Environ. Sci. Technol. 2003, 37, 256-260
Chorioallantoic Membranes Indicate Avian Exposure and Biomarker Responses to Environmental Contaminants: A Laboratory Study with White Leghorn Chickens (Gallus domesticus) T I M O T H Y A . B A R G A R , * ,† GEOFFREY I. SCOTT,‡ AND GEORGE P. COBB† The Institute of Environmental and Human Health, Department of Environmental Toxicology, Texas Tech University, Lubbock, Texas 79409, and USDOC/NOAA/NOS/CCEHBR, 219 Fort Johnson Road, Charleston, South Carolina 29412-9110
PCB and endosulfan concentrations in the chorioallantoic membrane (CAM) of white leghorn chickens (Gallus domesticus) were evaluated as indicators of hepatic cytochrome P450 isozyme activity in hens and chicks as well as toxicant concentrations in eggs and hens. Sixteen hens were randomly divided into four groups of four and dosed with a mixture of PCB105 (2,3,3′,4,4′-pentachlorobiphenyl), PCB156 (2,3,3′,4,4′,5-hexachlorobiphenyl), PCB189 (2,3,3′,4,4′,5,5′-heptachlorobiphenyl), and technical grade endosulfan (3:1 ratio of R and β isomers) at three different dose groups. The first 10 fertile eggs laid by each hen were collected, the even-numbered eggs incubated until hatched, and the odd numbered eggs were analyzed for test chemicals. Strong (r2), significantly positive (p value) relationships were found between total PCB mass (ng) in CAMs and both total PCB concentrations (ng/g wet wt) in adults (r2 ) 0.91, p ) 0.0001) and eggs (r2 ) 0.87, p ) 0.0001). The relationship between total PCB mass in CAMs and hepatic cytochrome p450 isozyme activity in chicks (r2 ) 0.49, p ) 0.0001) and hens (r2 ) 0.45, p ) 0.014) was also significant but not as strong. This study shows that CAMs can be used to estimate avian exposure to PCBs and resultant biological response.
Introduction Recently published studies evaluated chorioallantoic membranes (CAMs) as indicators of avian and reptilian exposure to chlorinated compounds with mixed conclusions. Some have shown that a significant positive relationship exists between chemicals in the CAM, remaining egg contents, and organ tissues for American alligators (Alligator mississippiensis) (1, 2), great blue herons (Ardea herodias) (3), and loggerhead sea turtles (Carretta carretta) (4), leading the authors to advocate CAM use in exposure assessments. Other studies have attempted to explain the relationship and found * Corresponding author telephone: (703)605-1531; fax: (703)3056309; e-mail:
[email protected]. † Texas Tech University. ‡ USDOC/NOAA/NOS/CCEHBR. 256
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
that neither lipid distribution nor simple chemical partitioning between the embryo and CAM adequately explained the relationship (5), leading the authors to conclude that caution should be exercised when using CAMs in exposure assessments. The CAM has been proposed for use as a non-lethal, noninvasive indicator of oviparous organism exposure to persistent organics for several reasons. First, the CAM is a tissue of embryonic origin. It is derived from each of the three primary germ layers and grows from the abdominal region of the developing embryo. Second, the respiratory function of the CAM is critical to embryo development. As such, CAMs are highly vascularized with considerable blood volume passing through from the embryo proper. As a result, the CAM is likely to contain detectable quantities of chemicals to which the embryo is exposed. Third, CAMs may be nonlethally and non-invasively collected from discarded eggshells. During the last day of development, the respiratory function of the CAM is transferred to the embryo, meaning that the CAM is discarded with the shell when the neonate hatches. Last, larger sample numbers may be obtained since CAMs can be collected non-lethally and non-invasively. Risk assessments would benefit from the greater precision afforded by larger sample numbers. Because of the apparent conflicting opinions over CAM use in assessing organism exposure to organochlorines (6, 7), two hypotheses were tested in this study. Our first hypothesis was that chemical concentrations in the CAM could be used to predict concentrations in the egg, chick, and adult. The second hypothesis was that chemical concentrations in CAMs could be used as an indicator of biomarker response in chicks and adults. In this study, CAMs were collected from newly hatched white leghorn (Gallus domesticus) chicks that had been maternally exposed to three different PCB congeners and endosulfan. Toxicant concentrations in the CAMs were compared to the dose administered to hens, to concentrations in laid eggs, and to biomarker responses in chicks and hens. The utility of the CAM in nonlethal assessment of avian exposure to chlorinated chemicals is discussed in light of data gained from this study.
