Gill Filament-Based EROD Assay for Monitoring Waterborne Dioxin

Jun 11, 2002 - (EROD) activities in gill filaments, using rainbow trout as ... Introduction. Fish gills are adapted for respiration in a medium wherei...
0 downloads 0 Views 95KB Size
Environ. Sci. Technol. 2002, 36, 3340-3344

Gill Filament-Based EROD Assay for Monitoring Waterborne Dioxin-like Pollutants in Fish E. MARIA JO ¨ NSSON, INGVAR BRANDT, AND BJO ¨ RN BRUNSTRO ¨ M* Department of Environmental Toxicology, Evolutionary Biology Centre, Uppsala University, Norbyva¨gen 18A, SE-752 36 Uppsala, Sweden.

A sensitive, accurate, and straightforward way to determine basal and induced ethoxyresorufin O-deethylase (EROD) activities in gill filaments, using rainbow trout as model species, is described. Tip pieces of primary gill filaments were incubated in tissue culture plate wells containing HEPES-Cortland buffer supplemented with 7-ethoxyresorufin. Each well was sampled on two occasions, and resorufin concentrations were determined by measuring the fluorescence with a plate reader. EROD activity was calculated from the difference in resorufin concentration and the interval between the two samplings. EROD activity was found to be significantly induced by 6 h of exposure to waterborne β-naphthoflavone (1 µM). EROD in gills was also induced by caging rainbow trout in a polluted river or by laboratory exposure of fish to water extracted from that river. There was no loss of EROD activity when gill tissue was kept in ice-cold buffer for up to 1 d, which promotes the use of the method for studying fish exposed in the wild. We propose this novel method as a way to monitor dioxin-like compounds in aquatic environments.

Introduction Fish gills are adapted for respiration in a medium wherein the oxygen concentration is 1/20-1/40 of that in air. Respiratory gas exchange is facilitated through the large epithelial surface in contact with water, the short diffusion distance, counter-current flows of water and blood, and a large water flow. These features make the gills efficient for extracting not only O2 but also organic pollutants from ambient water. Adult rainbow trout, having a ventilation rate of 118 mL/min, extracted approximately 60% of the aqueous O2 during a single pass through the gills (1). Likewise, about 60% of organic chemicals within a Kow range of 103-106 were extracted from the inspired water (1). As in many other organs, cytochrome P450 1A (CYP1A) is induced in fish gills following exposure to aryl hydrocarbon receptor (AhR) agonists, such as polychlorinated dibenzodioxins, dibenzofurans, and biphenyls and polycyclic aromatic hydrocarbons (PAHs). Immunohistochemical studies have shown that endothelial and epithelial gill cells in various fish species respond to CYP1A inducers (e.g., refs 2-4). Some AhR agonists, such as benzo[a]pyrene (BaP), are both inducers of and substrates for CYP1A enzymes. For readily metabolized AhR agonists (e.g., BaP), the exposure route can be crucial for the degree of CYP1A induction in various organs. * Corresponding author phone: +46-(0)18-471 26 26; fax: +46(0)18-51 88 43, e-mail: [email protected]. 3340

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 15, 2002

In the mummichog (Fundulus heteroclitus), exposure to waterborne BaP (10 µg/L) resulted in intense immunohistochemical staining of CYP1A in pillar cells as well as in vascular and cardiac endothelia, whereas hepatic staining of CYP1A was moderate (4). Following dietary exposure (10 µg of BaP/g of food), the CYP1A enzyme was localized primarily in the intestinal mucosa, while hepatic staining was relatively weak (4). This suggests that CYP1A-mediated metabolism in the cells of the intestinal mucosa functions as a barrier to the entry of BaP into the circulation (4). Similarly, although uptake via the gills is an important ingress into the body for waterborne AhR agonists, the absorbed compounds may undergo first-pass metabolism in the branchial cells. When rainbow trout (Oncorhynchus mykiss) gills were perfused and immersed in water containing radiolabeled BaP, more than half of the radioactivity measured in the effluent perfusate constituted BaP metabolites (5). CYP1A-catalyzed O-deethylation of ethoxyresorufin (EROD) is a widely used biomarker for AhR agonist exposure in wildlife. In fish, EROD induction is customarily determined in hepatic microsomes but has, for example, also been measured in branchial microsomes (6) and in epithelial gill cell cultures (7). However, it may be suspected that hepatic EROD induction does not invariably accurately reflect the presence of CYP1A inducing chemicals in the water since pollutants that are absorbed via the gills may be metabolized before reaching the liver. To account for readily metabolized chemicals, environmental monitoring should therefore include a tissue that is proximate to the environment, such as gill tissue. The objective of this study was to develop a sensitive and robust test system for determining EROD induction in fish gills, using rainbow trout as the model species. The system should be applicable both to fish exposed in the wild and to laboratory-exposed fish.

