF, a model PAH. Addison et al. (1978) fed brook trout (Salvelinus fontinalis) several doses of either Arochlor 1254 or 3-MC, also a PAH, and found both compounds to be strong MFO inducers; however, EROD activity was 2–3 fold higher in trout fed Arochlor 1254. Analysis of fluorescent aromatic metabolites (FACs) in the bile showed brown bullhead from the Buffalo River were exposed to much higher concentrations of PAHs than were bullhead from Love Canal. In addition, a strong association was shown between bile FACs and hepatic DNA adducts, while little association existed between bile FAC concentrations and levels of CYP induction. Thus, overall, these observations imply that PAHs were primarily responsible for the formation of hepatic DNA adducts in brown bullhead from the Buffalo River while other inducers were the apparent causal agents for the strong CYP1A induction seen in bullhead from Love Canal. Hepatic cytochrome P450 Values for hepatic EROD activity, immunoquantitated CYP1A and total CYP content were consistently higher in fish from the contaminated sites, Love Canal and Buffalo River, than those observed for fish from the reference site, Black Creek. Similar observations have been made by other investigators examining similarly contaminated sites (Klotz et al., 1983; Elskus and Stegeman, 1989; Collier et al., 1995). In the present study, median EROD activities from Love Canal fish were 6–8 fold higher than that of Black Creek fish. Similarly, median total CYP and CYP1A concentrations were 2–3 fold greater in Love Canal fish than in fish from Black Creek. Differences in monooxygenase induction between fish populations from Buffalo River and Black Creek were also observed, but were of much lesser magnitude; EROD activity, total CYP and CYP1A concentrations were 1.5–2 fold greater at Buffalo River than Black Creek. The smaller differenes in CYP1A measures between Buffalo River and Black Creek may be attributed in part to elevated contaminant concentration at the Black Creek reference site. Chemical analyses of sediment from Black Creek revealed a total PAH

28

Eufemia et al.

concentration of 390 ng gÿ1 (wet weight), well above concentrations typical in sediments from areas considered uncontaminated reference sites. Contaminant concentrations at Black Creek were apparently great enough to induce CYP content and activity into a range similar to the monooxygenase induction caused by Buffalo River contaminants, both of which were 3–10 fold higher than CYP content and activity in untreated and corn oil exposed control bullhead from the laboratory induction study. CYP1A induction in environmental samples is typically measured by immunoquantitation or catalytic activity (e.g. EROD activity). In the present study both methods were used, and we found that EROD activities and CYP1A levels tracked each other closely among sites and sampling periods. However, correlations between the two measures in individual fish were weak. EROD activities showed greater distinctions among sites than CYP1A quantitation, possibly providing a more sensitive measure of monooxygenase induction. These results are in contrast to several studies which have found analysis of CYP1A concentrations, typically quantitated with specific monoclonal antibodies, to be the more sensitive measure of induction (Stegeman et al., 1987; Kloepper-Sams and Stegeman, 1989). Other researchers have found immunoquantitation of CYP1A to have similar or even reduced sensitivity compared to catalytic assay (Varanasi et al., 1986; Goksøyr et al., 1987; Elskus and Stegeman, 1989; Collier et al., 1992, 1995). Spectral analysis revealed that in certain cases degradation of CYP did occur in samples from all three sites, potentially diminishing EROD activity. This was diagnosed by the characteristic shift from 450 to 420 nm in the absorption spectrum maximum of carbon monoxide difference spectra of CYP. Fo¨rlin and Andersson (1985) observed a time dependent decrease in P450 activity in rainbow trout livers stored in liquid nitrogen. Because of limited facilities at the field sites, liver samples in the present study were stored for up to 7 days in liquid nitrogen before hepatic microsomes were prepared. As a result, it is possible that EROD activity may have been decreased. While the results suggest a greater sensitivity with the catalytic assay, the potential loss of EROD activity, particularly with field samples, arguably supports the use of isozyme quantitation as a biomarker for contaminant exposure, since it is not as readily effected by sample degradation. Thus, it appears that both assays should be used to take advantage of the possibly greater sensitivity provided by the catalytic assays and the ability to detect enzyme deactivation provided by the immunoquantitation of CYP1A protein (Collier et al., 1995). The values for hepatic EROD activities and CYP1A concentrations observed in brown bullhead from both the laboratory and field studies are relatively low in comparison to values reported for other teleost fish, including Ictaluridae. For example, a recent study from this laboratory (Hasspieler et al., 1994) compared hepatic EROD activities in control and BNF-induced in brown bullhead with identically-treated channel catfish (Ictalurus punctatus). Mean EROD activities in control and BNF-induced bullhead were approximately 3 and 17 pmol mgÿ1 protein per min, while the analogous activities in channel catfish were 28 and 268 pmol mgÿ1 protein per min. Watson (1995) also observed a similar difference in EROD activities in these species that was consistent with differences in CYP1A concentrations. Fluorescent aromatic compounds in bile Buffalo River brown bullhead had significantly higher levels of biliary FACs than bullhead from either Love Canal or Black Creek. Sediment chemistry of the three sites

