www.elsevier.com/locate/chemosphere

Chemosphere 68 (2007) 1435–1441

Comparison of in vitro and in vivo acute fish toxicity in relationto toxicant mode of action

K. Knauer *, C. Lampert, J. Gonzalez-Valero

Syngenta Crop Protection AG, Schwarzwaldallee, 4002 Basel, Switzerland

Received 16 November 2006; received in revised form 31 March 2007; accepted 4 April 2007Available online 21 May 2007

Abstract

Acute toxicity to fish hepatoma cell line PLHC-1 and to juvenile rainbow trout was examined for 18 plant protection products. Themain objective was to explore whether hepatoma cells could be used to predict acute toxicity in fish taking into account the mode of toxicaction and compound properties. Acute fish toxicity was determined using the OECD guideline test 203 and compared to predicted base-line LC50 of acute fish toxicity calculated with a quantitative structure–activity relationship (QSAR) derived for guppy fish. Cytotoxicitywas determined through the inhibition of neutral red uptake (NR50) into lysosomes and compared to predicted baseline cytotoxicityderived for goldfish GFS cells. In general, NR50 values were higher by a factor ranging from 3 to 3000 than the corresponding acuteLC50. A weak correlation between NR50 and LC50 values was found (log/log: r2 = 0.62). Also the lipophilicity (logKow) was not a goodpredictor for cytotoxicity (r2 = 0.43) and lethality (r2 = 0.57) of these pesticides. The neutral red assay is detecting general baseline toxi-city only. Comparing LC50 data to QSAR results, the compounds can be classified to act as narcotics or reactive compounds with aspecific mode of toxic action in fish. The results indicate that limitation of the neutral red assay in predicting acute fish toxicity. A pro-mising alternative might be the assessment of toxicity in a set of in vitro systems addressing also cell-specific functions which are related tothe mode of toxic action of the compound.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Rainbow trout; Hepatoma cell line PLHC-1; QSAR; Pesticides

1. Introduction

Pesticides are designed to inhibit specific metabolic pro-cesses in target organisms. For an aquatic risk assessment,the toxicity of these compounds to non-target organisms isadditionally assessed to evaluate the toxic potential of thecompounds in an aquatic environment (91/414/EEC,1991). Traditionally, the environmental hazard on verte-brates in aquatic systems is evaluated by performing acute(OECD 203, 1992) and chronic fish experiments (OECD203, 1984; OECD 305, 1996). The most widely performedtest is the acute fish toxicity test.

0045-6535/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.chemosphere.2007.04.009

* Corresponding author. Tel.: +41 61 2670405; fax: +41 61 2670409.E-mail address: katja.knauer@unibas.ch (K. Knauer).

In the last decade, more than 150 continuous fish celllines have been established and a number of bioassayshave been developed in an attempt to replace acute toxic-ity tests with fish (Castano et al., 2003; Schirmer, 2006).Cell lines such as hepatoma cells PLHC-1 from the deserttopminnow (Poeciliopsis lucida) (Bruschweiler et al., 1995;Fent and Hunn, 1996), hepatocytes and gill epithelial cellsfrom rainbow trout (Oncorhynchus mykiss) (Castano et al.,1996; Sandbacka et al., 2000; Castano et al., 2003) wereshown to be suitable for the determination of cytotoxicityendpoints.

Currently, the use of fish cells in ecotoxicology isfocused on measurements of cytotoxicity, both basal andselective, genotoxicity and effects on cell-specific functionsand parameters (Castano et al., 2003; Bols et al., 2005).In most cytotoxicity studies with fish cells, basal toxicity

1436 K. Knauer et al. / Chemosphere 68 (2007) 1435–1441

has been determined, mainly by measuring membraneintegrity or energy metabolism (Babich and Borenfreund,1991; Segner, 1998; Segner and Braunbeck, 2003; Dayehet al., 2004). For example, the neutral red (NR) test indi-cates membrane permeability, particularly for the lyso-somal membrane (Borenfreund and Puerner, 1985), thetetrazolium salt reduction (MTT) assay measures the activ-ity of a mitochondrial succinate-dependent enzyme (Mos-mann, 1983), the bromodeoxyuridine (BrdU) assaydetermines cell proliferation (Gratzner, 1982), and the crys-tal violet (CV) assay measures protein content (Saad et al.,1993).