Experimental Section Animals. Two hundred juvenile rainbow trout (ca. 50 g) were purchased from Persbo-Klotens fiskodling AB, Kloten, Sweden. The fish were kept in the aquarium at the Evolutionary Biology Centre (Uppsala University) in a 1-m3 holding tank continuously supplied with aerated Uppsala tap water (1.5 L/min; 11-13 °C). The day/night cycle was adjusted automatically to the diurnal variations at latitude 52° N. The fish were fed pellets (EST40-4 from Aller Aqua, Denmark) once daily, at a ration corresponding to 1% of their body mass. Chemicals and Materials. Dicumarol, β-naphthoflavone (βNF), 7-ethoxyresorufin (ER), sodium salt of resorufin, and dimethyl sulfoxide (DMSO) were all obtained from SigmaAldrich (USA). A mixture of β-glucuronidase and arylsulfatase from Helix pomatia (EC 3.2.1.31/EC 3.1.6.1) was purchased from Roche (Germany). Twelve-well tissue culture plates were obtained from Falcon, Becton Dickinson (USA), and 96-well Fluoronunc plates were from Nunc (Denmark). Biocompatible, glass-encapsulated passive integrated transponder (PIT) tags (ID 100), 2.12 × 11.5 mm, were obtained from Trovan Inc. (Germany). Laboratory Exposure. Transparent polyethylene bags (85 × 60 cm) were placed in opaque plastic boxes (45 × 30 × 15 cm) and filled with 20 L of continuously aerated tap water. The boxes stood in a trough with running tap water to maintain the water temperature at 13 °C. For the methodological studies, groups of rainbow trout (n ) 6) were exposed to 10-6 M βΝF for 1-3 d. βΝF was dissolved in acetone and added to the water in the exposure bags to yield a final acetone concentration of 20 ppm. In an 10.1021/es015859a CCC: $22.00