Biochemical responses in Niagara River bullhead

29

shows a similar trend; PAH concentrations in sediments are 2–200 times greater in Buffalo River than Love Canal or Black Creek. Maccubbin et al. (1988) reported an association between FACs in bile of brown bullhead and sediment AH concentrations from the Niagara River region. A similar association has been observed in oyster toadfish from the PAH-contaminated Elizabeth River, Virginia (Collier et al., 1993). The present study supports the findings of both pieces of research. The biliary FACs data appear to give information comparable to reports from other field studies (Krahn et al., 1986; Maccubbin et al., 1988; Johnston and Baumann, 1989). Correlations of bile FAC concentrations with other indicators of monooxygenase induction and DNA adduct formation on an individual fish basis were not warranted due to sample size, although general trends between measures were detected. The bile metabolite assay does not identify and quantitate individual metabolites, but rather measures all fluorescent aromatic compounds that are not endogenous to the species. Therefore, this method may be measuring a different suite of aromatic compounds in fish from different sites. The qualitative nature of the method, with respect to identification of individual compounds, requires that the results be used with caution in assessing correlations (Krahn et al., 1987). Hepatic DNA alterations Although a link between sediment contaminant exposure and neoplasia has been reported in several fish species, the steps in the process of neoplasm development have yet to be elucidated fully (Myers et al., 1991). Biochemical indices detecting early steps in neoplasia may provide information on the etiology of cancer and further define the association between sediment contaminant exposure and tumor formation (Varanasi et al., 1989a). Of the two biological markers of genotoxicity examined, the 32 P-postlabelling assay demonstrated DNA-adduct formation, while no consistent increases in 8-oxo-dG levels were evident in fish from Love Canal and the Buffalo River. However, a previous study (Nishimoto et al., 1991) indicated that this lesion may be more transient in wild fish than DNA adducts measured by 32 P-postlabelling. DNA damage measured by the latter technique is persistent in fish exposed to PAHs (Stein et al., 1993). Elevated concentrations of 8-oxo-dG in fish hepatic tissue have been reported for acute (Nishimoto et al., 1991) and chronic (Kelly et al., 1992) laboratory exposures to pure compounds, and in fish found in contaminated marine ecosystems (Malins et al., 1990; Malins et al., 1990; Malins and Haimanot, 1991). Those publications report statistically significant increases in hepatic 8-oxo-dG concentrations, with absolute values ranging from 3–10-fold greater in exposed compared to control fish. Although there were no significant differences between any sites in the present study, rankings of sites in this study by absolute concentration of hepatic 8-oxo-dG were consistent in both June and September, as follows: Love Canal . Buffalo River . Black Creek. It is possible that the lack of significant differences is due to a species-specific variable, such as rate of 8-oxo-dG repair. The study of Nishimoto et al. (1991) showed that 8oxo-dG is relatively rapidly repaired in English sole; the rate of repair in brown bullhead is unknown. Another limitation in the 8-oxo-dG method is the difficulty at present in determining which xenobiotics in the environment are effectively inducing the damage observed with the assay. Regardless, the measurement of 8-oxo-dG as an indicator of biochemical injury has been only relatively recently examined, and further