These experiences called into question the use of cyto-toxicity instead of acute fish toxicity tests. However, themain drawback of in vitro cytotoxicity tests is their seem-ingly lower sensitivity compared to fish (Segner and Lenz,1993; Castano et al., 2003; Segner, 2004) which restrictstheir use as alternative to the acute fish test. However,for screening and ranking purposes of chemicals, cytotoxi-city tests might be still be a valuable approach.

In the present investigation, the responses of rainbowtrout and the fish cell line PLHC-1 towards 18 pesticideswere analyzed with respect to different modes of action.In vivo lethality and in vitro cytotoxicity were investigatedto receive comparable information in the fish and the cyto-toxicity test. Toxicological responses are discussed in termsof their non-specific sensitivity using lipophilicity (logKow)as molecular descriptor.

Table 1Chemical, class of chemical, CAS number, DMSO content in % in the in vitro

Chemical Pesticide CAS number DMSO (%)in vitro assay

P

Atrazine H 1912-24-9 0 PTerbutylazine H 5915-41-3 0 PDiafenthiuron I, A 80060-09-9 0.1 I

ind

Dichlornitroacetanilide I – 0 ?Dicyclanil I 112636-83-6 0 PStrobilurin-derivate F – 0.1 I

bc

Thiabendazole F 148-79-8 0.2 Iin

Fenpiclonil F 74738-17-3 0.5 Pa

Cyprodinil F 121552-61-2 0.5 IAcibenzolar-S-methyl F 135158-54-2 0.5 NAzole: F I

bin

Cyproconazole 94361-06-5 0.2Myclobutanil 88671-89-0 0.5Tebuconazole 107534-96-3 0.5Penconazole 66246-88-6 0.1Propiconazole 60207-90-1 0.2Imazalil 35554-44-0 0.1Prochloraz 67747-09-5 0.1Difenoconazole 119446-68-3 0.2

F = fungicide, H = herbicide, I = insecticide, A = acaricide.

2. Materials and methods

2.1. Chemicals

The test set of 18 chemicals consisted of herbicides,insecticides and fungicides, covering substances with adirect target site in fish such as mitochondrial respirationand those without target effects in fish (Table 1). As a posi-tive control for sensitivity of the fish cells, tributyltin-chlo-ride (TBT) was used (Bruschweiler et al., 1995).

For the determination of cytotoxicity, stock solutionswere prepared in Eagle’s Minimum Essential Medium(MEM, Biochrom KG, Berlin Germany) and dimethylsulf-oxide (DMSO, Fluka GmbH, Buchs, Switzerland). The finalDMSO concentration in the microtiter wells are reported inTable 1. For the acute fish toxicity, substances were dis-solved in de-chlorinated tap water and the final DMSO con-centration in the aquaria were far below 0.1%.

2.2. Fish

Juvenile rainbow trout (O. mykiss) were supplied by P.Hohler, 4314 Zeiningen, Switzerland. Fish were held forseveral days prior to testing in water of similar quality usedin the test. The fish were fed a commercially prepared troutfood diet (HOKOVIT No.: 502 (H.U. Hofmann, 4922-Butzberg, Switzerland) supplemented by living, frozen ordried organisms, e.g. daphnia and larvae of mosquitoes,

assay, and mode of action

rimary mode of action Reference

hotosystem II inhibitor Forney and Davis (1981)hotosystem II inhibitor Forney and Davis (1981)nhibitor of mitochondrial respiration,hibitor of oxidative phosphorylation,

isruptor of ATP formation

Dekeyser (2005)

–revent development of larvae and pupae Pesticide Manual (1997)nhibits mitochondrial respiration bylocking electron transport betweenytochrome b and cytochrome c1

Pesticide Manual (1997)

nhibition of fumarate reductase activitymitochondrial fraction

Fornelio et al. (1987)

ossibly acts by inhibition of transportssociated phosphorylation of glucose

Jespers and Dewaard (1995)

nhibits methionine biosynthesis Pesticide Manual (1997)o direct effect on target pests –

nhibits cell membrane ergosteroliosynthesis, sterol demethylationhibitor, stopping development of fungi

Rievere et al. (1984)

K. Knauer et al. / Chemosphere 68 (2007) 1435–1441 1437

until 24 h prior to the test. Total mortality did notexceed 3% during the 7 day period prior to the start ofthe test.