 2002 American Chemical Society Published on Web 06/11/2002

initial experiment, we found that this acetone concentration had no effect on gill filament EROD activity. One group of rainbow trout (n ) 6), taken directly from the 1-m3 aquarium tank represented the noninduced fish. The fish weighed 87.6 ( 23.5 g (mean ( SD). To study the time course for CYP1A induction in gills, groups of rainbow trout (n ) 5-6) were taken from the 1-m3 tank and exposed to 10-6 M βNF for 0, 3, 6, 12, 24, 48, or 96 h. Other groups of fish (n ) 6-7) were exposed to 10-6 M βΝF for 96 h followed by a recovery period of 1, 4, or 8 d in Uppsala tap water. The water was changed daily. The weight of the fish used in this part of the study was 121.4 ( 21.9 g (mean ( SD). The applicability of the gill filament EROD assay for testing polluted waters was studied by exposing rainbow trout to surface water extracted from the Viskan River near Ga¨sslo¨sa sewage treatment plant (STP) in Borås, Sweden. Water was taken 10 m and 2 km downstream of the STP outlet and at a reference location located about 500 m upstream of Lake O ¨ resjo¨ (which is the drinking water reservoir for Borås, located about 10 km north of the town). The water was transported in polyethylene bags to Uppsala and stored overnight at 4 °C. Next day, groups of 6 rainbow trout (59.1 ( 9.2 g) were transferred from the 1-m3 tank to four polyethylene bags, each containing 20 L of water from one of the locations in Viskan or 20 L of tap water. The fish were exposed for 24 h. As determined in a separate experiment, 24 h of exposure to tap water in these bags did not alter EROD activity. Exposure in the Wild. Rainbow trout were individually marked by implanting a PIT tag in the posterior part of the abdominal cavity. The fish were transported to Borås and caged (21 fish per cage) in Viskan River at five different locations: 10 m, 100 m, and 2 km downstream of the Ga¨sslo¨sa STP outlet, 10 m upstream of the outlet, and at a reference location about 500 m upstream of Lake O ¨ resjo¨. Following 28 d of caging, 5-6 fish from each cage were transported by car to Uppsala in 20 L of water from their respective caging location. The water was kept in polyethylene bags, continuously aerated, and cooled with cool packs during the 10-h transport. From each location, 20 L of extra water was brought in a polyethylene bag. On arrival in Uppsala, the bags with extra water were supplied with aeration, and the caged fish were transferred to their corresponding water and held overnight. Six fish from the 1-m3 tank were used as controls. The fish weighed 66.6 ( 15.9 g at the end of this part of the study. All remaining fish from the cages were transported to Uppsala in a tank with continuously aerated water (approximately 50 L) from the reference location. Of these fish, 5 from each of the cages placed 10 m and 2 km downstream of the STP outlet and 5 from the reference location were placed in a 200-L tank continuously supplied with aerated tap water and held for 8 d. The remaining caged fish were used for other purposes. Gill Filament Preparation and EROD Determination. The fish were struck on the head and decapitated. The gill arches were excised and placed in ice-cold HEPES-Cortland (HC) buffer (0.38 g of KCl, 7.74 g of NaCl, 0.23 g of MgSO4‚ 7H2O, 0.23 g of CaCl2‚2H2O, 0.41 g of NaH2PO4‚H2O, 1.43 g of HEPES, and 1 g of glucose per 1 L of dH2O; pH 7.7). Primary gill filaments were prepared according to McCormick and Bern (8); while immersed in HC buffer, the filaments were cut immediately above the septum, resulting in tip pieces about 2 mm long. From each fish, 2-mm tip pieces were carefully selected by comparison with a standard, and duplicate groups of 10 tip pieces were transferred (using a Pasteur pipet) to wells of a 12-well tissue culture plate containing HC buffer. The HC buffer was replaced with 0.5 mL of ‘‘reaction buffer” consisting of HC buffer supplemented with 10-6 M ER and 10-5 M dicumarol (using stock solutions

of 10-3 M ER and 10-2 M dicumarol dissolved in DMSO). Following 10 min of preincubation at 13 °C and room atmosphere, the buffer was replaced with 0.7 mL of fresh reaction buffer. After 10 and 30 min of incubation (as above), 0.2-mL aliquots were transferred from each well to a Fluoronunc 96-well plate. Resorufin standard solutions (0.5250 nM) were prepared from a stock solution (10 mM in methanol) by dilution in reaction buffer. Duplicate 0.2-mL aliquots of the resorufin standard solutions and of reaction buffer were included on each plate. The fluorescence was determined in a multi-well plate reader (FluostarP, SLT Labinstruments GmbH, Austria) using the wavelengths 544 (ex) and 590 nm (em). EROD activity was calculated and expressed as picomole of resorufin per filament tip and minute. To ascertain whether resorufin formation was linear over time, gill preparations from three βNF-exposed rainbow trout were used. Groups of 10 filament tips were placed in six wells per fish. From each fish, samples were taken from one well at a time at 10-min intervals during 60 min. The amount of resorufin formed per filament tip was plotted against the sampling time. The effect on resorufin concentration of various concentrations of the DT-diaphorase inhibitor dicumarol was studied using gill filament tips from six βΝF-exposed (3 d) fish. For each fish, the assay was run at dicumarol concentrations of 0, 10-6, 10-5, and 10-4 M. The possible reduction of resorufin concentration by its conjugation with glucuronic acid and sulfate was examined in the gill filaments, using a modified version of the method of Donato et al. (9). The assay was run with gill filaments from βΝF-exposed fish (n ) 6) according to the general procedure. Following sampling of duplicates (after 10 and 30 min), 75 µL of sodium acetate buffer (0.1 M; pH 4.6) was added to one of the duplicate samples, whereas 75 µL of the same buffer containing 45 Fisherman units of β-glucuronidase and 360 Roy units of arylsulfatase was added to the other. Following incubation for 2 h at 37 °C, the resorufin concentration was determined. Effects of Storage of Excised Gill Tissue on EROD Activity. The effect of storage in ice-cold buffer on branchial EROD activity was examined in gills from six nonexposed (from the 1-m3 tank) and six βΝF-exposed rainbow trout. Immediately after being excised the gill arches were placed in 50 mL of ice-cold (0 °C) HC buffer and kept for various periods (2 h-8 d) before EROD determination. To study the effect of freezing, gill tissue from βNF-exposed (n ) 3) and nonexposed (n ) 3) fish was frozen in liquid nitrogen directly after dissection. After 4 days of storage at -80 °C, the tissue was thawed in HC buffer, and the gill filament EROD assay was run according to the general procedure. Statistics. EROD activity was regularly measured in duplicate samples of gill filament tips from the same individual, which provided the means to estimate methodological variation. For each fish, the deviation of the duplicates from the duplicate mean was expressed as a percentage of this mean. The variation of the fish within an exposure group was described by the coefficient of variation (standard deviation as a percentage of the mean). Data of noninduced, βNF-induced, and caged fish are presented as means ( SD. One-way ANOVA followed by Dunnett’s t test was used to determine differences between exposure groups and the control/reference groups. Log-transformed values were used when variances differed significantly (Bartlett’s test, p < 0.05).