30

Eufemia et al.

studies are warranted before its utility as a biomarker of exposure to, and toxicity from, contaminants can be assessed. The level of hepatic DNA adducts determined by 32 P-postlabelling showed significantly higher levels of adduct formation at Buffalo River than at Love Canal and Black Creek for June and September. At present, it is difficult to relate levels of DNA adducts to a single group of environmental contaminants (Varanasi et al., 1989a). However, previous studies (Stein et al., 1993, 1994) show similarities in chromatographic characteristics of adducts in wild fish from contaminated sites and fish exposed in the laboratory to specific hydrophobic PAHs, such as BaP, chrysene and dibenz(a,h)anthracene, as well as in fish exposed to organic solvent extracts of sediments from PAH contaminated sites. Moreover, the xenobiotic compounds most prevalent in sediments and which form adducts with DNA bases are the PAHs, and the greatest levels of DNA adducts are seen at sites most heavily contaminated with these compounds (Varanasi et al., 1987; Stein et al., 1992; Collier et al., 1993; Wirgin et al., 1994). These findings suggest that anthropogenic PAHs are forming adducts with DNA of fish from contaminated sites (Stein et al., 1989). In the present study, comparison of total DNA adduct levels in brown bullhead to sediment concentrations of high molecular weight PAHs showed a general concordance between the two measures. Thus, these findings together with the findings from previous studies with other fish species support an association between DNA adducts detected by 32 P-postlabelling in brown bullheads and exposure to polycyclic aromatic compounds, including PAHs. The levels of hepatic DNA adducts followed a similar trend as biliary FACs among the three study sites. Studies of DNA adduct formation of Buffalo River brown bullhead by Dunn et al. (1987) reported adduct levels, as measured by 32 P-postlabelling, similar to the levels observed in the present study. Buffalo River, the site with the highest concentrations of sediment PAHs, was also the site where brown bullhead showed the highest concentrations of FACs in bile and levels of hepatic DNA adducts. The relationship among sediment PAH concentrations, biliary FACs, and DNA adducts lends additional support to the suggestion that DNA adduct formation, as measured by 32 Ppostlabelling, is linked to PAH exposure. Total hepatic glutathione TH-GSH concentrations were not significantly different between contaminated and reference sites, with the exception of June collected Buffalo River bullhead having higher concentrations of TH-GSH than Black Creek bullhead. Recent sudies have shown THGSH concentrations to be elevated in fish exposed to contaminated sediments (Stein et al., 1992; Di Giulio et al., 1993) or to extracts of contaminated sediments (Nishimoto et al., 1995). Additional research is required to elucidate chemicals that enhance TH-GSH concentrations in teleosts, as well as the utility of this response as a biomarker. Summary In summary, the ability to reach definitive conclusions about biochemical responses to contaminants is especially difficult with field samples because of the complex mixtures of contaminants found at polluted sites. The present study applied several commonly used biomarkers as well as some novel approaches to environmental monitoring. These indices showed strong associations between specific contaminant compounds and biochemical effects in exposed organisms, and suggest that the complexity of field