Materials and Methods Sixteen actively laying (approximately 1 egg/day) white leghorn chicken hens were randomly divided into four groups of four hens. Each group was designated as a control, low, medium, or high dose group. Subcutaneous dosing and artificial insemination of hens are described elsewhere (57). Dosing solutions contained a mixture of the following chemicals: PCB105 (2,3,3′,4,4′-pentachlorobiphenyl), PCB156 (2,3,3′,4,4′,5-hexachlorobiphenyl), PCB189 (2,3,3′,4,4′,5,5′heptachlorobiphenyl), and technical grade endosulfan (R and β isomers with an analytically determined ratio of 3:1). The anticipated individual chemical concentrations in the low, medium, and high dose solutions were 0.625, 1.25, and 1.875 µg/µL, respectively. Concentrations for each chemical were chosen based on two factors, avoiding embryo mortality and ensuring detectability in the CAMs. An LD50 for PCB105 to 11-day-old chicken embryos of 5592 µg/kg egg has been reported (8). This would be approximately 280 µg of PCB105 assuming an average egg size of 50 g for chickens giving whole egg concentrations of 5.6 µg/g, well above the instrument detection limits for these chemicals. Assuming that 1% of total chemical body burden is excreted into eggs (9), dose solution concentrations were chosen that were well below 5.6 µg/g, thereby avoiding excessive embryo mortality, 10.1021/es0257873 CCC: $25.00
2003 American Chemical Society Published on Web 12/13/2002
lessening the likelihood of adult mortality, and ensuring analyte detection in CAMs. Sample Collection. Ten eggs were collected from each hen once egg fertility was confirmed. Each egg was labeled with a hen identifier and a number indicating the position in the laying sequence. All odd-numbered eggs were incubated at 99 °F and 60% humidity until hatched. All evennumbered eggs were held at 0 °C until extraction and analyses. Care and monitoring of eggs are also described elsewhere (6, 7). Eggs that had not developed noticeable circulatory systems by incubation day 6 were considered infertile. As soon as hatchlings were discovered, spent shells were removed from the incubator; CAMs were removed using solvent-rinsed forceps. Dried CAMs were moistened with distilled water to facilitate their removal from the shells. All CAMs were placed into preweighed, solvent-rinsed scintillation vials. Hatched chicks were removed from the incubator, weighed, banded to denote the dose group, and placed into a brooder with their dose group until day 7 post-hatch. Livers were removed from chicks and hens for cytochrome P450 assays. Tissue was removed from each hen after it had laid the 10th egg and from each chick day 7 post-hatch. Livers were weighed, wrapped in aluminum foil, and placed into liquid nitrogen before storage at -80 °C. Extractions. Analyte isolation and quantification procedures are described elsewhere (6, 7). Briefly, tissues were mixed with Na2SO4 and spiked with decachlorobiphenyl (DCBP). Samples were Soxhlet extracted for 16 h in dichloromethane (MeCl2). Total extract volumes for each egg were rotary evaporated to 5 mL. One milliliter was removed for gravimetric lipid determination, and 400 µL of the crude extract was cleaned by GPC and silica gel chromatography. Recoveries of test compounds from spiked samples were 92% for PCB105, 101% for both PCB156 and PCB189 ((1.3% for all), 69% ( 8.5% for R-endosulfan, and 83% ( 7.2% for β-endosulfan. CAMs were homogenized with Na2SO4 and spiked with DCBP. The mixture was extracted by elution with 150 mL of 10% MeCl2 in hexane followed by 100 mL of 25% ethyl acetate in hexane. The cleanup procedure for the CAM extracts was the same as for the egg extracts. Fully processed extracts were analyzed by GC/ECD using dual-column confirmation: 30 m DB-1701 (0.25 mm i.d., 0.25 µm