Results and Discussion Methodology. The fish gill filament method developed in this study proved to be sensitive, accurate, and straightforward for determination of basal and induced EROD activities. VOL. 36, NO. 15, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3341

FIGURE 1. Time course of resorufin formation in primary filaments of rainbow trout gills following 24 h of exposure to 1 µM βNF. Each point indicates the amount of resorufin formed per filament tip piece in a single well. Rings, triangles, and stars represent three different individuals. Whereas traditional, microsome-based techniques require several steps of tissue preparation, this method utilizes intact gill tissue. In addition to saving time, this procedure avoids the risk of methodological variation due to, for example, release of proteolytic enzymes upon homogenization or variable recovery of membrane-bound enzymes upon centrifugation. In the assay presented here, we express the EROD activity as the amount of resorufin formed per ‘‘filament tip”. To accurately determine the activity, filament tips of similar size were carefully selected by comparison with a 2-mm standard. Ten filament tips were included per well to level out differences in size between the filament tips. The small difference in activity obtained between the duplicate measurements supported this methodology. Alternatively, the activity could be related to the amount of protein or to the mass of the gill filaments. Gill filaments are hard to homogenize or sonicate because of the presence of cartilage and connective tissue; therefore lysis of the tissue might be a better choice if proteins should be measured. Ten 2-mm filament tips weigh only about 2 mg, and it is difficult to standardize how to dry them before weighing, which makes reference to mass less attractive. In the present study, we choose to express the activity per filament tip because of the simplicity and accuracy of this methodology. The duplicate deviations in noninduced, βΝF-induced, and caged fish were 6.4 ( 4.6% (n ) 25), 4.5 ( 3.3% (n ) 72), and 5.4 ( 4.8% (n ) 29), respectively. The coefficients of variation for EROD activity in gills of fish from the same box or cage were 40 ( 28% in noninduced fish (n ) 3 boxes), 13 ( 5% (n ) 11 boxes) in βΝF-induced fish, and 33 ( 15% (n ) 5 cages) in fish exposed in the wild. The reason for the greater duplicate variation and coefficient of variation in noninduced vis-a`-vis induced fish was probably that the former had very low gill filament EROD activities (as low as 0.003 pmol of resorufin filament-1 min-1). The greater coefficient of variation in wild-exposed vis-a`-vis laboratoryexposed fish could be due to the uncontrolled exposure situation in the wild, resulting in greater differences in exposure than in the laboratory. For example, a social hierarchy was probably established within the group during the month of caging. Differently ranked individuals in a social hierarchy of fish may vary in stress level, behavior, and feeding (10-12). Resorufin formation in gill filament tips from βΝF-exposed rainbow trout was linear from 10 min until 60 min of incubation (r2 ) 0.98-1.00; Figure 1). However, we noticed that when reaction buffer was added to filament tips, the fluorescence initially declined below the background fluorescence of the buffer. This was probably due to absorption of ethoxyresorufin by the gill tissue. To saturate the tissue with ethoxyresorufin, we preincubated gill filaments in 3342