Biochemical responses in Niagara River bullhead

31

studies necessitates the use of several measures of contaminant exposure and effects to differentiate site-specific alterations. Integrating biochemical and physiological biomarkers with other approaches to assessing environmental impact of chemical contaminant, such as sediment chemical analyses toxicity bioassays and benthic community analyses, would appear to significantly enhance monitoring capabilities. Acknowledgements We thank Dr John Hickey, Chuck Merkel, Chris Zogby and Betsy Kozuchowski for their assistance in sample collecting and processing, Tom Burns and Margaret Kroen for conducting 8-oxo-dG analyses, William L. Reichert and Barbara French for conducting the 32 P-postlabelling assays, Bernadita F. Anulacion for immunoquantitation of CYP1A, and Rebecca Yang for assisting with the laboratory induction study. This work was supported by grants from the US EPA Office of Exploratory Research (R-817301-01-1), and a joint program between the US Geological Survey and the University of North Carolina Water Resources Research Institute (No. 20158). References Addison, R.F., Zinck, M.E., Willis, D.E. (1978) Induction of hepatic mixed-function oxidase (MFO) enzymes in trout (Salvelinus fontinalis) by feeding Arochlor 1254 or 3-methylcholanthrene. Comp. Biochem. Physiol. 61, 323–5. Bayne, B.L., Addison, R.F., Capuzzo, J.M., Charke, K.R., Gray, J.S., Moore, M.N. and Wanwick, R.M. (1988) An overview of the GEEP workshop. Man. Ecol. Prog. Ser. 46, 235–243. Black, J.J. (1983) Field and laboratory studies of environmental carcinogenesis in Niagara River fish. J. Great Lakes Res. 9, 326–34. Black, J.J., Holmes, M., Dymerski, P.R. and Zapisek, W.F. (1980) Fish tumor pathology and aromatic hydrocarbon pollution in a Great Lake estuary. Environ. Sci. Res. 16, 559–65. Brigelius, R.A., Hashem, A. and Lengfelder, E. (1981) Paraquat-induced alterations of phospholipids and GSSG release in the isolated perfused rat liver, and the effect of SOD-active copper complexes. Biochem. Pharmacol. 30, 349–54. Burke, D.M. and Mayer, R.T. (1974) Ethoxyresorufin: direct fluorometric assay of a microsomal Odealkylation which is preferentially inducible by 3-methylcholanthrene. Drug Metab. Dispos. 2, 583–8. Collier, T.K., Eberhart, B.L., Connor, S.D., Goksøyr, A. and Varanasi, U. (1992) Using cytochrome P450 to monitor the aquatic environment: initial results from regional and national surveys. Mar. Environ. Res. 34, 195–9. Collier, T.K., Stein, J.E., Goksøyr, A., Myers, M.S., Gooch, J.W., Huggett, R.J. and Varanasi, U. (1993) Biomarkers of PAH exposure in oyster toadfish (Opsanis tau) from the Elizabeth River, Virginia. Environ. Sci. 2, 161–77. Collier, T.K., Anulacion, B.F., Stein, J.E., Goksøyr, A. and Varanasi, U. (1995) A field evaluation of cytochrome P4501A as a biomarker of contaminant exposure in three species of flatfish. Environ. Toxicol. Chem. 14, 143–52. Di Giulio, R.T., Habig, C. and Gallagher, E. (1993) Effects of Black Rock Harbor sediments on indices of biotransformation, oxidative stress and DNA integrity in channel catfish. Aquatic Tox. 26, 1–22. Dunn, B.P., Black, J.J. and Maccubbin, A. (1987) 32 P-postlabelling analysis of aromatic DNA adducts in fish from polluted areas. Cancer Res. 47, 6543–8. Elskus, A.A. and Stegeman, J.J. (1989) Induced cytochrome P450 in Fundulus heteroclitus associated

32

Eufemia et al.