2.3. Cell culture

Fish hepatoma cells PLHC-1 from the desert topmin-now (P. lucida) were kindly supplied by K. Fent, HBB,Switzerland, from a culture originally isolated by High-tower, University of Connecticut, USA (1988). The cellswere maintained as described in Bruschweiler et al.(1995). For the cytotoxicity tests, cells were incubated on96-well microtiter plates containing 100 ll medium, L-glu-tamine, and 10% fetal calf serum (FCS). Cells were incu-bated until approximately 80% confluency was reached ata temperature of 30 �C and 5% CO2 atmosphere.

2.4. Acute fish test

Rainbow trout (N = 7 per aquarium) were exposed for96 h to each of the test substances under static conditions(OECD 203, 1992). At least five concentrations and a dilu-tion water control were tested to determine the LC50 (con-centration at which 50% of the fish population dies). Theaquaria had a working volume of 20 l. The loading of theaquaria did not exceed 1 g fish/l. The average length ofthe fish was 5.0 ± 1.0 cm. Mortalities were recorded follow-ing the guideline for fish acute toxicity OECD 203.

2.5. Neutral red assay

For the determination of in vitro toxicity, the quantita-tive neutral red assay was applied as described in Borenfre-und and Puerner (1985). The accumulation of the dye isinhibited if a chemical caused lysosomal or membranedamage. Cells were incubated with various substances for24 h and cytotoxicity was determined using 12 replicatewells per toxicant concentration. As a positive control,tributyltin-chloride (TBT) was used. Average values ofthe control cytotoxicity were EC50 of 2.3 ± 0.59 · 10�7 Mand these were comparable to published values (EC50 forTBT of 1.1 · 10�7 M) (Bruschweiler et al., 1995).

2.6. Data analysis

Results of the toxicity test were interpreted by standardstatistical techniques using SAS Version 6.11. and the pro-cedure ‘‘ECOS’’ by Fisch and Strehlau (1998a,b) using SASPROC GLM and PROC NLIN (SAS, 1990). Mortalitydata or vitality data were used to statistically estimate amedian lethal or effect concentration (LC50, NR50) after96 and 24-h exposure. The LC50 or NR50 values were cal-culated according to the maximum likelihood method, pro-bit model (Finney, 1971).

2.7. Quantitative structure–activity relationships (QSAR)

To determine if the pesticides tested act as narcotic sub-stances only or interfere more specifically with metabolicfunctions in fish, we compared the LC50 data to the QSARmodel for guppies describing baseline toxicity only (Eq. (1):Konemann, 1981).

LogðLC50 baselineÞ ¼ �0:87 � log Kow þ 1:87 ð1Þ

Comparisons were further performed to other earlier pub-lished QSARs for baseline toxicity in fathead minnow(Eq. (2): McCarty et al., 1991, Eq. (3): Russom et al., 1997,Eq. (4) ECB, 2003, Eq. (5): Maeder et al., 2004).

LogðLC50 baselineÞ ¼ �0:87 � log Kow þ 1:87 ð2ÞLogðLC50 baselineÞ ¼ �0:94 � log Kow þ 1:75 ð3ÞLogðLC50 baselineÞ ¼ �0:85 � log Kow þ 1:61 ð4ÞLogðLC50 baselineÞ ¼ �0:87 � log Kow þ 1:68 ð5Þ

The cytotoxicity values (NR50) of the pesticides tested werecompared to baseline toxicity (NR50 baseline) as calculatedfrom the Saito equation derived for NR50 (24 h) of goldfishGFS cells and chlorophenols (Eq. (6): Saito et al., 1993).The logKow’s of the substituted phenols were corrected tothe logDlipw at pH 7 taking into account the distributionratio between liposomes and water (Dlipw) as descriptor(Escher and Schwarzenbach, 1996, 2002).