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 15, 2002

FIGURE 2. Effect of 1, 10, and 100 µM dicumarol on βNF-induced (1 µM) gill filament EROD activity in rainbow trout. Mean ( SD, n ) 6. Dunnett’s t test, repeated measurements. Significant difference from control (clear bar) is denoted by asterisk (*) (p < 0.05). reaction buffer for 10 min before starting the assay by renewing the reaction buffer. As evident in Figure 1, this procedure did not completely prevent initial ethoxyresorufin absorption; if the regression lines were extrapolated toward zero time, they would cross the y-axis at a negative value. As resorufin formation was linear over time from 10 min until 1 h of incubation, the problem could be avoided by sampling each well on two occasions within this period. EROD activity was then calculated from the difference in resorufin concentration and the interval between the sampling occasions. When measuring EROD activity in gill filaments ex vivo, all essential components are present in the tissue except the substrate (7-ethoxyresorufin), i.e., there is no need to add cofactors. On the other hand, ex vivo (as well as whole cell, homogenate, and post-mitochondial supernatant) EROD determination may lead to underestimation of the activity, as the resorufin formed is a substrate for other enzymes, such as DT-diaphorase and conjugation enzymes. Accordingly, Lubet et al. (13) recorded 5-fold higher EROD activities in 9000g supernatants of Arochlor-induced rat and hamster liver, following addition of the DT-diaphorase inhibitor dicumarol to the reaction buffer. In the present assay, inclusion of 10-5 M dicumarol in the reaction buffer increased the resorufin concentration by 33%, whereas 10-6 and 10-4 M dicumarol had less effect (Figure 2). The relatively small effect of dicumarol may have been due to inherent differences between fish and rodents or differences in DT-diaphorase activity between gill and liver. However, despite the quite small effect, we chose to regularly include dicumarol in the assay as it is easily added to the reaction buffer. Fish gills express the conjugation enzymes UDP-glucuronyl transferase, sulfotransferase, and glutathione S-transferase, but the catalytic activities are generally lower in gills than in liver (14, 15). We found a slight increase (6.1 ( 1.0%) in fluorescence following treatment of the reaction products with a mixture of β-glucuronidase and arylsulfatase. The procedure was time-consuming and, as the reaction was carried out at pH 4.6 (optimal for β-glucuronidase activity), the fluorescence was less intense. Because of the methodological inconveniences and the minor effects, hydrolysis of conjugates was not included in the routine assay. In the field study, fish that had been caged were transported to Uppsala in bags of water from the caging location, and EROD was measured within 2 h after excising the gills. However, for practical reasons, it is not always possible to bring live fish from the ‘‘field” to the laboratory. Therefore, we investigated the possibility to store gill tissue; i.e., gill tissue was frozen (-80 °C) or kept for various periods in ice-cold buffer. Following freezing and thawing, EROD activity could not be detected with the gill filament assay. This could be due to cell rupture when freezing and thawing the tissue. Following 25 h of storage of βΝF-induced gill filaments in ice-cold buffer, EROD activity was slightly higher

FIGURE 3. Effect of storage of gill tissue in ice-cold buffer for 2 h-8 d on EROD activity in gill filaments from (a) βNF-exposed (1 µM) and (b) nonexposed rainbow trout. Mean ( SD, n ) 6. than the activity measured 2 h after dissection (118% of the 2-h mean; Figure 3a). Following 3, 5, and 7 d of storage in ice-cold buffer, EROD activity had decreased to 78, 28, and 6% of the 2-h mean, respectively (Figure 3a). In nonexposed fish, 25 h of storage of gill tissue in icecold buffer more than doubled gill filament EROD activity (236% of the 2-h mean; Figure 3b). In these filaments, EROD activity remained elevated after 2 and 4 d (221 and 160% of the 2-h mean; Figure 3b). After 8 d, the tissue appeared partly decomposed, and the EROD activity had fallen to below the initial mean (61% of the 2-h mean; Figure 3b). In filaments stored for 1 d, resorufin formation was linear from 10 until 60 min of incubation (not shown), and deviation between duplicates was low (4% in βΝF-induced and 15% in noninduced). The reason for the increase in EROD following storage is not known. Nevertheless, having the option to postpone measurement up to 1 d following killing of fish is sometimes so advantageous that these variations can be considered acceptable (provided that the gill samples within a single study are stored for a similar length of time). EROD Activity following Caging in Viskan and Exposure to Viskan Water. Gill filament EROD activity was higher in fish from the reference location in the Viskan river system than in those from the tank with Uppsala tap water (Figure 4a). This reference location (500 m upstream of Lake O ¨ resjo¨) was chosen because it was located in a nature conservation area and therefore not directly affected by industrial or municipal effluents. Hence, the increased EROD activity in the fish caged at this location may be assumed to reflect the degree of pollution in the area generally. In all groups of rainbow trout caged downstream of the STP in Borås, gill filament EROD activity was markedly induced (Figure 4a). It was also induced by short-term (24-h) laboratory exposure to water collected downstream of the STP outlet (Figure 4b). The results from the short-term exposure show that EROD is rapidly induced in gills of fish exposed to polluted water. In this experiment, transport of water as well as exposure of fish was carried out in plastic bags. It should be noted that hydrophobic chemicals can adsorb to plastic, which may have resulted in loss of inducing capacity. Borås has for many decades been the location for several textile and engineering firms, and the area is heavily polluted by a wide variety of chemicals. However, since fish