film thickness) and DB-5 (same dimensions). ECD data were adjusted for recovery of DCBP but not for individual analyte extraction efficiencies. Cytochrome P450 Assays. Activities were determined for mono-oxygenase enzymes that are important for xenobiotic metabolism: ethoxyresorufin- (EROD), methoxyresorufin(MROD), and butoxyresorufin-O-dealkylase (BROD). Details of these assays are described elsewhere (13). Briefly, a portion of each liver was weighed and homogenized on ice in buffer A (Tris buffer containing 250 mM sucrose and adjusted to pH 7.4). The homogenate was centrifuged three times at 4 °C using a clean centrifuge tube each time. Centrifugation conditions speeds were 10000g for 10 min, 15000g for 20 min, and 105000g for 1 h. The pellet of microsomes was washed with chilled (4 °C) Tris buffer adjusted to pH 7.4 and containing sucrose and KCl. Microsomes were resuspended by homogenization in 80% buffer A+ 20% aqueous glycerol. Activities within the isolated microsomes were assayed in triplicate in the presence of buffer, the appropriate substrate, and NADPH. Enzyme activity was measured with a fluorescence plate reader after a period of incubation. Results are reported in pmol min-1 (mg of protein)-1. Statistical Analyses. One-way ANOVA (10) (R ) 0.05) was used to compare among dose groups (i) chemical concentrations in dosing solutions, (ii) mean hepatic cytochrome P450 isozyme activity (log pmol min-1 (mg of protein)-1) in chicks and adult hens, (iii) mean PCB concentrations in eggs
FIGURE 1. PCB concentrations (ng/g tissue) in white leghorn hens and eggs and PCB mass (ng) in CAMs. Variance bars are standard errors of the mean. (ng/g), and (iv) mean PCB mass (ng) extracted from CAMs. Model I regression analyses were used to investigate the relationship between total PCB mass (ng) extracted from CAMs and (i) total PCB concentrations (ng/g tissue) in hens and whole eggs and (ii) hepatic cytochrome P450 isozyme activity (pmol min-1 (mg of protein)-1) in hens and chicks. In these comparisons, PCB mass in the CAMs was the independent variable while PCB concentrations (eggs or adults) and P450 isozyme activity were dependent variables. PCB mass in CAMs was utilized in all statistical analyses rather than concentrations to eliminate potential bias in tissue weight caused by moistening of CAMs before removal from the shells. Chemical concentrations were not measured in hens but were estimated on the basis of the mass injected, hen weight, and a lipid content of 14% for white leghorn chickens.
Results Chemical Concentrations. PCB concentrations in hens, eggs, and CAMs are reported in Figure 1. The increase in PCB concentrations among dose groups was consistently significant among dose groups for eggs and generally consistent among dose groups for CAMs. It should be noted that a positive relationship between egg laying order and total PCB mass excreted into the eggs was evident (Figure 2). However, PCB concentrations in eggs were not significantly different among dates within a given dose group during the study. No endosulfan was detected in either eggs (MDL ) 0.009 pg/g wet weight) or CAMs (MDL ) 0.7 pg/g wet weight). Enzyme Activities. Mean hepatic cytochrome P450 activities were not significantly different among dose groups for hens but were significantly different among dose groups for chicks (Figure 3). Correlations. Significant relationships were evident among PCB concentrations in CAMs, eggs, and hens. Mean total PCB mass among CAMs of sibling chicks was highly related to both the total PCB concentration within the respective VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
257
FIGURE 2. Temporal pattern of PCB concentrations in eggs produced by hens (Gallus domesticus) exposed to a PCB mixture (PCB 105, PCB152, and PCB189).
