Fish Embryo Toxicity Test: Identification of Compounds with Weak

May 4, 2015 - Published fish embryo LC50 values were confirmed by experimental analysis of zebrafish embryo LC50 according to OECD guideline 236. Exce...
0 downloads 7 Views 2MB Size
Article pubs.acs.org/est

Fish Embryo Toxicity Test: Identification of Compounds with Weak Toxicity and Analysis of Behavioral Effects To Improve Prediction of Acute Toxicity for Neurotoxic Compounds Nils Klüver,*,† Maria König,†,§ Julia Ortmann,† Riccardo Massei,†,§ Albrecht Paschke,‡ Ralph Kühne,‡ and Stefan Scholz† †

Department of Bioanalytical Ecotoxicology, UFZ-Helmholtz Centre for Environmental Research, Permoserstr. 15, 04318 Leipzig, Germany ‡ Department of Ecological Chemistry, UFZ-Helmholtz Centre for Environmental Research, Permoserstr. 15, 04318 Leipzig, Germany S Supporting Information *

ABSTRACT: The fish embryo toxicity test has been proposed as an alternative for the acute fish toxicity test, but concerns have been raised for its predictivity given that a few compounds have been shown to exhibit a weak acute toxicity in the fish embryo. In order to better define the applicability domain and improve the predictive capacity of the fish embryo test, we performed a systematic analysis of existing fish embryo and acute fish toxicity data. A correlation analysis of a total of 153 compounds identified 28 compounds with a weaker or no toxicity in the fish embryo test. Eleven of these compounds exhibited a neurotoxic mode of action. We selected a subset of eight compounds with weaker or no embryo toxicity (cyanazine, picloram, aldicarb, azinphos-methyl, dieldrin, diquat dibromide, endosulfan, and esfenvalerate) to study toxicokinetics and a neurotoxic mode of action as potential reasons for the deviating fish embryo toxicity. Published fish embryo LC50 values were confirmed by experimental analysis of zebrafish embryo LC50 according to OECD guideline 236. Except for diquat dibromide, internal concentration analysis did not indicate a potential relation of the low sensitivity of fish embryos to a limited uptake of the compounds. Analysis of locomotor activity of diquat dibromide and the neurotoxic compounds in 98 hpf embryos (exposed for 96 h) indicated a specific effect on behavior (embryonic movement) for the neurotoxic compounds. The EC50s of behavior for neurotoxic compounds were close to the acute fish toxicity LC50. Our data provided the first evidence that the applicability domain of the fish embryo test (LC50s determination) may exclude neurotoxic compounds. However, neurotoxic compounds could be identified by changes in embryonic locomotion. Although a quantitative prediction of acute fish toxicity LC50 using behavioral assays in fish embryos may not yet be possible, the identification of neurotoxicity could trigger the conduction of a conventional fish acute toxicity test or application of assessment factors while considering the very good fish embryo−acute fish toxicity correlation for other compounds.



ment 126),7 sequential testing or by predicting acute fish toxicity using fish cell lines and fish embryos.1 Particularly, the use of fish embryos has been discussed as a promising alternative to predict acute fish toxicity, since fish embryo mortality shows a high correlation to the AFT with similar sensitivity.4,8−10 Fish embryos until the onset of independent feeding are considered as nonprotected life stages by the current European Directive Directive 2010/63/EU on the protection of animals used for scientific purposes and are therefore considered as an alternative to the testing of (juvenile or adult) animals.11,12 Recently, a guideline for the 96 h fish embryo acute toxicity test (FET) has been adopted by the

INTRODUCTION Acute fish toxicity data are required for the registration and authorization of industrial chemicals, plant protection products, biocides, veterinary pharmaceuticals, and feed additives and in several countries and regions the assessment of effluent toxicity.1 The acute fish toxicity test (AFT) is usually conducted according to the OECD testing guideline (TG) 2032 and represents the most commonly conducted vertebrate test for environmental hazard and risk assessment. The end point assessed in the AFT is mortality resulting in severe suffering and distress of the test animals.3,4 Up to 7−10 fish per concentration are exposed to estimate the 50% lethal concentration (LC50). Various approaches to reduce or replace the number of test animals for the acute fish toxicity test have been proposed, such as changes in the protocol,5,6 the application of a threshold approach (OECD guidance docu© 2015 American Chemical Society

Received: April 15, 2015 Accepted: May 4, 2015 Published: May 4, 2015 7002

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology OECD (OECD TG 236).13 Currently, the most frequently used embryos are those of the zebrafish,8−10 and the OECD validation study with zebrafish embryos has indicated a high inter- and intralaboratory reproducibility of the acute toxicity testing.14 Zebrafish has a high fecundity and is easy to maintain in the lab. Their embryos are optically transparent and develop rapidly, and the stage of independent feeding is reached at around 5 days post fertilization (dpf). They also offer the possibility to perform large-scale, high-throughput analyses and are already used as a screening model in various applications.15 Given the availability of an OECD guideline (TG 236), this now provides an opportunity to reduce acute fish toxicity testing for regulatory purposes.8,10,14 For instance, in the context of the European Regulation on industrial chemicals (REACH), the FET could be considered as a suitable alternative by REACH registrants in their testing proposals for aquatic toxicity.16 Nevertheless, concerns of limitations in the application domain and/or predictive capacity for specific compounds have been raised. For instance, anecdotal evidence has been provided that a low toxicity in the FET, if compared to the AFT, might be provoked by limitations in the metabolic capacity.9,17 Furthermore, a reduced uptake of compounds, if compared to adult fish, could reduce the sensitivity of fish embryos. For instance, it was reported that highly hydrophobic compounds did not reach equilibrium internal concentrations in early life stages18 or that incorporation of highly hydrophilic compounds could not be detected until 7 dpf.19 Instability of exposure concentrations could also result in a weaker sensitivity of fish embryos. So far, most of the fish embryo tests have been conducted with a static or semistatic exposure setup without analytical verification of exposure concentration, in contrast to many acute fish toxicity tests conducted for regulatory purposes. Finally, the first evidence has been provided that fish embryos show a weak sensitivity for neurotoxic compounds with respect to acute toxicity.9 The systematic analysis of outliers, i.e., compounds with a lower acute toxicity in embryos than in the AFT, and the identification of mechanisms leading to a deviating toxicity is crucial to define and/or improve the application domain, reduce the probability for false negatives, and increase acceptance of the FET for regulatory applications. Therefore, in this study, we made use of a previously established FET database.20 From an initial list of 26 compounds with a potentially lower toxicity in the FET, we prioritized 8 compounds with different physicochemical properties and different modes of action (MoA) for further analysis using a strategy outlined in Figure 1. For the 8 prioritized outliers that were experimentally verified using the zebrafish FET, we investigated (i) stability of exposure concentrations, (ii) a potential limited uptake, and/or (iii) a specific MoA as potential reasons for the deviation of fish embryo from juvenile or adult fish LC 50 . Therefore, exposure and internal concentrations were determined using chemical analyses. Furthermore, behavioral changes were assessed by tracking embryonic movement as an additional end point to predict or identify neurotoxic compounds.

