Environ. Sci. Technol. 2005, 39, 8083-8089
Toxicokinetics of 14C-Atrazine and Its Metabolites in Stage-66 Xenopus laevis ANDREA N. EDGINTON* Department of Environmental Biology, University of Guelph, Guelph, Ontario N1G 2W1, Canada CLAUDE ROULEAU Department of Fisheries and Oceans, Institut Maurice-Lamontagne, 850 route de la mer, C. P. 1000, Mont-Joli, Quebec G5H 3Z4, Canada
Several recent studies have focused on the toxicodynamic implications of amphibian exposure to the commonly used herbicide atrazine (2-chloro-4-ethylamine-6isopropylamino-s-triazine). These studies are an important part of the risk assessment process; however, the underlying mechanisms of atrazine toxicodynamics are lacking. In an attempt to more fully describe atrazine exposure, the toxicokinetics of atrazine were studied in stage-66 Xenopus laevis larvae. The absorption, distribution, and excretion capacity of these larvae were found to be comparable to those observed in fish. The calculated bioconcentration factor (BCF) was 1.5-1.6 mLwater/glarvae, and by use of whole-body autoradiography, the radiolabel was found to be concentrated in the gall bladder and gastrointestinal tract. Elimination of atrazine was rapid with a half-life of 48 min. The high metabolic capacity of stage66 X. laevis larvae was demonstrated where, following 8 h of exposure to 14C-atrazine, the percentages of atrazine and its metabolites deethyldeisopropylatrazine (DACT), deisopropylatrazine (DIA), and deethylatrazine (DEA) in larvae, determined by thin-layer chromatography, were 49.8% ( 3.3%, 9.8% ( 2.1%, 16.1% ( 2.5%, and 15.6% ( 2.0%, respectively. An unknown metabolite(s) was also produced and accounted for the remaining proportion of the total body radioactive residues. This metabolite(s) is hypothesized to be a conjugate of either atrazine or one of its metabolites. These metabolites, namely, DIA, were responsible for the long elimination half-life (72 h) of the total body radioactive residues. These toxicokinetics data would provide better insights in the interpretation of toxicodynamic data.
Introduction Atrazine (2-chloro-4-ethylamine-6-isopropylamino-s-triazine) is a widely used herbicide for the pre- and postemergent control of broadleaf weeds and grasses. It is a small (215.7 g/mol) molecule with moderate water solubility (33 mg/L) and lipophilicity (log Kow ) 2.7). Contamination of environmental matrices by atrazine and its metabolites has been reported in surface waters, groundwaters, and soil in North * Corresponding author address: Biophysics, Bldg E41, Bayer Technology Services, Leverkusen, Germany 51368; phone: 011 49 214 30 46034; fax 011 49 214 30 50698; e-mail:
[email protected]. 10.1021/es050295m CCC: $30.25 Published on Web 09/17/2005
2005 American Chemical Society
America. Atrazine concentrations can reach g20 µg/L in streams and ponds but are generally observed at less than 5 µg/L (1). The atrazine metabolites deethylatrazine (DEA) and deisopropylatrazine (DIA) have been found to range in concentration from 12% to 39% and from 4.9% to 24%, respectively, of the atrazine concentration in rivers, streams, and reservoirs (1). Hayes et al. (2) also reported concentrations of metabolites in reservoirs from which amphibians were sampled and were able to quantify DEA, DIA, and deethyldeisopropylatrazine (DACT) in four of seven sites. The toxicokinetics of atrazine in amphibians has not been previously reported; however, atrazine toxicodynamics have been the subject of several recent publications. Some studies have evaluated end points associated with unspecific mechanisms of action including growth, larval developmental time, and survival (3-5), while others have focused on end points specific to endocrine-mediated effects (2, 6-10) or genotoxicity (11). Atrazine concentrations shown to produce abnormal gonadal development, such as hermaphroditism (2, 8, 9), testicular dysgenesis (2, 7), and increased follicular atresia (6), have been inconsistent among studies. Debate has surfaced regarding the dependence of individual larval exposure on the density of larvae within the experimental unit (12). None of the above studies measured bioconcentration or metabolite concentrations in the larvae and all effects were attributed to the parent compound, atrazine. This approach may be appropriate for the purposes of exposure characterization and risk assessment, though for greater understanding, atrazine concentrations in water should be linked to the atrazine and metabolite profiles within the amphibians. The objective of the following studies was to define the toxicokinetics of atrazine in stage-66 Xenopus laevis, a species commonly used in endocrine disruption studies as a representative amphibian, to better define absorption, distribution, metabolism, and excretion of atrazine and its metabolites. These toxicokinetics data may provide better insights in the interpretation of toxicodynamic data.