FIGURE 4. Relationship of total PCB mass, averaged among sibling CAMs, with total PCB concentration in parental hens [b] (r2 ) 0.91) and average PCB concentration among same clutch eggs [+] (r2 ) 0.87).
TABLE 1. Regression Analyses for Hen and Chick Hepatic Cytochrome P450 Isozyme Activity with Dose Group and with Total PCBs in Whole Eggs and CAMs enzyme
FIGURE 3. Hepatic cytochrome P450 isozyme activity in hens and chicks in each dose group. Enzyme activity was not significantly different among dose groups for hens (r ) 0.05). Enzyme activity was significantly different among dose groups for chicks. Variance bars are standard errors of the mean. maternal hen (r2 ) 0.91) and the mean total PCB concentration in eggs from the same clutch (r2 ) 0.87) (Figure 4). In addition, individual PCB congener masses in CAMs and concentrations in eggs were significantly related. The strongest relationship was for hexachlorinated PCB156 (r2 ) 0.68), followed by heptachlorinated PCB189 (r2 ) 0.57) and pentachlorinated PCB105 (r2 ) 0.37). Generally, hepatic enzyme activity and PCB burdens were significantly correlated. Total PCB concentrations in whole eggs were significantly related to hepatic enzyme activity for chicks (R ) 0.05) but not for hens (Table 1). Total PCB mass 258
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
EROD MROD BROD EROD MROD BROD
hen r2 (p) P450 vs Dose Group 0.39 (0.043) 0.01 (0.810) 0.22 (0.115)
chick r2 (p) 0.69 (0.0001) 0.56 (0.0001) 0.50 (0.0001)
P450 vs Whole Egg Total PCBs 0.18 (0.169) 0.65 (0.003) 0.00 (0.846) 0.61 (0.004) 0.21 (0.251) 0.49 (0.024)
EROD vs CAM Total PCBs, Average of Sibling CAMs (ng) EROD 0.45 (0.014) 0.49 (0.0001)
in CAMs was significantly related to hepatic enzyme activity in both hens and chicks (Figure 5).
Discussion PCB burdens in CAMs, eggs, and adults are highly related. In our study, the coefficients of determination (r2) for the
FIGURE 5. Correlation between total PCB mass extracted from CAMs and both chick [+] (r2 ) 0.49) and hen [b] (r2 ) 0.45) hepatic ethoxyresorufin-O-dealkylase (EROD) activity. CAM to egg comparison (0.87) and the CAM to hen comparison (0.91) indicated that the CAM accurately estimates PCB concentrations in eggs and adults. No other study has evaluated PCB burdens between CAMs and adults. Strong relationships for total PCB concentrations have also been observed between alligator CAMs and eggs (r2 ) 0.67) (2) and loggerhead sea turtle CAMs and eggs (r2 ) 0.78) (4). Although not specifically evaluated by Pastor et al. (5), a bivariate plot of their organochlorine data for CAMs and embryos of audouin’s sea gulls (Larus audouinii) also revealed a strong relationship (r2 ) 0.77). It is evident that organochlorine contamination in CAMs, adults, and their eggs covary to a significant extent regardless of the species. Chemical structure appears to affect the distribution of individual congeners in CAMs and eggs. As noted earlier, the coefficient of determination for CAM vs egg comparisons in this study was greater for the hexachlorinated PCB156 (r2 ) 0.677) than for either the pentachlorinated PCB105 (r2 ) 0.366) or heptachlorinated PCB189 (r2 ) 0.573). A similar relationship has been reported for CAM vs egg comparisons in alligators wherein regressions were stronger for hexachlorinated biphenyls (r2 ) 0.92) than for either pentachlorinated (r2 ) 0.72) or heptachlorinated (r2 ) 0.68) biphenyls (2). Similar trends were also observed for loggerhead sea turtles where the coefficient for heptachlorinated biphenyls (r2 ) 0.46) was again less than the coefficient for hexachlorinated biphenyls (r2 ) 0.75) (4). However, the coefficient for pentachlorinated biphenyls (r2 ) 0.78) was slightly greater than the coefficient for the hexachlorinated biphenyls. These data from our study and from previous studies demonstrate a relationship between chemical structure and the strength of the regression between PCBs in the CAM and embryo. The potential influence of PCB chemical structure on chemical partitioning between adipose tissues