Figure 1. Strategy to identify and evaluate compounds with a lower acute toxicity in the fish embryo test if compared to adult fish. (A−E) Boxes include information on the number of compounds used or identified at that particular stage.

to identify corresponding AFT data that were matching certain quality criteria (e.g., nominal exposure concentrations analytically confirmed, water quality parameters measured; see the Supporting Information for details). For the AFT, data of the five most commonly used species (fathead minnow, medaka, bluegill sunfish, zebrafish, or rainbow trout) were considered. In case that an LC50 was available for more than one species, the lowest value was used for the regression analysis with the FET. If FET effect concentrations were available for more than one study, all values were included individually in the regression analysis. However, if an individual study investigated different exposure windows or durations, only the lowest LC50 was considered. Regression analysis was conducted using a Deming (type II) regression in order to consider variability for both the fish embryo and juvenile or adult acute toxicity. The regression analysis was performed using the Software Sigma Plot 12.0 (Systat Software GmbH, Erkrath, Germany). Outliers were identified using a box plot analysis for the residuals of the regression analysis with the software IBM SPSS (IBM, Ehningen, Germany). Outliers represented values with a more than 1.5-fold of the 25−75% percentile distance below or above the 25% percentile (lower whisker) or 75% percentile (upper whisker). Subsequently, the regression analysis was repeated excluding the previously identified outliers. Chemicals and Stock Solutions. All tested chemicals and solvents were ordered from Sigma-Aldrich (Deisenhofen, Germany) at the highest available purity (≥97%). Stock solutions in embryo medium (see the Supporting Information)



MATERIALS AND METHODS Data Selection and FET-AFT Regression Analysis. The comparison of FET and AFT data was limited to organic compounds given the principally unlimited number of new organic compounds that will require toxicity assessment in the future. An existing data set20 of fish embryo mortality was used 7003

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology

analysis, we did not exclude embryos with malformations (e.g., axis deformations). Embryonic movement was tracked using the ZebraBox video tracking system (Viewpoint, Lyon, France) for 20 min at a temperature of 26 ± 1 °C. For details, see the Supporting Information.

were prepared the day prior to initiation of an exposure experiment in closed graduated flasks, sealed with a glass lid and Parafilm (VWR, Darmstadt, Germany). The solutions were stirred overnight and protected from light. For dieldrin, endosulfan, and esfenvalerate, stock solutions were prepared in acetone with 0.01% solvent concentrations in exposure media. Controls conducted for these compounds also contained 0.01% acetone. For all test solutions, pH was measured. The pH generally ranged from pH 6.5 to 8.5 and was only adjusted for picloram to pH 6.8 using 0.2 M NaOH. Oxygen levels of test solutions did not fall below 80% saturation during the exposure. Zebrafish Embryo Exposure. Fish maintenance, egg production, and range-finding tests were conducted according to Knöbel et al.9 Exposure of fish embryos was performed for 96 h (2−98 h post fertilization [hpf]) as outlined in the OECD TG 236 for the Fish Embryo Acute Toxicity (FET) Test.13 Standard FET tests were performed in 24-well plates, with one embryo per well and 2 mL of exposure volume. In contrast to the OECD TG 236, only ten embryos were used per concentrations, but the tests were repeated at least twice with adjusted concentration and dilution ranges to improve modeling of concentration−response curves. In all exposure solutions, the pH and oxygen levels during the testing period were in the range of acceptance criteria for the OECD TG 236. Analytical measurements of exposure concentrations were performed at the beginning and end of the test for aldicarb, azinphos-methyl, dieldrin, diquat dibromide, and esfenvalerate. On the basis of the analytical data we have determined, FET tests with esfenvalerate, endosulfan, and dieldrin were conducted using a semistatic design with 24 h renewal intervals. Details of the chemical analysis and sampling of embryos for internal concentration analysis are described in the Supporting Information. Internal Concentration Modeling and Bioconcentration Factor (BCF) Estimation. Model fits of the internal concentration were based on a one-compartment model (Cint = (k1/k2) × Cw × (1 − e(−k2 × t)).21 Cint refers to the internal concentration of the organism; Cw to the exposure concentration; k1 to the uptake rate; k2 to the depuration rate; t represents the corresponding time point. The (dimensionless) BCF of embryos was estimated on the basis of the steady-state concentration and assuming an embryo volume of 0.2 μL. The BCFs of diquat dibromide and dieldrin were based on the internal concentration after 96 h of exposure. If no experimental values for BCFs of adult fish were available, they were estimated using the EPI Suite22 BCFBAF module (Version 4.1; log BCF = 0.6598 log Kow − 0.333). Behavioral Changes Determined by Locomotor Response (LMR). For the LMR, 25 embryos per concentration were exposed in 50 mL exposure media in glass crystallization dishes covered with watchmaker glasses. At 96 hpf, before analysis of locomotion, embryos were transferred to a 96 well plate with rectangular wells (one embryo per well, with 630 μL). To avoid any interference of behavioral analysis caused by the position of the embryo within the plate, a randomized plate layout for the distribution of treatments was used. On each plate, one control treatment and three treatments with different exposure concentrations were distributed (24 embryos per treatment). In case of the three highest concentrations for diquat dibromide (350, 425, and 500 mg/L), nonhatched embryos were manually dechorionated with forceps prior to locomotion analysis. For the LMR