with environmental contamination by polychlorinated biphenyls and polycyclic aromatic hydrocarbons. Mar. Environ. Res. 27, 31–50. Erikkson, L.C., DePierre, J.W., Pallner, G. (1978) Preparation and properties of microsomal fractions. Pharmacol. Ther. A. 2, 281–317. Fo¨rlin, L. and Andersson, T. (1985) Storage conditions of rainbow trout liver cytochrome P450 and conjugating enzymes. Comp. Biochem. Physiol. 80B, 569–72. Gallagher, E.P., Canada, A.T. and Di Giulio, R.T. (1992) The protective role of glutathione in chlorothalonil-induced toxicity to channel catfish. Aquat. Toxicol. 23, 155–68. Goksøyr, A. (1991) A semi-quantitative cytochrome P450IA1 ELISA: a simple method for studying the monooxygenase induction response in environmental monitoring and ecotoxicological testing of fish. Sci. Total Environ. 101, 255–62. Goksøyr, A., Andersson, T., Hansson, T., Klungsøyr, J., Zhang, Y. and Fo¨rlin, L. (1987) Species characteristics of the hepatic xenobiotic and steroid biotransformation systems of two teleost fish, Atlantic cod (Gadus morhua) and rainbow trout (Salmo gairdneri). Toxicol. Appl. Pharmacol. 89, 347–60. Griffith, O.W. (1980) Determination of glutathione and glutathione disulfide using glutathione reductase and 2-vinylpyridine. Anal. Biochem. 106, 207–12. Gupta, R.C. (1984) Nonrandom binding of the carcinogen N-hydroxy-2-acytylaminofluorene to repetitive sequences of rat liver DNA in vivo. Proc. Natl. Acad. Sci. 81, 6943–7. Gupta, R.C. and Randerath, K. (1988) Analysis of DNA adducts by 32 P-labelling and thin layer chromatography. In: DNA Repair, E.C. Friedberg and P.H. Hanawalt, (eds). pp. 399–418. New York, NY: Marcel Dekker. Hasspieler, B.M., Behar, J.V., Carlson, D.B., Watson, D.E. and Di Giulio, R.T. (1994) Susceptibility of channel catfish (Ictalurus punctatus) and brown bullhead (Ameriurus nebulosus) to oxidative stress: a comparative study. Aquat. Toxicol. 28, 53–64. Johnston, E.P. and Baumann, P.C. (1989) Analysis of fish bile with HPLC-fluorescence to determine environmental exposure to benzo(a)pyrene. Hydrobiologia 188/189, 561–6. Kelly, J.D., Orner, G.A., Hendricks, J.D. and Williams, D.E. (1992) Dietary hydrogen peroxide enhances hepatocarcinogenesis in trout: correlation with 8-hydroxy-29-deoxyguanosine levels in liver DNA. Carcinogenesis 13, 1639–42. Kloepper-Sams, P.J., Park, S.S., Gelboin, H.V. and Stegeman, J.J. (1987) Specificity and crossreactivity of monoclonal and polyclonal antibodies against cytochrome P-450E of the marine fish scup. Arch. Biochem. Biophys. 253, 268–78. Kloepper-Sams, P.J. and Stegeman, J.J. (1989) The temporal relationship between P-450E protein content, catalytic activity and mRNA levels in the teleost Fundulus heteroclitus following treatment with B-naphthoflavone. Arch. Biochem. Biophys. 268, 525–35. Klotz, A.V., Stegeman, J.J. and Walsh, C. (1983) An aryl hydrocarbon hydroxylating hepatic cytochrome P-450 from the marine fish scup (Stenotomus chrysops). Arch. Biochem. Biophys. 226, 578–92. Krahn, M.M., Burrows, D.G., MacLeod, W.D. and Malins, D.C. (1987) Determination of individual metabolites of aromatic compounds in hydrolyzed bile of English sole (Parophrys vetulus) from polluted sites in Puget Sound, Washington. Arch. Environ. Contam. Toxicol. 16, 511–22. Krahn, M.M., Myers, M.S., Burrows, D.G. and Malins, D.C. (1984) Determination of metabolites of xenobiotics in bile of fish from polluted waterways. Xenobiotica 14, 633–46. Krahn, M.M., Rhodes, L.D., Myers, M.S., Moore, L.K., MacLeod, Jr., W.D. and Malins, D.C. (1986) Association between metabolites of aromatic compounds in bile and the occurrence of hepatic lesions in English sole (Parophrys vetulus) from Puget Sound. Arch. Environ. Contam. Toxicol. 15, 61–7. Larsson, A., Orrenius, S., Holmgren, A. and Mannervik, B. eds. (1983) Functions of Glutathione. New York: Raven Press. Maccubbin, A.E., Black, J.J. and Dunn, B.P. (1990) 32 P-postlabelling detection of DNA adducts in fish