and plasma has been reviewed (16). Twenty-eight different parameters cited in the literature that might influence fat to plasma partitioning (Kfp) have been reported as factors that could influence PCB excretion from the body (16). Three of these parameters explained the majority of the variance in partitioning; the degree of non-chlorination at neighboring meta-para positions (UNS); the degree of nonplanarity (NPL), which is an indirect measure of ortho-chlorination; and polarity (DIFF2), an indirect measure of chlorine “balance” between the phenyl rings. Fat to plasma partitioning was inversely proportional to UNS and proportional to both NPL and DIFF2. The congeners utilized in this study differed with regards to only one of the previously mentioned structural parameters. The congeners were selected to minimize the likelihood of metabolism; none had neighboring, non-chlorinated meta-para positions. In addition, they were all mono-orthochlorinated. However, DIFF2 did vary among the congeners. PCB156 had a DIFF2 of two (four chlorines vs two chlorines)
while both PCB105 and PCB189 had a DIFF2 of one. Since PCB156 may have a greater Kfp than either PCB105 or PCB189 and since its burdens in CAMs and eggs are more strongly related than for other PCBs, the relationship between PCB concentrations in CAMs and eggs might be stronger for PCBs with a greater Kfp. The Kfp of one congener versus another was not addressed in this study, but the relationship between PCBs burdens in CAMs and eggs was addressed. The role DIFF2 may play in congener movement should be evaluated. Our study not only supported the results of previous studies that showed a positive relationship of chemical concentrations in CAMs and eggs but also demonstrated a significant relationship between chemical mass in the CAM and biological response in chicks and adults. Cytochrome P450 isozyme activity is well-known to be responsive to certain PCBs (9, 12). Indeed this was the case in our study as EROD response, in both chicks and adults, was related to PCB mass in CAMs. It should also be noted that differences in temporal PCB deposition into eggs likely contributed to within clutch variability in both CAM PCB burdens and chick EROD activity. This could have decreased the strength of the relationship between total PCB burdens among CAMs, hens, and eggs, and with biological response in the chicks and adults. Hepatic EROD activity in the hens and chicks in our study were generally greater than the activity in other species. Previous evaluation of daily oral administration of Aroclor 1254 to white leghorn hens, at doses ranging from 5 to 40 mg/day, induced mean hepatic EROD activities ranging from 100 to about 380 (11). These were slightly lower activities than we found. However, PCB absorption from the gastrointestinal tract and the presence of other congeners in the Aroclor mixture may alter the EROD response in their study relative to the response found in our study. Japanese quail (Coturnix coturnix japonica) were orally dosed over a 2-week period with an average daily exposure of 3 mg kg-1 day-1 of PCB105 and subsequently exhibited mean EROD activity of 730 ( 250 (15). While this was near the activity found for hens in our high dose group, the average daily exposure for the quail was considerably higher than the average daily exposure for these hens (0.21 mg kg-1 day-1). This indicates a lower PCB induced monooxygenase activity in quail as compared to white leghorn hens. In fact, domestic species are often more susceptible to chlorinated contaminant exposures than are wild species. This has been documented with rodents (12) and several bird species, such as the barn owl (Tyto alba), African bulbul (Pycnonotus capensis), and house sparrow (Passer domesticus) (17). EROD activity in the 1-week-old chicks from our study was considerably greater than the activity in 10-day-old white leghorn embryos exposed to PCB156 through injection into the air sac of a 7-day-old incubating egg. The low concentration in their study (0.1 