RESULTS Identification of FET Outliers. FET effect concentrations were obtained from a previously established data set containing 641 compounds.20 The chemicals in the FET database (Figure 1, section A) and the subset with corresponding AFT data (Figure 1, section B) used in this study were compared with the European Inventory of Existing Commercial Chemical Substances (EINECS). With regard to chemical composition in terms of occurrence, proportion, and combination of functional groups such as amide, benzene ring, thiol groups, and polarity distributions, a similar pattern was observed in our data sets if compared to the EINECS list (Table S1, Supporting Information). The data set used for FET-AFT analysis comprised 225 FET and 604 AFT database entries (Tables S2 and S3, Supporting Information). The FET was mainly performed with zebrafish embryos (219 values; 97%). The analyzed compounds covered a range of molecular weights from 30 to 552 g/mol, of log Kows from −4.6 to 7.6 and of water solubilities from 0.24 μg/L to 1000 g/L (Table S2, Supporting Information). The data set used for the comparative analysis covered 153 compounds, of which 22 compounds did not provoke mortality in the FET in the tested range of concentrations. These 21 compounds were not included in the linear regression analysis but were considered as compounds with a potential weaker sensitivity in the FET (Figure 1, section C). The regression analysis of FET and AFT-LC50s revealed a regression slope of 1.13 (significantly different from 1) and an intercept of −0.95 with a correlation coefficient of 0.91 (Figure 2). Box plot analysis revealed ten compounds (allyl alcohol, azinphosmethyl, chlorothalonil, chlorpyrifos-ethyl, cyclohexane, dieldrin, endosulfan, N-methylanilin, technical nonylphenol, and per-

Figure 2. Regression analysis of fish embryo test (FET) vs acute fish toxicity (AFT) LC50. FET data represent 202 database entries referring to 132 different compounds. AFT data represented the lowest available LC50. Compounds with no toxicity in the FET in the tested range of the concentrations have not been included in the regression analysis. Closed circles represent geometric means of AFT and corresponding FET data for each chemical. Open circles correspond to FET outliers identified by a box blot analysis. The black solid line represents the regression line obtained by exclusion of outliers. The dotted black line represents the line of unity. APM, azinphos-methyl; CPY, chlorpyrifosethyl. 7004

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

7005

660 0.424 255 3.88 279.9 [95% CI: 252.80−307.30] 1.096 8−144 no tox >1000 1.13 aldicarb

60571

116063

a

dieldrin

66230044

Experimental data determined within this study are shown in bold font. MoA information was derived from a literature search, and classification of MoAs was conducted mainly on the basis of primary, application-related bioactivity. CI, confidence interval; n.d., not determined; n.a., not applicable. bMoA not applicable in fish. cNo stable exposure concentrations. dNo mortality in FET and therefore the ratio of (limit of water solubility)/EC50 was used.

1.87d − 5.4

0.195

7.99

6−96

0.00225

no mortality



0.104

14.29d − −

azinphosmethyl esfenvalerate 86500

neurotoxic (voltage gated Na channels)29 neurotoxic (inhibit GABA receptors)52 neurotoxic (AChE inhibition)45

6.22

0.002−0.006

no tox

0−48

0.00002

no mortalityc

0.00014

7.21 0.4701 328 3.02 3.39 [95% CI: 2.83−4.07] 0.00909 0−48 3.47 20 2.75

56.84d − 3.83

0.324

3.63

8−144

0.00062

no mortality



0.006

− − n.a. n.d. n.d. n.a. 19 7 31 3.82 3.94 3.19

115297

LC50 mg/L

75.91 [95% CI: 64.14−90.54] 112.56 [95% CI: lower 90.09] 555.29 [95% CI: 459.41−657.82] 3.98 15.49 18.2 8−144 8−144 8−144 no tox no tox no tox 169 426 >1000 2.22 0.3 −4.6

herbicide PS II inhibitorb,49 auxinic herbicideb,50 non selective (herbicide) - redox cycling51 neurotoxic (GABA-gated chloride channel antagonist)23 neurotoxic (AChE inhibitor)17

MoA compound

cyanazine picloram diquat dibromide endosulfan

CAS#

21725462 1918021 85007

EC50 mg/L FET exposure/ hpf FETLC50 mg/L water solubility mg/L log Kow

Table 1. Summary of FET-LC50 and LMR-EC50 Values Determined after 96 h Exposurea

AFTLC50 mg/L

ZFET (0−96 hpf)

Hillslope

ZFET/ AFT ratio

ZFET behavior (LMR)

methrin) with FET toxicities deviating from the FET-AFT regression (Figure S1, Supporting Information). The FET/ AFT ratios for these compounds were in the range of 134− 5830 (Table S2, Supporting Information). However, for four compounds (chlorothalonil, N-methylanilin, technical nonylphenol, and permethrin), there were also studies with a lower LC50 for fish embryos with only 0.68- to 58-fold deviation from the FET-AFT regression. Since it was questionable whether these compounds represent true outliers from the FET-AFT correlation, they were not prioritized for subsequent experimental analyses. Thus, we identified 28 potential FET outliers (Figure 1, section D and Table S4, Supporting Information). The outliers covered a similar range of compound properties if compared to the entire data set; i.e., they exhibited a molecular weight range of 58 to 451 g/mol, a water solubility range of 2 × 10−3 to 1.00 × 106 mg/L, and a log Kow range of −4.6 to 6.22 (Figure S2, Supporting Information). Interestingly, 11 of these 28 outliers (39%) represented neurotoxic compounds. For 7 of these 11 outliers with neurotoxic MoA, no lethality has been observed in the FET (aldicarb, cyfluthrin, endrin, esfenvalerate, flucythrinate, oxamyl, and resmethrin). Four compounds exhibited higher FET-LC50s if compared to AFT (azinphos-methyl, chlorpyifos-ethyl, dieldrin, and endosulfan). These neurotoxic outliers comprised compounds of different mechanisms of action such as acetylcholine esterase inhibition (aldicarb, azinphos-methyl, chlorpyrifos-ethyl, oxamyl), blocking of voltage-gated sodium channels (cyfluthrin, esfenvalerate, flucythrinate, resmethrin), or interference with GABA-gated chloride channels (dieldrin, endosulfan, endrin). For nonoutliers (148 compounds) of the FET-AFT regression, 13 compounds had a neurotoxic mechanism of action (9%). FET Outlier Prioritization and Determination of FETLC50s. For the experimental validation and characterization, outliers were prioritized on the basis of two criteria: 1. Compounds with no mortality in the FET but that were tested only in a concentration range below the baseline toxicity and where AFT LC50 values were in the range of the predicted baseline toxicity were not prioritized for subsequent analysis. It was considered as likely that these outliers would not be confirmed if the range of concentrations were extended to the baseline concentration range. 2. FET outliers selected for subsequent experimental analyses should represent different compound classes (according to ECOSAR) and different modes of action (MoA). After applying the two criteria, 16 compounds (4 compounds with a weak embryo toxicity and 12 compounds with no FET mortality) were prioritized (Figure 1, section E). Subsequently, eight compounds (aldicarb, azinphos-methyl, cyanazine, dieldrin, diquat dibromide, endosulfan, esfenvalerate, and picloram) reflecting the different physicochemical properties and MoAs of the outliers were selected for further experimental analysis (Table 1). A zebrafish FET test was conducted for these compounds and confirmed the weaker sensitivity for these compounds, albeit with partially lower LC50 as reported in earlier studies. For cyanazine, picloram, azinphos-methyl, aldicarb, and diquat dibromide, we obtained LC50 values deviating 7-, 19-, 31-, 255-, and 328-fold, respectively, from the AFT (Table 1 and Figure S3, Supporting Information).