Materials and Methods Test Substances. Analytical-grade atrazine and the associated metabolites 4-(ethylamino)-2-hydroxy-6-isopropylamino-striazine (G34048, HA, 97% purity), 2-chloro-4,6-diamino-striazine (G28273, DACT, 97% purity), 2-chloro-4-amino-6(ethylamino)-s-triazine (G28279, DIA, 96% purity), 4-amino2-chloro-6-isopropylamino-s-triazine (G30033, DEA, 94% purity), 4-amino-6-[(1-methylethyl)amino]-1,3,5-triazin2(1H)-one (GS17794), and 2-amino-4-(ethylamino)-6-hydroxy-s-triazine (GS17792), were obtained from Syngenta (Syngenta Crop Protection, Greensboro, NC). Radiolabeled atrazine (G30027) [triazinyl-U-14C]atrazine, was also provided by Syngenta. It had a radiochemical purity of 96.1%, a chemical purity of 96.3%, and a specific activity of 9.04 mCi/ mmol. All solvents that were used were HPLC grade or better. Larval Culturing. Stage-66 X. laevis larvae that were used for preliminary testing and method development were University of Guelph, Guelph, ON, Canada stock. Parental X. laevis were induced to breed by the injection of 600 and 800 IU of human chorionic gonadotrophin (Sigma-Aldrich Canada, Oakville, ON, Canada) into the dorsal lymph sac of males and females, respectively. Amplexus, egg laying, and fertilization occurred within 12 h. Embryos and larvae were cultured in flow-through tanks at 20 °C and fed ground spirulina flakes daily. When they reached stage-66, in about 3 months, the larvae were moved to another tank and fed chironimids every other day. For the definitive studies, stageVOL. 39, NO. 20, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8083
TABLE 1. Thin-Layer Chromatographic Analyses of Atrazine and Its Metabolites TLC Rf value chemical
solvent system 1
solvent system 2
atrazine DEA DIA DACT HA 4-amino-6-[(1-methylethyl)amino]-1,3,5-triazin-2(1H)-one 2-amino-4-(ethylamino)-6-hydroxy-s-triazine
0.85 0.55 0.42 0.18 0 0 0
1 1 1 1 0.60 0.32 0.13
66 larvae were obtained from Xenopus I (Xenopus I, Dexter, MI). Prior to use, they were acclimated in FETAX culture water for 1 week and fed chironimids every other day. Water used in all culturing tanks, controls, and treatments conformed to the American Society of Testing and Materials guideline for the performance of the Frog Embryo Teratogenesis Assay - Xenopus (13). The pH of the FETAX culture water was adjusted to 7.2-7.4 with 1 N HCl. Larvae were not fed 12 h prior to the beginning of any study and were not fed during the studies. Kinetic Study. Experimental units used in the exposure phase consisted of 10 glass beakers containing 100 mL of FETAX culture water spiked with 1.9 nmol/mL 14C-atrazine. Six larvae were added to each experimental unit. The 10 experimental units represented five time points for the exposure phase (1, 2, 4, 6, and 8 h) with two replicates. Two additional beakers contained no larvae and were used to quantify 14C-atrazine removal from the FETAX culture water due to chemical degradation and/or volatilization during the 8 h exposure phase. To quantify elimination, two experimental units contained 42 larvae at the same larval/ FETAX culture water/chemical ratio as in the exposure phase. Following 8 h of exposure, these larvae were moved to two 40-L aquaria each containing 20 L of clean FETAX culture water. Sample time points for the elimination phase were 1, 2, 4, 14, 40, and 64 h posttransfer. Half of the volume of the aquarium was replaced with fresh FETAX culture water at each time point during the elimination phase to ensure the maintenance of good water quality and to dilute any excreted radioactivity. At each time point, the concentration of atrazine and its metabolites were assessed. Larvae were removed from the experimental unit, rinsed for 5 s in deionized water, and submerged in 1000 mg/L tricaine methanesulfonate (MS222; Sigma-Aldrich Canada, Oakville, ON, Canada) for 5 min. Larvae were then transferred to an ice bath for 2 min to ensure complete euthanasia, which was based on the cessation of the heartbeat. To determine total body radioactive residues, three of the six larvae from each experimental unit were weighed and freeze-dried for at least 17 h. Freezedried larvae were processed in a Harvey biological oxidizer