mg/kg egg), which approximated that of the high concentration determined in eggs from the high dose group in our study (0.12 mg/kg egg), led to an EROD activity of approximately 20 pmol min-1 (mg of protein)-1 (18). EROD activity in 10-day-old white leghorn embryos similarly exposed to 0.1 mg/kg egg of PCB169 was approximately 60 pmol min-1 (mg of protein)-1 (19). Obviously, enzyme activity in white leghorn embryos has not developed to the level of recently hatched chicks. EROD activity in 1-week-old white leghorn chicks is considerably greater than the activity in chicks of wild species. Twelve-hour-old common tern chicks (Sterna hirudo), while exposed to total PCB concentrations (400 µg/g egg) much higher than the chicks in our study (432 ng/g egg, high dose group) as well as to a variety of dioxins and furans, had a lower average EROD activity (≈400 as compared to ≈2900 pg min-1 (mg of protein)-1) (20, 21). One-day-old bald eagle chicks (Haliaeetus leucocephalus) containing total PCB VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
259
concentrations of 559 mg/kg egg lipid in yolk sacs had an average EROD activity of 9.3 pmol min-1 (mg of protein)-1 (22). This exposure was much greater than the exposure level in our study (4.0 mg/kg egg lipid, high dose group) that resulted in an EROD activity of ≈2900 pg min-1 (mg of protein)-1. The eagle chicks also contained a mean nonortho PCB concentration of 61.5 µg/kg egg lipid, which was slightly lower than the total PCB concentrations observed in our low dose group (Figure 1), which induced EROD to ≈500 pg min-1 (mg of protein)-1. By either comparison, the white leghorn chickens utilized in our study have much more responsive hepatic cytochrome P450 enzymes than wild bird species. The relationship between CAM contamination and biological response in both chicks and hens, while significant, was not as strong as that of chemical burdens in CAMs, eggs, and adults. As indicated by the coefficients of determination, approximately 90% of the variation in total PCB concentrations in the hen (r2 ) 0.91) or egg (r2 ) 0.87) were described by variation in total PCB mass in the CAMs. Approximately 50% of the variation in chick (r2 ) 0.50) and hen (r2 ) 0.45) hepatic EROD activity was described by variation in total PCB mass in the CAM. Thus, contaminant quantity in the CAM is more accurate in estimating contaminant concentrations in the egg or adult than in estimating biological response. Averaging PCB exposures within a given nest provided a reasonable estimate of maternal exposure since the average chick exposure to PCBs should approximate maternal exposure, which controls cytochrome response characteristics. Regression of average responses or average exposures would mask individual variability in biomarker response and exposure in the field. Individual responses may be critical factors in evaluating differences among sites with potentially different exposure conditions. The results from this study support the use of CAMs to assess oviparous organism exposure to persistent chlorinated organics. Chemical burdens in adults and in eggs were accurately described by chemical burdens in CAMs. In addition, biological response in adults and neonates were significantly related to chemical burdens in CAMs. Therefore, comparison of chemical burdens in CAMs collected from two different sites should provide a relative indication of adult and neonatal exposure to chemicals as well as the relative biological response of organisms at the two sites.
Acknowledgments Funding for this research was provided by Dupont de Nemours, ECORISK Inc., and The National Institute of Environmental Health Sciences (grant number P42 ES04696). We are grateful to the National Ocean Service for providing
260
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
analytical equipment used in this project. Special thanks are extended to Dr. Milton Taylor for assistance in determining cytochrome activities and to Dr. David Otis for providing guidance in statistical analyses.