ZFET LC50/ EC50

Environmental Science & Technology

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology

Figure 3. Internal concentration−time profiles for zebrafish embryos exposed for 96 h (2−98 h post fertilization) to sublethal concentrations. (A) aldicarb (106 mg/L measured exposure concentration; n = 2; LC10 = 159 mg/L), (B) azinphos-methyl (0.66 mg/L measured exposure concentration; n = 1; LC10 = 1.64 mg/L), (C) dieldrin (0.2 mg/L nominal exposure concentration; n = 3), (D) diquat dibromide (353 mg/L, measured exposure concentration; n = 1; LC10 = 280 mg/L), and (E) esfenvalerate (0.55 μg/L, measured exposure concentration; n = 4). ∗: The BCF of adult fish [L/kg wet weight] was calculated with EPI Suite (0.6598 log Kow − 0.333). ∗∗: The BCF of embryos was estimated on the basis of the internal concentration after 96 h of exposure or, if available, using steady-state concentration. Model fit of the internal concentration is represented by solid lines, based on a one-compartment model.

weight-based BCFs observed for adult fish (Figure 2A,B). No steady state was reached for dieldrin and diquat dibromide. For all other compounds, steady-state concentrations could be modeled (Figure 2C,E). Modeling indicated that the internal concentrations at the end of the embryonic period or predicted for the steady state were close to the steady-state concentrations with BCFs similar or above those reported or estimated for adult fish (Figure 2C,E). These findings suggest that limitations in compound uptake for fish embryos were not the major driver for the observed discrepancies of FET and AFT data. Analysis of Locomotor Response Increased FET Sensitivity. The high proportion of neurotoxic compounds leading to no or low mortality in zebrafish embryos compared to AFT suggested a potential weak sensitivity of the FET for this mode of action if mortality is used as an end point. However, for these compounds (aldicarb, azinphos-methyl, dieldrin, endosulfan, and esfenvalerate), embryos may exhibit changes in behavior, i.e., changes in embryonic movement patterns, at concentrations close to the AFT-LC50. Therefore, we studied locomotion of individual zebrafish embryos by video tracking and determined swimming activity by analysis of the mean distance moved during a specified period. We included the herbicide diquat dibromide as a non-neurotoxic outlier in order to test for the specificity of behavioral responses for neurotoxic compounds. The EC50 values of the locomotor response (LMR) for neurotoxic chemicals were 3-fold (azinphos-methyl, LC10 = 1.64 mg/L) to 375-fold (aldicarb, LC10 = 159 mg/L) lower than the LC10 and could also be obtained for compounds with no mortality in the FET (Table 1 and Figure 4). Furthermore, effects on the LMR for neurotoxic compounds were observed at concentrations which did not induce morphological changes such as a curved body axis (Figure S4, Supporting Information).

Dieldrin, endosulfan, and esfenvalerate did not produce mortality up to the range of solubility (Table 1). External Concentration Analysis and Compound Uptake in Fish Embryos. For further subsequent analysis, we selected the 6 compounds with the highest deviation from the AFT based on the results of our FETs (aldicarb, azinphosmethyl, dieldrin, endosulfan, esfenvalerate, and diquat dibromide). For these compounds, exposure concentrations and/or internal concentration time courses were analyzed to exclude compound instability or a limited uptake of the compound as a source for the low toxicity in the FET. The exposure concentrations of dieldrin and endosulfan could not be detected in the HPLC analysis, since the highest test concentrations were close to the limits of detection. For aldicarb, azinphos-methyl, and diquat dibromide, chemical analysis of the exposure media revealed stable exposure concentrations in the static test (0−96 h), referring to 80− 120% of the nominal concentrations (Table S5, Supporting Information). Measured esfenvalerate concentration deviated by 79% from the nominal concentrations and decreased to 8.5% of the nominal concentrations within 24 h of exposure (Table S5, Supporting Information). For five of the outliers (aldicarb, azinphos-methyl, dieldrin, esfenvalerate, and diquat dibromide), the compound’s internal concentration, as an indicator for the uptake, was investigated (Figure 3 and Table S6, Supporting Information). For endosulfan, a previous study has already shown stable and close to nominal exposure concentrations, and it was demonstrated that equilibrium of internal concentrations was approached in zebrafish embryos and a BCF similar to the one observed for adult fish was reached.23 Internal concentration analysis of embryos indicated that a steady state was reached for aldicarb and azinphos-methyl within 48 h (Figure 3). The deduced (dimensionless) bioconcentration factors (BCFs) for aldicarb and azinphosmethyl in zebrafish embryos were in a similar range as wet7006

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology

Figure 4. Concentration−response analysis of the locomotor response in zebrafish embryos 96 h post fertilization. LMR of (A) aldicarb, (B) azinphos-methyl, (C) dieldrin, (D) endosulfan, (E) esfenvalerate, and (F) diquat dibromide. The frequency of the moved distance categories was compared between controls and treatments. The overlapping area (OA) of the density curves of controls and treatments were calculated (identical density curves would result in an OA of 1). The mean OAs observed for permutations of the control experiment are indicated by the horizontal dotted line and are used as the no effect offset for concentration response modeling (see Materials and Methods for further details). For compounds that did provoke mortality in the fish embryo test, the corresponding LC10 and LC50 values are depicted by black dotted lines.

The EC50 values of the LMR for neurotoxic compounds in the FET were closer to the AFT-LC50. However, for azinphosmethyl and dieldrin, EC50 concentrations were still about 46- to 51-fold higher than AFT-LC50s.