and counted in a Beckman LS6K-SC scintillation counter (Beckman Instruments Inc., Fullerton, CA). The mean recovery ( SEM for spiked larvae samples in the biological oxidizer was 98.3% ( 2.1% (n ) 5). The other three larvae were homogenized individually by use of a Polytron (Brinkmann Instruments, Mississauga, ON, Canada) tissue homogenizer in 10 mL of methanol/dimethylformamide/water (100: 0.5: 0.5). The homogenate was shaken for 30 min and centrifuged at 6000g for 30 min, and the supernatant was collected. Following one additional 10 mL methanol/dimethylformamide/water extraction, the pooled supernatant was evaporated to near dryness under vacuum at 45 °C by use of a rotary evaporator. The contents were brought to 1 mL with the extraction solvent. A 100 µL sample was taken and counted by liquid scintillation counting (LSC) to generate another measure of total body radioactive residues. The mean 8084
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 20, 2005
recovery ( SEM for spiked larval samples following extraction was 101.0% ( 3.0% (n ) 8) for 14C-atrazine. Thin-layer chromatography (TLC) was used to separate 14 C-atrazine from its radioactive metabolites. Fifty and 100 µL aliquots of the larvae extracts and FETAX culture water, respectively, were loaded onto aluminum-backed silica gel TLC plates (Kieselgel 60 F254, Merck, NJ) with ethyl acetate/ hexane/acetic acid (10:10:0.05) (solvent system 1) or chloroform/ethanol/acetic acid (100:10:10) (solvent system 2) as mobile phases. Standards of nonlabeled atrazine and its metabolites were run on all plates for comparison, and Rf values were determined as shown in Table 1. Areas corresponding to atrazine and standard metabolites were cut out and placed in 7 mL glass scintillation vials containing CytoScinct ES (ICN Biomedicals, Irvine, CA). The presence of the plate did not alter the counting efficiency of LSC. Qualitative recovery of the six metabolites was assessed after plating homogenates that had been spiked with metabolites prior to homogenization. Spots were compared to metabolite standards under UV light and all metabolites could be recovered. Metabolism Studies. Twelve larvae were exposed to 1.92.0 nmol/mL 14C-atrazine. Two exposure units consisted of 100 mL of FETAX culture water containing six larvae each. Larvae were exposed for 24 h and aeration was provided throughout the exposure period. All larvae and FETAX culture water samples were processed for TLC as described above. Whole-Body Autoradiography. Whole-body autoradiography, as described by Ullberg et al. (14), was used to determine the internal distribution of the radiolabel. Fifteen stage-66 larvae were exposed to 0.53 nmol/mL 14C-atrazine for 8 h in 250 mL of FETAX culture water. At 8 h, larvae were removed from the exposure units, placed in clean FETAX culture water for 20 min, removed, rinsed for 5 s in deionized water, and euthanized as described above. Twelve larvae were embedded in a 5% carboxymethylcellulose gel and rapidly frozen in a -80 °C hexane/dry ice slurry. Frozen blocks were sectioned on tape (20 µm sections) at -25 °C by use of a specially designed cryomicrotome (Jung Cryomacrocut, Leica, San Diego, CA). Sections were freeze-dried at -25 °C for 16 h and placed on flexible storage phosphor screens for 330 h. Digital scans of the phosphor screens were produced with a Cyclone storage phosphor system (Packard BioScience Co., IL) and visualized and quantified with Optiquant software (Optiquant v 3.1, Packard Instrument Co.). The distribution of the radiolabel was expressed as a concentration index relative to the whole body (IC) (15):
IC )
(DLU/mm2)tissue (DLU/mm2)wholebody
DLU/mm2 represents the digital light units per square millimeter of the section surface. False color images were produced with the software Dplot (www.dplot.com). Three larvae were counted by LSC, as previously described, to determine total body radioactive residues.
At 8 h, the atrazine concentration was set to 0. Minimization was performed by the Gauss-Newton (Levenberg and Hartley) process with a maximum of 50 iterations and a convergence criterion of 0.0001. All reported values were expressed as mean ( SEM.