Logð1=NR50 baselineÞ ¼ 1:42 � log Dlipw � 0:99 ð6Þ

Compounds were further classified by calculating a toxicratio (TR = NRbaseline/NR50 and LC50 baseline /LC50) andattributed to the four chemical classes non-polar and polarnarcotics (class 1 and 2), and compounds with a reactive ora specific mode of toxic action (class 3 and 4) (Verhaaret al., 1992). Narcotics 1 and 2 are not distinguished herebecause their mode of toxic action is the same, namelynon-specific disturbance of structure and function of bio-logical membranes (Vaes et al., 1998; Escher and Schwar-zenbach, 2002).

3. Results

3.1. Toxicity evaluation

The acute toxicity data reveal substantial differencesin the sensitivity of the two test systems (Table 2, ratioNR50/LC50). In general, NR50 values were higher thanLC50 demonstrating the lower sensitivity of the in vitro

assay compared to the in vivo toxicity test.A comparison of NR50 values to LC50 for the 18 com-

pounds and for the azole only resulted in a poor concordance(Fig. 1) demonstrating the limitation of prediction of theneutral red assay for in vivo effects. The outlier can be attrib-uted to class 3 and 4 compounds (TR > 10) (Table 2).

Further, the logKow values were compared to NR50 orLC50 values resulting in a weak correlation r2 of 0.43(slope = 0.54) and 0.57 (slope = 0.95) (N = 18), respec-

Table 2The ratio between LC50 baseline baseline toxicity derived for guppy fish and the experimental determined acute toxicity LC50 in rainbow trout and betweenthe NR50 baseline baseline toxicity derived for goldfish GFS cells and NR50 values in the PLHC-1 fish cell line for pesticides and the ratio of NR50/LC50.

Chemical LogKow Log(1/LC50 baseline)[M]

Log(1/LC50) [M]

RatioLC50 baseline/LC50

Log(1/NR50 baseline)[M]

Log(1/NR50)[M]

RatioNR50 baseline/NR50

NR50/LC50

Guppy Rainbowtrout

Goldfish GFScells

Desert topminnowPLHC-1 cells

Atrazine 2.50 3.32 4.29 9.49 2.55 <3.30 0.77 10Terbutylazine 3.10 3.84 4.78 8.82 3.41 <2.30 1.48 303Diafenthiuron 5.76 6.15 8.73 379.56 7.19 5.22 1.38 3208Dichlornitroacetanilide 4.91 3.46 28Dicyclanile 3.50 <3.00 3Strobulin-derivate 7.44 6.83 4Thiabendazole 2.30 3.14 5.56 264.76 2.28 <3.40 0.67 146Fenpiclonile 3.86 4.50 5.47 9.42 4.49 3.89 1.16 39Cyprodinile 4.00 4.62 5.36 5.51 4.69 3.59 1.31 60Acibenzolar-S-Methyl 3.10 3.84 5.72 76.58 3.41 <4.85 0.70 7Cyproconazole 2.91 3.67 4.19 3.27 3.14 3.20 0.98 10Myclobutanil 2.94 3.70 4.85 14.32 3.18 3.32 0.96 34Tebuconazole 3.7 4.40 4.68 2.1 4.26 3.59 1.18 12Penconazole 3.72 4.38 4.82 2.78 4.29 4.00 1.07 7Propiconazole 3.72 4.38 4.88 3.18 4.29 3.74 1.14 13Imazalile 3.82 4.46 5.30 6.88 4.43 3.89 1.14 26Prochloraz 4.12 4.72 5.40 4.72 4.86 3.96 1.23 28Difenoconazole 4.2 4.79 5.57 5.93 4.97 4.21 1.18 23

LC50 baseline = baseline toxicity with the equation log(1/LC50 baseline) = 0.87 * logKow + 1.14 (Konemann, 1981).NR50 baseline = baseline toxicity with the equation log(1/NR50 baseline) = 1.42 * logDlipw � 0.99 (Saito et al., 1993).LC50 and NR50 = experimental determined data.