The FET-AFT relationship (Figure 3) obtained in this study was similar to those obtained from other studies.8−10 The most comprehensive study of Belanger and colleagues8 revealed a regression equation of AFT log LC50 = 1.026 × ET log LC50 − 0.307, on the basis of 144 chemicals. Our own calculation was based on 132 compounds with a regression slope of 1.13 and an intercept of −0.951. Both data sets have 106 chemicals in common. Twenty-six chemicals included in this study had not been previously considered in the FET-AFT regression, and 10 of these compounds showed a lower sensitivity in the FET with an FET/AFT ratio in the range of 27−5830. Since a high variability of the AFT might be associated with performance limitations,25,26 we used more stringent criteria for the acceptance of AFT data; i.e., AFT data were only accepted if exposure concentrations had been analytically confirmed. These differences may explain the slight deviation of regression coefficients from previous analyses.8,10,14 The high correlation

DISCUSSION

An OECD testing guideline (TG 236) has recently been established for the FET, and it is expected that it could increasingly be used to provide data on aquatic toxicity, for instance, in the REACH registration process.13,14,24 Albeit very rarely observed, compounds with a weaker toxicity in the FET have raised concerns of a potentially limited predictivity of the FET for AFT.9,17 Therefore, in this study, we systematically analyzed existing FET and AFT data to identify outliers. The aim was to better characterize the applicability domain and/or develop measures for further improvement of the FET. 7007

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology of fish embryo and acute fish toxicity tests apparently applies to a wide range of physicochemical properties, as has been already shown by Belanger et al.8 However, deviation from the correlation for some of the compounds might be related to either their high hydrophobicity (diquat dibromide) or high hydrophobicity (esfenvalerate). The time course of diquat dibromide internal concentration is likely associated with the slow uptake of the double positively charged diquat cation across cellular membranes. This is in line with the lack of mortality in early stages and an increase of toxicity at later stages (≥72 hpf). We can exclude the chorion as a barrier, as we did not detect an increased toxicity when zebrafish embryos were manually dechorionated at 24 hpf (data not shown). Similar observations were described for other compounds, e.g., the two cationic polymers merquat and luviquat.14 So far, a strong impact of the chorion on the compound uptake has only been shown for high molecular weight compounds.27,28 Two of the outliers, chlorothalonil and technical nonylphenol, revealed a higher toxicity in studies with prolonged exposure (96−120 h), but whether this is associated with limited uptake has not been investigated. The 96 h exposure period implemented in the OECD TG 236 included posthatched embryonic stages and would already prevent limitation to predict mortality due to a barrier function of the chorion and inadequately short exposure periods for compounds such as those described above. For very hydrophobic compounds, the difficulties to establish stable exposure concentration may result in lower toxicities in FETs conducted with a static exposure setup. In our study, 18 compounds with a log Kow > 5 were included. Eleven were classified as potential FET outliers because of no observed or low toxicity. This suggests that compounds with high log Kow might be poorly predictable from the FET. However, data on the stability of the exposure concentrations are lacking, and six of these chemicals represented pyrethroids, i.e., neurotoxic compounds acting via interference with sodium-gated voltage channels.29 One of these pyrethroids, esfenvalerate, was studied in greater detail in this study. Indeed, a 6.5- to 46-fold deviation of the measured concentration from nominal concentrations was observed for esfenvalerate. However, this deviation could not fully explain the low toxicity, since the AFT-LC50 was 27-fold below the geometric mean of the measured exposure concentration in the FET. Furthermore, internal concentrations of esfenvalerate did not indicate a weaker bioconcentration in fish embryos. A recent study with synthetic pyrethroids and zebrafish embryos observed a similar decline of the exposure concentrations during the first 24 h, which might be due to adsorption to glass surfaces. Further, a fast accumulation of synthetic pyrethroids in fish embryos has been observed, whereas elimination of these compounds was slow.30 The acute toxicity of neurotoxic compounds in fish is mediated by respiratory failure, i.e., inability to supply sufficient oxygen levels for essential life functions.31,32 In the case of AChE inhibitors, this has mainly been attributed to perturbations of cholinergic signaling, leading to, e.g., spasmassociated hemorrhages in blood vessels, decreased heart rates, and/or vasoconstriction of vessels in the gill lamellae. In contrast, in fish embryos and early life stages, in zebrafish until approximately 7−14 dpf, oxygen supply is not dependent on a proper function of the cardiovascular system and is provided mainly via diffusion through the skin.33,34 Hence, while concentrations impairing neural function in adults may ultimately cause death, the same concentrations may be

inefficient to provoke mortality in embryos. However, the impairment of neuronal signaling could be identified by estimation of effect concentrations for alterations in embryonic movements and used to predict acute toxicity in adult fish. Therefore, we determined the EC50 of the LMR of five neurotoxic outliers. For the selected outliers, the behavioral end point LMR in zebrafish embryos was effective at concentration up to 660-fold below FET LC50 or the maximum tested concentration, generally the limit of water solubility (Table 1). This ratio indicates that targets of neuronal signaling are present in the zebrafish embryos. Not only neurotoxic compounds can be expected to affect behavior but also compounds which induce gross malformations such as a curved body axis have been shown to be associated with aberrant embryonic behavior.35 However, neurotoxic compounds, as indicated in our study, affect behavior at concentrations that did not induce malformations, and hence, behavior can be considered as an indicator of a neurotoxic mode of action. The ability to demonstrate functional interaction with the neuronal system is associated with the differentiation status of the nervous system in fish embryos. Voltage-gated sodium channels (target of pyrethroids) are already expressed at 24 hpf, and acetylcholine esterase activity (target or organophosphates and carbamate pesticides) could be measured from 27 hpf.36,37 GABA-activated channels (targets for endosulfan and dieldrin) are present during early zebrafish embryogenesis and treatment with GABA pathway inhibitors increased locomotor activity in zebrafish larvae.38 Differences in the metabolic capacity of fish embryos have been discussed as a further potential limitation of the fish embryo test that could result in weaker toxicity. In zebrafish embryos, however, the majority of cytochrome P450 (CYP) genes, major classes of biotransformation enzymes, are already expressed indicating a principal capacity of CYP-mediated biotransformation in embryonic stages.39 Metabolic activation has been also demonstrated for various compounds,40 and internal concentration analysis provided further evidence for a metabolic capacity of fish embryos.41−43 A limited metabolic capacity as a cause for low fish embryo acute toxicity has so far only been demonstrated for one compound, allyl alcohol.9,17 Given that some of the identified outliers represent organophosphate AChE inhibitors, which are known to require activation by CYP to their oxon-metabolites,44 a limited metabolic activation to the oxon-derivate may have contributed to the lower toxicity in fish embryos for these compounds. However, also the oxon metabolites of several AChE inhibitors revealed no or only weakly higher toxicities in the FET if compared to the parent compounds (0.17- to 17-fold, Table S7, Supporting Information). Metabolites for other FET outliers such as endosulfan and aldicarb were previously studied in zebrafish embryos and did not reveal an increased toxicity if compared to the parent compound.23,45 Hence, for most of the compounds with weaker embryo toxicity, a limited metabolic activation cannot explain deviation of fish embryo from adult acute fish toxicity. The present study has provided strong evidence that particularly neurotoxic compounds may not sufficiently be predicted by the fish embryo test. Behavior could be analyzed as an additional end point to indicate neurotoxicity and to predict acute toxicity of adult fish. Further analysis is needed to clarify which method is appropriate to detect alterations in behavior. Recently, the touch-evoked response (e.g., Stanley et al.23), the photomotor response (PMR),46 and the LMR47,48 7008