Results
FIGURE 1. Hematoxylin- and eosin-stained 4 µm longitudinal section of a stage-66 Xenopus laevis larva. The lower photograph is the abdominal section at a higher resolution. This and similar slides were used to identify organs by location, size, cell type, and shape. Organs in the scanned sections were identified by comparison to 4 µm longitudinal sections that had been previously prepared for paraffin sectioning and stained with hematoxylin and eosin (Figure 1). Deterministic factors were location, cell type, organ contents, shape, and size (16). Data Analyses. The time course of 14C-atrazine and metabolite concentrations were modeled, from 8 to 72 h, by monophasic and biphasic exponential decay equations. Model choice was dependent on visual inspection of the observed vs predicted graphs and reasonable coefficients of variation of the parameter estimates. Atrazine and DACT elimination were best described by a monophasic decay equation and DEA, DIA, and the origin on the TLC plate were best described by a biphasic decay equation. For the generation of an atrazine water clearance rate, the following model was used to describe the larval atrazine concentrations over time with the nonlinear regression software WinNonlin (v 3.1; Pharsight, Mountain View, CA): from 0 to 8 h
dCATR,LAR ) (CATR,WATERCLU) - (CATR,LARKE) dt and from 8 to 72 h
dCATR,LAR ) -(CATR,LARKE) dt CATR,LAR was the concentration of atrazine in the larvae (nanomoles/gram), CATR,WATER was the concentration of atrazine in the water (nanomoles/milliliter), CLU was the atrazine water clearance rate (milliliterswater hour-1 gramtissue-1), and KE was the atrazine elimination rate (hours-1), inclusive of both the excretion and metabolism rates. Any change in the water concentration of atrazine during the exposure phase was also determined and included in the above model. At each iteration of the model, the water concentration of atrazine was recalculated and a new, time-sensitive atrazine concentration was produced. This accounted for the actual exposure of the larvae to 14C-atrazine over the 8 h of exposure.
Metabolism Study. Recovery of the samples with solvent system 1 was 101.0% ( 3.1%, which included atrazine, DACT, DIA, DEA, and the origin on the TLC plate. Following 24 h of exposure, atrazine, DACT, DIA, and DEA accounted for 36.7% ( 1.7%, 13.6% ( 0.80%, 15.5% ( 0.88%, and 13.6% ( 1.2%, respectively, of the total body radioactive residues in the 12 larvae (weight 419.4 ( 33.4 mg). An atrazine BCF of 1.6 ( 0.081 mLwater/gtissue (n ) 10) was derived at 24 h by dividing the concentration of atrazine in the larvae (nanomoles/gram) by the concentration in the water (nanomoles/ milliliter). The BCF for the total body radioactive residues was calculated at 4.4 ( 0.22 mLwater/gtissue. At the end of 24 h, the concentration of 14C-atrazine in the exposure water was 52% of the initial concentration. None of the standard metabolites that were separated under solvent system 2 (4-amino-6-[(1-methylethyl)amino]1,3,5-triazin-2(1H)-one, 2-amino-4-(ethylamino)-6-hydroxys-triazine, or HA) were produced by the larvae. There was 13.2% ( 0.84% more radioactivity found on the origin of solvent system 1 than on solvent system 2. The origin on solvent system 2 accounted for only 7.4% ( 0.40% of the total spotted sample, whereas it accounted for 20.5% ( 1.0% with solvent system 1. This suggests that another compound, a likely metabolite, was stationary in solvent system 1 and was mobile in solvent system 2. It may also suggest that the stationary radioactivity (7.4%) on solvent system 2 was a second unknown metabolite. Alternatively, it may have represented radiolabel that was bound to organic matter in the spotted sample. Kinetic Study. FETAX culture water controls containing no larvae were used to determine if degradation occurred in the exposure water. Previous reports have demonstrated slow degradation with half-lives in water of at least 285 days (17, 18). Control FETAX culture water contained an average of 95.9% ( 0.24% (n ) 12; six time points, two repetitions) 14Catrazine over all of the exposure time points. The chemical purity of 14C-atrazine was 96.3%; thus chemical degradation of 14C-atrazine did not occur in this time frame. The remainder of the radioactivity (3%) was DEA, suggesting that the stock solution minimally contained this metabolite. The percentage of DEA did not change over the 8 h period and no metabolites were formed in the FETAX culture water controls. The FETAX culture water was sampled at each time point in the exposure units containing larvae, and at 8 h, 88% of the initial 14C-atrazine water concentration remained. 14CAtrazine loss was fit to a linear-, monophasic exponential and a biphasic exponential decay. Based on the visual fit of the data, the biphasic decay was the best descriptor for atrazine loss in FETAX culture water, where
[atrazine (nanomoles/milliliter)] ) 0.27e-2.5t + 1.55e-0.013t (r2 ) 0.90) The biphasic decay adequately described this process and that also of the metabolism study (same larvae loading and similar concentrations used): observed at 24 h, 1.07 nmol/ mL; predicted by equation, 1.14 nmol/mL. No growth or weight loss of the larvae was observed during the study. The mean weight of the 119 larvae was 437.7 ( 11.4 mg. The atrazine water clearance rate was calculated at 1.33 ( 0.16 mLwater h-1 gtissue. The elimination rate of atrazine was calculated at 0.86 ( 0.11 h-1 and represented a half-life of elimination of 48 ( 6.1 min. Figure 2 represents the VOL. 39, NO. 20, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8085
FIGURE 2. Concentration-time data for total body radioactive residues and atrazine in stage-66 Xenopus laevis larvae. The predicted atrazine curve was generated from a user-defined WinnonLin model. Larvae were exposed to 14C-atrazine for 8 h, followed by a 64 h elimination phase in clean FETAX culture water. Observed values are mean ( SEM (n ) 6 except at 9 h, where n ) 3).