Azole: y = 1.0637x + 0.9836, R2 = 0.6857

Pesticides: y = 0.9432x + 1.6486, R2 = 0.6234

2

4

6

8

log(1/NR50)

log(

1/LC

50)

Pesticides Azole

Linear (Azole) Linear (Pesticides)

2 3 4 5 6 7

Fig. 1. The relationship between logKow and acute toxicity values(log(1/NR50) and (log(1/LC50) determined in PLHC-1 fish cells andrainbow trout, respectively.

2

2 3 4 5 6

4

6

8 baseline toxicity, GFS cells neutral red, PLHC-1

log Kow, logDlipw

log

(1/N

R50

)

Fig. 2. The relationship between NR50 values obtained from cytotoxicityin PLHC-1 fish cells and logKow investigating 18 pesticides. The line isbased on the effect data of chlorphenolic compounds determined with theneutral red assay and is considered as the baseline toxicity derived forgoldfish GFS cells (Saito et al., 1993).

1438 K. Knauer et al. / Chemosphere 68 (2007) 1435–1441

tively. Considering the data of the azoles only, a strongercorrelation was determined (Eq. (7)) demonstrating anappropriate predictability of the baseline toxicity usinglogKow for this class of substances. But for the lethalitydata of the azole again a weak relationship was found(Eq. (8)).

log 1=NR50 ¼ 0:68 log Kow þ 1:28; n ¼ 8; r2 ¼ 0:87 ð7Þlog 1=LC50 ¼ 0:75 log Kow þ 2:22; n ¼ 8; r2 ¼ 0:66 ð8Þ

These results demonstrate that for the group of pesticidesinvestigated lipophilicity is not a good predictor for theacute toxicity in fish. All slopes are lower than 1 indicatingthat octanol is not an ideal surrogate of cellular lipids asalso reported by others (McCarty et al., 1985, 1991, 1993).

Slope and intercept of the fit of the LC50 data for rain-bow trout (Fig. 2, Eq. (8)) demonstrated comparable

K. Knauer et al. / Chemosphere 68 (2007) 1435–1441 1439

results to earlier published QSARs for baseline toxicity inother fish species like guppy (Eq. (1)) and fathead minnow(see Eqs. (2)–(5)).

3.2. Cytotoxicity and baseline toxicity (QSAR)

Comparing the NR50 values to the calculated NR50 base-

line data for narcotic substances, it can be concluded thatmembrane disruption in PLHC-1 is an effective endpointfor determining general baseline toxicity of the compoundstested (0.1 < EC50 baseline/NR50 < 10) (Fig. 2, Table 2). Theline drawn in Fig. 2 is based on the effect data of chloro-phenolic compounds determined with the neutral red assayand is considered as the baseline toxicity derived for gold-fish GFS cells (Saito et al., 1993; Escher and Schwarzen-bach, 2002). Despite the fact that this analysis has someshortcomings, e.g. different cells used and a different setof compounds, the agreement between the baseline QSARand all experimental data is close enough to conclude thatthe neutral red assay exclusively detects baseline toxicity.

3.3. Lethality and baseline toxicity (QSAR)

Comparing the LC50 values to the calculated LC50 baseline

data, four of the 18 compounds were classified as class 3 or4 compounds since more than a factor of 10 between LC50

and LC50 baseline was calculated (Verhaar et al., 1992)(Fig. 3, Table 2). The TR of 10 compounds ranged between1 and 10, indicating that the compounds act as narcotics.The lines drawn in Fig. 3 are based on the data of non-polar narcotic compounds and are considered as the base-line toxicity for guppy fish (Eq. (1)) and fathead minnow(Eqs. (2)–(5)).