Biochemical responses in Niagara River bullhead

33

from chemically contaminated waterways. Sci. Total Environ. 94, 89–104. Maccubbin, A.E., Chidamvaram, S. and Black, J.J. (1988) Metabolites of aromatic hydrocarbons in the bile of brown bullheads (Ictalurus nebulosus). J. Great Lakes Res. 14, 101–8. Malins, D.C. and Haimanot, R. (1990) 4,6-Diamino-5-formamidopyrimidine, 8-hydroxyguanine and 8hydroxyadenine in DNA from neoplastic liver of English sole exposed to carcinogens. Biochem. and Biophys. Res. Comm. 173, 614–9. Malins, D.C. and Haimanot, R. (1991) The etiology of cancer: hydroxyl radical-induced DNA lesions in histologically normal livers of fish from a population with tumors. Aquat. Toxicol. 20, 123–30. Malins, D.C., Ostrander, G.K., Haimanot, R. and Williams, P. (1990) A novel DNA lesion in neoplastic livers of feral fish: 2,6-diamino-4-hydroxy-5-formamidopyrimidine. Carcinogenesis 11, 1045–7. Matsubara, T., Koike, M., Touchi, A., Tochino, Y. and Sugeno, K. (1976) Quantitative determination of cytochrome P-450 in rat liver homogenate. Anal. Biochem. 75, 596–603. Myers, M.S., Landahl, J.T., Krahn, M.M. and McCain, B.B. (1991) Relationships between hepatic neoplasms and related lesions and exposure to toxic chemicals in marine fish from the US west coast. Environ. Health Perspect. 90, 7–15. Nishimoto, M., Roubal, W.T., Stein, J.E., Baird, W.M. and Varanasi, U. (1991) Oxidative DNA damage in tissues of English sole (Parophrys vetulus) exposed to nitrofurantoin. Chem.-Biol. Interact. 80, 317–26. Nishimoto, M., Le Eberhart, B-T., Sanborn, H.R., Krone, C., Varanasi, U. and Stein, J.E. (1995) Effects of a complex mixture of contaminants on hepatic glutathione, L-cysteine, and g-glutamylcysteine synthetase in English sole (Pleuronectes vetulus). Environ. Toxicol. Chem. 14, 461–9. NRTC, (1984) Report of the Niagara River Toxics Committee. US EPA, Environment Canada, New York State Department of Environmental Conservation, and the Ontario Ministry of the Environment. Park, J.W., Cundy, K.C. and Ames, B.N. (1989) Detection of DNA adducts by high-performance liquid chromotography with electrochemical detection. Carcinogenesis 10, 827–32. Pesonen, M., Goksøyr, A. and Andersson, T. (1992) Expression of cytochrome P4501A1 in a primary culture of rainbow trout hepatocytes exposed to -naphthoflavone or 2,3,7,8 tetrachlorodibenzodioxin. Arch. Biochem. Biophys. 292, 228–33. Reddy, M.V. and Randerath, K. (1986) Nuclease P1-mediated enhancement of sensitivity of 32 Ppostlabelling test for structurally diverse DNA adducts. Carcinogenesis 7, 1543–51. Reddy, M.V. and Randerath, K. (1987) 32 P-analysis of DNA adducts in somatic and reproductive tissues of rats treated with the anticancer antibiotic, mitomycin C. Mutat. Res. 179, 75–88. Reichert, W.L., Stein, J.E., French, B., Goodwin, P. and Varanasi, U. (1992) Storage phosphor imaging technique for detection and quantitation of DNA adducts measured by the 32 P-postlabelling assay. Carcinogenesis 13, 1475–9. SAS Institute (1986) SAS User’s Guide: Statistics. SAS Institute. Cary, N.C. Sloan, C.A., Adams, N.G., Pearce, R.W., Brown, D.W. and Chan, S-L. (1993) Northwest Fisheries Science Center Organic analytical Procedures. in: G.G. Lauenstein and A.Y. Cantillo (eds), Sampling and Analytical Methods of the National Status and Trends Program National Benthic Surveillance and Mussel Watch Projects 1984–1992, Comprehensive Descriptions of Trace Organic Analytical Methods, Vol. 4 pp. 53–97 US Department of Commerce, NOAA Technical Memorandum NOS/ORCA/CMBAD-71, Seattle, WA. Stegeman, J.J. and Lech, J.J. (1991) Cytochrome P450 monooxygenase systems in aquatic species: carcinogen metabolism and biomarkers for carcinogen and pollutant exposure. Environ. Health Perspect. 90, 101–9. Stegeman, J.J., Teng, F.Y. and Snowberger, E.A. (1987) Induced cytochrome P-450 in winter flounder (Pseudopleuronectes americanus) from coastal Massachusetts evaluated by catalytic assay and monoclonal antibody probes. Can. J. Fish. Aquat. Sci. 44, 1270–7. Stein, J.E., Collier, T.K., Reichert, W.L., Casillas, E., Hom, T. and Varanasi, U. (1992) Bioindicators of