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology have been studied most frequently in fish embryos. The analysis of touch-evoked response does not require any specific analytical equipment but depends on the subjective assessment and scoring by an individual observer. In contrast, LMR and PMR provide genuine quantitative methods but need specific equipment and software. At present, the different assays provide qualitative information on the potential neurotoxicity of the test compound and an associated higher acute toxicity. It requires further analysis, whether the different assays provide complementary information and allow one to distinguish between different modes of action. The PMR has already been shown that specific patterns in the response of embryos to a light stimulus can provide information to deduce a mechanism of action.46 A systematic analysis for other assays and their combinations is yet lacking. The FET-AFT comparative analysis conducted in this study clearly indicated an enrichment of a neurotoxic mode of action for compounds with low toxicity in the FET. However, a weaker acute toxicity was not observed for all neurotoxic compounds independent of their mode of action (e.g., disulfoton; malathion). This may have been caused by an enhanced metabolic degradation in adults, and analysis of behavioral end points in embryos could have led to an overprediction of acute toxicity. Further research is needed to identify the factors responsible for similar or deviating toxicities of neurotoxic compounds in fish embryos and adult fish and to compensate for, e.g., different metabolic capacities. While a quantitative prediction of acute fish toxicity LC50 using behavioral assays of the FET appears not yet possible, behavioral end points in fish embryos could already be used now to identify potential neurotoxic compounds as an intermediate step until more quantitative approaches have been developed. The identification of neurotoxicity could trigger the conduction of conventional fish acute toxicity tests or the application of assessment factors, while considering that the very good FET-AFT relationship for non-neurotoxic compounds of fish embryo LC50 is capable of predicting acute fish toxicity.



Uwe Schröter for his help with GC-MS analysis, and Rolf Altenburger for recommendations on internal concentration modelling. Ralf-Uwe Ebert is acknowledged for performing calculations regarding the chemical domain. We thank the EPAA Science Award Scientific Advisory Committee Members (Thomas Braunbeck, University of Heidelberg; Marlies Halder, JRC-IHCP; Gabriele Küsters, PIC G.Küsters; Marc Léonard, L’ORÉAL; Peter Maier, University of Zürich; Tzutzuy Ramirez, BASF; Belen Tornesi, Abbvie; Susanne Walter-Rohde, German Federal Environmental Agency) for critical supervision of the project, discussion of results, and internal review of the manuscript. L’ORÉAL is also acknowledged for funding the generation of a FET database and was involved as the primary industry partner in this project. This research has been financially supported by the EPAA 3Rs Science Award 2012 to N.K.



(1) Scholz, S.; Sela, E.; Blaha, L.; Braunbeck, T.; Galay-Burgos, M.; Garcia-Franco, M.; Guinea, J.; Klüver, N.; Schirmer, K.; Tanneberger, K.; Tobor-Kaplon, M.; Witters, H.; Belanger, S.; Benfenati, E.; Creton, S.; Cronin, M. T.; Egqgen, R. I.; Embry, M.; Ekman, D.; Gourmelon, A.; Halder, M.; Hardy, B.; Hartung, T.; Hubesch, B.; Jungmann, D.; Lampi, M. A.; Lee, L.; Leonard, M.; Küster, E.; Lillicrap, A.; Luckenbach, T.; Murk, A. J.; Navas, J. M.; Peijnenburg, W.; Repetto, G.; Salinas, E.; Schüürmann, G.; Spielmann, H.; Tollefsen, K. E.; Walter-Rohde, S.; Whale, G.; Wheeler, J. R.; Winter, M. J. A European perspective on alternatives to animal testing for environmental hazard identification and risk assessment. Regul. Toxicol. Pharmacol. 2013, 67 (3), 506−530. (2) OECD. OECD guideline for testing of chemicals. Test No. 203: Acute fish test; Organization for Economic Cooperation and Development: Paris, France, 1992. (3) Braunbeck, T.; Böttcher, M.; Hollert, H.; Kosmehl, T.; Lammer, E.; Leist, E.; Rudolf, M.; Seitz, N. Towards an alternative for the acute fish LC (50) test in chemical assessment: The fish embryo toxicity test goes multi-speciesan update. ALTEX 2005, 22 (2), 87−102. (4) Nagel, R. DarT: The embryo test with the zebrafish Danio rerio A general model in ecotoxicology and toxicology. ALTEX 2002, 19 (Suppl1/02), 38−48. (5) Rufli, H. Introduction of moribund category to OECD fish acute test and its effect on suffering and LC50 values. Environ. Toxicol. Chem. 2012, 31 (5), 1107−1112. (6) Rufli, H.; Springer, T. A. Can we reduce the number of fish in the OECD acute toxicity test? Environ. Toxicol. Chem. 2011, 30 (4), 1006−1011. (7) OECD. Guideline No. 126: Short guidance on the threshold approach for acute fish toxicity; Organization for Economic Cooperation and Development: Paris, France, 2010. (8) Belanger, S. E.; Rawlings, J. M.; Carr, G. J. Use of fish embryo toxicity tests for the prediction of acute fish toxicity to chemicals. Environ. Toxicol. Chem. 2013, 32 (8), 1768−1783. (9) Knöbel, M.; Busser, F. J.; Rico-Rico, A.; Kramer, N. I.; Hermens, J. L.; Hafner, C.; Tanneberger, K.; Schirmer, K.; Scholz, S. Predicting adult fish acute lethality with the zebrafish embryo: Relevance of test duration, endpoints, compound properties, and exposure concentration analysis. Environ. Sci. Technol. 2012, 46 (17), 9690−9700. (10) Lammer, E.; Carr, G. J.; Wendler, K.; Rawlings, J. M.; Belanger, S. E.; Braunbeck, T. Is the fish embryo toxicity test (FET) with the zebrafish (Danio rerio) a potential alternative for the fish acute toxicity test? Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 2009, 149 (2), 196−209. (11) EU. Directive 2010/63/EU of the European parliament and of the council of 22 September 2010 on the protection of animals used for scientific purposes. Off. J. Eur. Union, L 2010, L 276, 33−79. (12) EU. Commission Implementing Decision 2012/707/EU establishing a common format for the submission of the information

ASSOCIATED CONTENT

S Supporting Information *

Description of data compilation for FET and AFT, analysis of the chemical space, outlier identification and prioritization for validation, sampling procedure for chemical analysis of internal concentration, chemical analysis, and details of behavioral analysis. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/ acs.est.5b01910.



REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: +49(0)341 235 1558; e-mail: [email protected]. Present Address §

M.K. and R.M.: Department of Effect-Directed Analysis, UFZHelmholtz Centre for Environmental Research, Permoserstr. 15, 04318 Leipzig, Germany. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Hao Yu for support of esfenvalerate exposure experiments, Agata Turek for support with the HPLC analysis, 7009

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology pursuant to Directive 2010/63/EU of the European Parliament and of the Council on the protection of animals used for scientific purposes. Off. J. Eur. Union L 2012, L 320, 33−50. (13) OECD. OECD guideline for testing of chemicals. Test No. 236: Fish embryo acute toxicity (FET) test; Organization for Economic Cooperation and Development: Paris, France, 2013. (14) Busquet, F.; Strecker, R.; Rawlings, J. M.; Belanger, S. E.; Braunbeck, T.; Carr, G. J.; Cenijn, P.; Fochtman, P.; Gourmelon, A.; Hubler, N.; Kleensang, A.; Knobel, M.; Kussatz, C.; Legler, J.; Lillicrap, A.; Martinez-Jeronimo, F.; Polleichtner, C.; Rzodeczko, H.; Salinas, E.; Schneider, K. E.; Scholz, S.; van den Brandhof, E. J.; van der Ven, L. T.; Walter-Rohde, S.; Weigt, S.; Witters, H.; Halder, M. OECD validation study to assess intra- and inter-laboratory reproducibility of the zebrafish embryo toxicity test for acute aquatic toxicity testing. Regul. Toxicol. Pharmacol. 2014, 69 (3), 496−511. (15) Strahle, U.; Scholz, S.; Geisler, R.; Greiner, P.; Hollert, H.; Rastegar, S.; Schumacher, A.; Selderslaghs, I.; Weiss, C.; Witters, H.; Braunbeck, T. Zebrafish embryos as an alternative to animal experiments–A commentary on the definition of the onset of protected life stages in animal welfare regulations. Reprod. Toxicol. 2012, 33 (2), 128−132. (16) Sobanska, M. A.; Cesnaitis, R.; Sobanski, T.; Versonnen, B.; Bonnomet, V.; Tarazona, J. V.; De Coen, W. Analysis of the ecotoxicity data submitted within the framework of the REACH Regulation. Part 1. General overview and data availability for the first registration deadline. Sci. Total Environ. 2014, 470−471, 1225−1232. (17) Klüver, N.; Ortmann, J.; Paschke, H.; Renner, P.; Ritter, A. P.; Scholz, S. Transient overexpression of adh8a increases allyl alcohol toxicity in zebrafish embryos. PLoS One 2014, 9 (3), No. e90619. (18) Van Leeuwen, C. J.; Griffioen, P. S.; Vergouw, W. H. A.; MaasDiepeveen, J. L. Differences in susceptibility of early life stages of rainbow trout (Salmo gairdneri) to environmental pollutants. Aquat. Toxicol. 1985, 7 (1−2), 59−78. (19) Berghmans, S.; Butler, P.; Goldsmith, P.; Waldron, G.; Gardner, I.; Golder, Z.; Richards, F. M.; Kimber, G.; Roach, A.; Alderton, W.; Fleming, A. Zebrafish based assays for the assessment of cardiac, visual and gut function–Potential safety screens for early drug discovery. J. Pharmacol. Toxicol. Methods 2008, 58 (1), 59−68. (20) Scholz, S.; Ortmann, J.; Klüver, N.; Léonard, M. Extensive review of fish embryo acute toxicities for the prediction of GHS acute systemic toxicity categories. Regul. Toxicol. Pharmacol. 2014, 69 (3), 572−579. (21) Gobas, F. A.; Zhang, X. Measuring bioconcentration factors and rate constants of chemicals in aquatic organisms under conditions of variable water concentrations and short exposure time. Chemosphere 1992, 25 (12), 1961−1971. (22) US EPA. Estimation Programs Interface Suite for Microsoft Windows, v. 4.11; U. S. Environ. Prot. Agency, Washington, DC, USA, 2014. (23) Stanley, K. A.; Curtis, L. R.; Simonich, S. L.; Tanguay, R. L. Endosulfan I and endosulfan sulfate disrupts zebrafish embryonic development. Aquat. Toxicol. 2009, 95 (4), 355−361. (24) Tarazona, J. V.; Sobanska, M. A.; Cesnaitis, R.; Sobanski, T.; Bonnomet, V.; Versonnen, B.; De Coen, W. Analysis of the ecotoxicity data submitted within the framework of the REACH Regulation. Part 2. Experimental aquatic toxicity assays. Sci. Total Environ. 2014, 472, 137−145. (25) Hrovat, M.; Segner, H.; Jeram, S. Variability of in vivo fish acute toxicity data. Regul. Toxicol. Pharmacol. 2009, 54 (3), 294−300. (26) Raimondo, S.; Vivian, D. N.; Barron, M. G. Standardizing acute toxicity data for use in ecotoxicology models: Influence of test type, life stage, and concentration reporting. Ecotoxicology 2009, 18 (7), 918− 928. (27) Creton, R. The calcium pump of the endoplasmic reticulum plays a role in midline signaling during early zebrafish development. Brain Res. 2004, 151 (1−2), 33−41. (28) Cheng, J.; Flahaut, E.; Cheng, S. H. Effect of carbon nanotubes on developing zebrafish (Danio rerio) embryos. Environ. Toxicol. Chem. 2007, 26 (4), 708−716.