FIGURE 3. Concentration-time curves of DACT, DIA, DEA, and an unknown metabolite(s), produced by stage-66 Xenopus laevis larvae. Larvae were exposed to 14C-atrazine for 8 h, followed by a 64 h elimination phase in clean FETAX culture water. Observed values are mean ( SEM (n ) 6 except at 9 h, where n ) 3). The elimination phase was modeled for each metabolite by a monophasic (CATR,LAR ) C(t)0) exp-βt) or biphasic (CATR,LAR ) A exp-rt + B exp-βt) exponential decay model. The predicted curve for the most appropriate model is shown, as well as half-lives (t1/2) of the first and terminal elimination phases. concentration-time data for total body radioactive residues and atrazine. Notice that total body radioactive residue and atrazine concentrations at the 8 h time point are lower than those at the 6 h time point. This potentially suggests that steady state was reached. Figure 3 represents the concentration-time curves for each of the observed metabolites. To summarize, following 8 h of exposure, the percentages of atrazine, DACT, DIA, and DEA in larvae were 49.8% ( 3.3%, 9.8% ( 2.1%, 16.1% ( 2.5%, and 15.6% ( 2.0%, respectively. In addition, 13.5% ( 2.3% did not migrate from the origin of the TLC plate in solvent system 1. It was apparent from Figure 3, Unknown Metabolite(s), that the amount that remained on the origin changed over time and was likely at least one other metabolite for which a standard was not available. On the basis of the movement of this compound 8086
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 20, 2005
with the solvent front in the more polar solvent system 2, it was deduced to be more polar than 4-amino-6-[(1-methylethyl)amino]-1,3,5-triazin-2(1H)-one, 2-amino-4-(ethylamino)-6-hydroxy-s-triazine, or HA but less polar than atrazine, DACT, DIA, and DEA. Whole-Body Autoradiography. Serial sections (5-12, median 8), which were 140 µm apart, were examined for atrazine distribution in each of 12 larvae. Representative pictures of larval sections and their corresponding phosphor screen images are shown in Figure 4. Atrazine accumulation, with a concentration index (IC) greater than 1, occurred in the gall bladder, stomach, intestines, and liver (Figure 4). The IC value (mean ( SEM) for the gall bladder was 6.9 ( 0.65 (n ) 12 larvae; min ) 4.0, max ) 10.2), for the stomach was 2.4 ( 0.35 (n ) 11 larvae; min ) 0.80, max ) 4.5), for the
FIGURE 4. Scanned sections of stage-66 Xenopus laevis larvae (A) and their subsequent phosphor screen images (B). The greatest radiolabel concentrations correspond to red areas. Units on the color scale are digital light units (DLU)/mm2. Units on axis are in millimeters. Significant accumulation can be seen in the gall bladder and gastrointestinal tract. VOL. 39, NO. 20, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8087
intestines was 3.2 ( 0.17 (n ) 11 larvae; min ) 2.1, max ) 4.2), and for the liver was 1.3 ( 0.11 (n ) 10 larvae; min ) 0.83, max ) 1.9). Other organs, including the heart, kidneygonad complex, eyes, brain, pancreas, and muscle, did not accumulate the radiolabel above unity. Total radioactivity per larvae, as determined by use of the biological oxidizer and LSC, was 0.95 ( 0.16 nmol/g (n ) 3).
Discussion The atrazine BCF in this study was 1.6 mLwater/glarvae, calculated following 24 h of exposure in the metabolism study, and 1.5 mLwater/glarvae, calculated on the basis of the equation CLU/KE adopted by Landrum et al. (19). These were within the BCF range reported for fish (