34

Eufemia et al.

contaminant exposure and sublethal effects: studies with benthic fish in Puget Sound, Washington. Environ. Toxicol. Chem. 11, 701–14. Stein, J.E., Reichert, W.L., French, B. and Varanasi, U. (1993) 32 P-postlabelling analysis of DNA adduct formation and persistence in English sole exposed to benzo[a]pyrene and 7Hdibenzo[c,g]carbazole. Chem. Biol. Interact. 88, 55–69. Stein, J.E., Reichert, W.L., Nishimoto, M. and Varanasi, U. (1989) 32 P-postlabelling of DNA: a sensitive method for assessing environmentally induced genotoxicity. Proceedings, Oceans ’89, Vol. 2 pp. 385–390. IEEE Service Center, Piscataway, NJ, Sept. 18–21, 1989. Stein, J.E., Reichert, W.L., French, B. and Varanasi, U. (1994) Molecular epizootiology: assessment of exposure to genotoxic compounds in teleosts. Environ. Health Perspect. 102 (12), 19–23. Varanasi, U., Collier, T.K., Williams, D.E. and Buhler, D.R. (1986) Hepatic cytochrome P-450 isozymes and aryl hydrocarbon hydroxylase in English sole (Parophrys vetulus). Biochem. Pharmacol. 35, 2967–71. Varanasi, U., Reichert, W.L. and Stein, J.E. (1989a) 32 P-postlabelling analysis of DNA adducts in liver of wild English sole (Parophrys vetulus) and winter flounder (Pseudopleuronectes americanus). Cancer Res. 49, 1171–7. Varanasi, U., Stein, J.E. and Nishimoto, M. (1989b) Biotransformation and disposition of polycyclic aromatic hydrocarbons in fish. In: Metaboism of Polycyclic Aromatic Hydrocarbons in the Aquatic Environment, U. Varanasi (ed) pp. 93–149. Boca Raton, FL: CRC Press. Varanasi, U., Stein, J.E., Nishimoto, M., Reichert, W.L. and Collier, T.K. (1987) Chemical carcinogenesis in feral fish: uptake, activation and detoxification or organic xenobiotics. Environ. Health Perspect. 71, 155–70. Vodicnik, M.J., Elcombe, C.R. and Lech, J.J. (1981) The effects of various types of inducing agents on hepatic microsomal monooxygenase activity in rainbow trout. Toxicol. Appl. Pharmacol. 59, 364– 74. Watson, D.E. (1995) Bioactivation and genotoxicity of 2-aminoanthracene in two Ictalurid species. Ph.D. Thesis, Duke University, Durham, NC. Williams, D.E. and Buhler, D.R. (1984) Benzo(a)pyrene-hydroxylase catalyzed by purified isozymes of cytochrome P-450 from -naphthoflavone-fed rainbow trout. Biochem. Pharmacol. 33, 3743–53. Wirgin, I.I., Grunwald, C., Courtenay, S., Kreamer, G-L., Reichert, W.L. and Stein, J.E. (1994) A biomarker approach in assessing xenobiotic exposure in cancer-prone Atlantic tomcod from the North American Atlantic coast. Environ. Health Perspect. 102, 764–9.