(29) Vijverberg, H. P.; vanden Bercken, J. Neurotoxicological effects and the mode of action of pyrethroid insecticides. Crit. Rev. Toxicol. 1990, 21 (2), 105−126. (30) Tu, W.; Lu, B.; Niu, L.; Xu, C.; Lin, C.; Liu, W. Dynamics of uptake and elimination of pyrethroid insecticides in zebrafish (Danio rerio) eleutheroembryos. Ecotoxicol. Environ. Saf. 2014, 107, 186−191. (31) Bradbury, S. P.; Carlson, R. W.; Henry, T. R.; Padilla, S.; Cowden, J. Toxic responses of the fish nervous system. In The toxicology of fishes; Taylor & Francis, Boca Raton, FL, 2008; pp 417− 456. (32) Russom, C. L.; LaLone, C. A.; Villeneuve, D. L.; Ankley, G. T. Development of an adverse outcome pathway for acetylcholinesterase inhibition leading to acute mortality. Environ. Toxicol. Chem. 2014, 33 (10), 2157−2169. (33) Jacob, E.; Drexel, M.; Schwerte, T.; Pelster, B. Influence of hypoxia and of hypoxemia on the development of cardiac activity in zebrafish larvae. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2002, 283 (4), R911−R917. (34) Rombough, P. Gills are needed for ionoregulation before they are needed for O2 uptake in developing zebrafish, Danio rerio. J. Exp. Biol. 2002, 205 (12), 1787−1794. (35) Ducharme, N. A.; Peterson, L. E.; Benfenati, E.; Reif, D.; McCollum, C. W.; Gustafsson, J. A.; Bondesson, M. Meta-analysis of toxicity and teratogenicity of 133 chemicals from zebrafish developmental toxicity studies. Reprod. Toxicol. 2013, 41, 98−108. (36) Novak, A. E.; Taylor, A. D.; Pineda, R. H.; Lasda, E. L.; Wright, M. A.; Ribera, A. B. Embryonic and larval expression of zebrafish voltage-gated sodium channel α-subunit genes. Dev. Dyn. 2006, 235 (7), 1962−1973. (37) Teixido, E.; Pique, E.; Gomez-Catalan, J.; Llobet, J. Assessment of developmental delay in the zebrafish embryo teratogenicity assay. Toxicol. in Vitro 2013, 27 (1), 469−478. (38) Baxendale, S.; Holdsworth, C. J.; Santoscoy, P. L. M.; Harrison, M. R.; Fox, J.; Parkin, C. A.; Ingham, P. W.; Cunliffe, V. T. Identification of compounds with anti-convulsant properties in a zebrafish model of epileptic seizures. Dis. Models & Mech. 2012, 5 (6), 773−784. (39) Goldstone, J. V.; McArthur, A. G.; Kubota, A.; Zanette, J.; Parente, T.; Jonsson, M. E.; Nelson, D. R.; Stegeman, J. J. Identification and developmental expression of the full complement of cytochrome P450 genes in zebrafish. BMC Genomics 2010, 11, 643. (40) Weigt, S.; Huebler, N.; Strecker, R.; Braunbeck, T.; Broschard, T. H. Zebrafish (Danio rerio) embryos as a model for testing proteratogens. Toxicology 2011, 281 (1−3), 25−36. (41) Alderton, W.; Berghmans, S.; Butler, P.; Chassaing, H.; Fleming, A.; Golder, Z.; Richards, F.; Gardner, I. Accumulation and metabolism of drugs and CYP probe substrates in zebrafish larvae. Xenobiotica 2010, 40 (8), 547−557. (42) Carlsson, G.; Patring, J.; Kreuger, J.; Norrgren, L.; Oskarsson, A. Toxicity of 15 veterinary pharmaceuticals in zebrafish (Danio rerio) embryos. Aquat. Toxicol. 2013, 126, 30−41. (43) Kühnert, A.; Vogs, C.; Altenburger, R.; Küster, E. The internal concentration of organic substances in fish embryosA toxicokinetic approach. Environ. Toxicol. Chem. 2013, 32 (8), 1819−1827. (44) Guo, J.-X.; Wu, J. J. Q.; Wright, J. B.; Lushington, G. H. Mechanistic insight into acetylcholinesterase inhibition and acute toxicity of organophosphorus compounds: A molecular modeling study. Chem. Res. Toxicol. 2006, 19 (2), 209−216. (45) Kuster, E.; Altenburger, R. Suborganismic and organismic effects of aldicarb and its metabolite aldicarb-sulfoxide to the zebrafish embryo (Danio rerio). Chemosphere 2007, 68 (4), 751−760. (46) Kokel, D.; Bryan, J.; Laggner, C.; White, R.; Cheung, C. Y. J.; Mateus, R.; Healey, D.; Kim, S.; Werdich, A. A.; Haggarty, S. J. Rapid behavior-based identification of neuroactive small molecules in the zebrafish. Nat. Chem. Biol. 2010, 6 (3), 231−237. (47) Selderslaghs, I. W.; Hooyberghs, J.; Blust, R.; Witters, H. E. Assessment of the developmental neurotoxicity of compounds by measuring locomotor activity in zebrafish embryos and larvae. Neurotoxicol. Teratol. 2013, 37, 44−56. 7010

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011

Article

Environmental Science & Technology (48) Selderslaghs, I. W.; Hooyberghs, J.; De Coen, W.; Witters, H. E. Locomotor activity in zebrafish embryos: A new method to assess developmental neurotoxicity. Neurotoxicol. Teratol. 2010, 32 (4), 460− 471. (49) Prosch, S.; Voss, S. Interaction of glyphosate with photosystem II inhibitor herbicides as a selection tool for roundup ready events. Patent WO2006121902 A3, 2006. (50) Savaldi-Goldstein, S.; Baiga, T. J.; Pojer, F.; Dabi, T.; Butterfield, C.; Parry, G.; Santner, A.; Dharmasiri, N.; Tao, Y.; Estelle, M.; Noel, J. P.; Chory, J. New auxin analogs with growth-promoting effects in intact plants reveal a chemical strategy to improve hormone delivery. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (39), 15190−15195. (51) Dodge, A.; Harris, N. The mode of action of paraquat and diquat. Biochem. J. 1970, 118, 43P−44P. (52) Zhao, X.; Salgado, V. L.; Yeh, J. Z.; Narahashi, T. Differential actions of fipronil and dieldrin insecticides on GABA-gated chloride channels in cockroach neurons. J. Pharmacol. Exp. Ther. 2003, 306 (3), 914−924.

7011

DOI: 10.1021/acs.est.5b01910 Environ. Sci. Technol. 2015, 49, 7002−7011