FIGURE 4. Gill filament EROD activity in rainbow trout: (a) Fish caged (28 d) in Viskan River system; at a reference location 500 m upstream of Lake O2 resjo1 (Ref.), 10 m upstream (-10 m), and 10 m (+10 m), 100 m (+100 m), and 2 km (+2 km) downstream of the STP in Borås or fish from a tap water tank in the aquarium facility (Tap w.). (b) Fish exposed (24 h) to water collected from the locations 10 m (+10 m) and 2 km (+2 km) downstream of the STP, to water from the reference location (Ref.), or to tap water (Tap w.). Mean ( SD, n ) 5-6. Dunnett’s t-test. Significant differences from the reference (clear bar) are denoted by three asterisks (***) (p < 0.001).

FIGURE 5. Gill filament EROD activity in rainbow trout following 0-96 h of exposure to 1 µM βNF (b). Groups of fish were also exposed to 1 µM βNF for 96 h and then allowed to recover in tap water for 1-8 d (O). Mean ( SD, n ) 5-7. caged 10 m upstream of the STP outlet showed weaker EROD induction than those caged downstream, the major source of EROD-inducing compounds was probably the STP. Engwall et al. (16) showed that organic extracts of sewage sludge collected from Borås STP strongly induced EROD in vitro, thus demonstrating that the water entering the STP is polluted with dioxin-like chemicals. The results of the present study imply that water released from the STP in Borås also contains EROD-inducing chemicals. EROD Induction and Recovery. The induction of branchial CYP1A following exposure to waterborne βΝF (1 µM) was quite rapid. After 6 h of βΝF exposure, EROD activity had increased to 14% of the maximum level (mean at 96 h), and maximum induction was reached within 24-96 h (Figure 5). The time required to reach maximum induction apparently varied between fish from different boxes (Figure 5). The VOL. 36, NO. 15, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3343

concentration of dissolved oxygen in the water could affect chemical uptake through the gills (17). When aqueous oxygen pressure is lower, more water must be filtered through the gills, with greater exposure to waterborne pollutants in consequence (18). Although attempts were made to balance the aeration pressure between the boxes in our study, differences in resistance of the air stone diffusers could have caused unequal aeration. Branchial EROD activity still remained maximally induced after a 24-h recovery period in tap water (i.e., after a 4-d βΝF-exposure period followed by 24 h in tap water) but had declined to 14 and 7% of the maximum level after 4 and 8 d of recovery (Figure 5). Similarly, 8 d in tap water reduced the branchial EROD activity in fish that had been caged 10 m and 2 km downstream of the Borås STP and at the control location to 6.1 ( 1.3%, 4.9 ( 1.2%, and 13.1 ( 5.0%, respectively, of the mean activity of the corresponding group directly after caging. The decline following transfer to clean water probably reflected clearance of EROD-inducing substances from the gills. Levine and Oris (19) studied the course of CYP1A induction in gills and liver of rainbow trout following exposure to waterborne BaP (1.23 µg/L) and found that CYP1A mRNA in gills and liver was maximally induced after 6 and 12 h, respectively. Whereas the CYP1A-catalyzed BaP hydroxylase activity in gills reached its maximum after 24 h and then remained at this level throughout the experiment (120 h), hepatic EROD activity peaked at 24 h, and then decreased to about one-third of maximum at 120 h (19). As shown by these authors, branchial biotransformation by CYP1A can function as a first-pass defense against readily metabolized waterborne AhR agonists, such as BaP. This metabolic efficiency could be due to expression of CYP1A in both the epithelial and the endothelial/pillar cells through which waterborne compounds have to pass when absorbed by the gill. The first-pass metabolic capability of the gill has implications for the interpretation of EROD activity measured in other organs, e.g., in environmental monitoring (19). It must be considered that environmental pollution most often represents a mixture of chemicals and that EROD activity is an integrated measure that reflects the concentration of various CYP1A inducers and inhibitors reaching the monitored organ. Hepatic EROD activity can act as an appropriate biomarker of persistent AhR agonists but is of limited value as an indicator of exposure to readily metabolized AhR agonists (19). The gill filament-based EROD assay described in this paper constitutes a new measure with which to reveal waterborne CYP1A inducers, both in the laboratory and in the field. EROD induction was recorded in fish caged at the reference location in the Viskan water system, suggesting that the concentrations