Literature Cited (1) Bargar, T. A.; Sills-McMurry, C.; Dickerson, R. L.; Rhodes, W. E.; Cobb, G. P. Arch. Environ. Contam. Toxicol. 1999, 37, 364-368. (2) Cobb, G. P.; Wood, P. D.; O’Quinn, M. Environ. Toxicol. Chem. 1997, 16, 1456-1462. (3) Cobb, G. P.; Norman, D. M.; Miller, M. W.; Brewer, L. W. Chemosphere 1995, 30, 151-164. (4) Cobb, G. P.; Wood, P. D. Chemosphere 1997, 34, 539-549. (5) Pastor, D.; Ruiz, X.; Jover, L.; Albaiges, J. Environ. Toxicol. Chem. 1996, 15, 167-171. (6) Bargar, T.A.; Scott, G. I.; Cobb, G. P. Environ. Toxicol. Chem. 2001, 20, 61-67. (7) Bargar, T. A.; Scott, G. I.; Cobb, G. P. Arch. Environ. Contam. Toxicol. 2001, 41, 508-514. (8) Powell, D. C.; Aulerich, R. J.; Stromberg, K. L.; Bursian. S. J. J. Toxicol. Environ. Health 1996, 49, 319-338. (9) Nosek, J. A.; Craven, S. R.; Sullivan, J. R.; Olson, J. R.; Peterson, R. E. J. Toxicol. Environ. Health 1992, 35, 153-164. (10) SAS. Statistical Application Software, Version 8.0; SAS Institute Inc.: Cary, NC, 1996. (11) Chen, S. W.; Dziuk, P. J.; Francis, B. M. Environ. Toxicol. Chem. 1994, 13, 789-796. (12) Burke, M. D.; Thompson, S.; Elcombe, C. R.; Halpert, J.; Haaparanta, T.; Mayer, R. T. Biochem. Pharmacol. 1985, 34, 3337-3345. (13) Trust, K. A.; Rummel, K. T.; Scheuhammer, A. M.; Brisbin, I. L.; Hooper, M. J. Arch. Environ. Contam. Toxicol., 2000, 38, 107113. (14) Dickerson, R. L.; Hooper, M. J.; Gard, N. W.; Cobb, G. P.; Kendall, R. J. Environ. Health Perspect. 1994, 102, 65-69. (15) Elliot, J. L.; Kennedy, S. W.; Peakall, D. B.; Won., H. Comp. Biochem. Physiol. 1990, 96C, 205-210. (16) Parham, F. M.; Kohn, M. C.; Matthews, H. B.; Derosa, C.; Portier, C. J. Toxicol. Appl. Pharm. 1997, 144, 340-347. (17) Yawetz, A.; Agosin, M.; Perry, A. S. Comp. Biochem. Physiol. 1978, 59C, 45-49. (18) Brunstrom, B. Arch. Toxicol. 1990, 64, 188-192. (19) Brunstrom, B.; Andersson, L. Arch. Toxicol. 1988, 62, 263-266. (20) Bosveld, A. T. C.; Gradener, J.; Murk, A. J.; Brouwer, A.; Van Kampen, M.; Evers, E. H. G.; Van Den Berg, M. Environ. Toxicol. Chem. 1995, 14, 99-115. (21) Bosveld, A. T. C.; Gradener, J.; Van Kampen, M.; Murk, A. J.; Evers, E. H. G.; Van Den Berg, M. Chemosphere 1993, 27, 419427. (22) Elliot J. E.; Norstrom, R. J.; Lorenzen, A.; Hart, L. E.; Philibert, H.; Kennedy, S. W.; Stegeman, J. J.; Bellward, G. D.; Cheng, K. M. Environ. Toxicol. Chem. 1996, 15, 782-793.
Received for review May 14, 2002. Revised manuscript received October 21, 2002. Accepted November 12, 2002. ES0257873