3344

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 15, 2002

of dioxin-like pollutants in areas lacking point sources also are high enough to be detected by this assay. The small degree of variation, observed between duplicates, suggests excellent reproducibility. We are currently investigating whether the assay can be applied to species other than rainbow trout. In conclusion, we propose this straightforward and robust assay for use as a means to reveal and monitor dioxin-like pollution in aquatic environments.

Acknowledgments We thank Kifle Ghebreab, Elin Wallin, and Jan O ¨ rberg for collaboration during the field experiment. Economic support was provided by the Swedish Environmental Protection Agency (ReproSafe) and the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas).

Literature Cited (1) McKim, J.; Schmieder, P.; Veith, G. Toxicol. Appl. Pharmacol. 1985, 77, 1-10. (2) Smolowitz, R. M.; Hahn, M. E.; Stegeman, J. J. Drug Metab. Dispos. 1991, 19, 113-123. (3) Husøy, A. M.; Myers, M. S.; Goksøyr, A. Aquat. Toxicol. 1996, 36, 53-74. (4) Van Veld, P. A.; Vogelbein, W. K.; Cochran, M. K.; Goksøyr, A.; Stegeman J. J. Toxicol. Appl. Pharmacol. 1997, 142, 348-359. (5) Andersson, T.; Pa¨rt, P. Mar. Environ. Res. 1989, 28, 3-7. (6) Lindstro¨m-Seppa¨, P.; Oikari, A. Aquat. Toxicol. 1990, 16, 187204. (7) Carlsson, C.; Pa¨rt, P.; Brunstro¨m, B. Aquat. Toxicol. 1999, 47, 117-128. (8) McCormick, S. D.; Bern, H. A. Am. J. Physiol. 1989, 256, R707R715. (9) Donato, M. T.; Go´mez-Lecho´n, M. J.; Castell, J. V. Anal. Biochem. 1993, 213, 29-33. (10) McCarthy, D.; Carter, C. G.; Houlihan, D. F. J. Fish Biol. 1992, 41, 257-263. (11) Alana¨ra¨, A.; Bra¨nna¨s, E. J. Fish Biol. 1996, 48, 242-254. (12) Gregory, T. R.; Wood, C. M. Physiol. Biochem. Zool. 1999, 72, 286-295. (13) Lubet, R.A.; Nims, R. W.; Mayer, R. T.; Cameron, J. W.; Schechtman, L. M. Mutat. Res. 1985, 142, 127-131. (14) Lindstro¨m-Seppa¨, P.; Koivusaari, U.; Hanninen, O. Comp. Biochem. Physiol. 1981, 69C, 259-263. (15) Leguen, I.; Carlsson, C.; Perdu-Durand, E.; Prunet, P.; Pa¨rt, P.; Cravedi, J. P. Aquat. Toxicol. 2000, 48, 165-176. (16) Engwall, M.; Brunstro¨m, B.; Na¨f, C.; Hjelm, K. Chemosphere 1999, 38, 2327-2343. (17) Kim, J. M.; Goeden, H. M. Comp. Biochem. Physiol. 1982, 72C, 65-74. (18) Lloyd, R. J. Exp. Biol. 1961, 38, 447-455. (19) Levine, S. L.; Oris, J. T. Aquat. Toxicol. 1999, 46, 279-287.

Received for review December 19, 2001. Revised manuscript received March 28, 2002. Accepted May 7, 2002. ES015859A