2.0 3.0 4.0 5.0 6.0-6

-5

-4

-3

-2

-1

0

rainbow trout (pesticides this study)rainbow trout (triazole this study)guppy (Könemann 1981)

fathead minnow (Russom et al. 1997) fathead minnow (McCarty et al. 1991)fathead minnow (ECB 2003)fathead minnow (Maeder et al. 2004)

logL

C50

(mM

)

logKow

Fig. 3. The relationship between LC50 values obtained from acute fishtoxicity tests and logKow for 18 pesticides. The lines are based on the dataof non-polar narcotic compounds and are considered as the baselinetoxicity for guppy fish (Konemann, 1981) and fathead minnow (McCartyet al., 1991; Russom et al., 1997; ECB, 2003; Maeder et al., 2004).

4. Discussion

The acute fish test was more sensitive and valuable todetermine effects of specific acting compounds and couldnot be simply replace by the neutral red cytotoxicity assay.The differences in the sensitivity of the two tests may inpart be a consequence of the different durations of expo-sure used for the fish cell line (24 h) and the in vivo fish test(96 h) which might potentially lead to higher internal com-pound concentrations in the fish compared to the fish cells.Since the response of an organism is best related to theinternal concentration of a toxic compound (Escher andHermens, 2002), the observed differences in sensitivitymight therefore be explained by different target site concen-trations. Further, we have to consider that compoundsmight be captured by media components and that thereforethe bioavailable concentration might differ in the two testsystems and lead to different toxicities (Heringa et al.,2004). Also the endpoints lethality and membrane disrup-tion are not directly comparable. Lethality as an endpointcan be seen as an integrative response of an organism to achemical stressor, whereas the neutral red assay can detectonly injury specific to the lysosomes. Lower sensitivities offish cell lines compared to in vivo fish exposure have alsobeen reported by others (Saito et al., 1991; Castanoet al., 1996; Schirmer, 2006).

The next question was, whether and how these differ-ences in biological response of the two test systems canbe tracked back to specific sensitivity differences for certaingroups of chemicals corresponding to a certain mode oftoxic action. Greatest toxicity was observed for diafenthiu-ron, and thiabendazole in the acute fish test (TR > 100)which are known to inhibit mitochondrial respiration infish (Fornelio et al., 1987; Dekeyser, 2005). It is thereforenot surprising that for these compounds, the neutral redassay was not suitable to determine effect concentrationscomparable to the acute fish test. An appropriate in vitro

test to detect effects on respiration is the MTT assay (Mos-mann, 1983) which investigates mitochondrial dehydroge-nase activity. For compounds such as the fungicideacibenzolar-S-methyl for which the mode of toxic actionhas not yet been described (EPA, 2000) and which is highlytoxic to fish (10 < TR < 100), we recommend the use of aset of in vitro cytotoxicity tests such as neutral red uptake,MTT assay, Alarma Blue CFDA-AM (Schirmer, 2006).Effects should further not only be investigated in a singlecell line but in different cell cultures reflecting different tar-get sites in fish (e.g. liver, gonads, gill, brain). A valuablecombination of suitable endpoints and cell culture modelsmight be a promising alternative for the acute in vivo fishtest and an appropriate tool in toxicity screening and rank-ing of compounds.

All other compounds tested were less toxic in fish andTR ranged between 1 and 10 indicting that their acute toxi-city is well described by baseline toxicity. However, forsome of these chemicals effects on specific physiologicalpathways have been reported in chronic fish studies. It is

1440 K. Knauer et al. / Chemosphere 68 (2007) 1435–1441

beyond the scope of this article to provide a comprehen-sive, in-depth evaluation of bioassay’s for metabolic end-points in fish cell cultures but a few aspects should bementioned. Atrazine e.g. is discussed to have endocrine-dis-rupting potential (Bisso and Hontela, 2002; Chang et al.,2005) and e.g. prochloraz has been found to reduced thefecundity of fish (Ankley et al., 2005). To identify possiblelong-term effects of chemicals on populations, other end-points than vitality are important to be considered. Hence,in vitro assays that measure cell-specific functions like vitel-login expression (Navas and Segner, 2006) or the bindingto specific receptors like e.g. the androgen receptor mightbe more sensitive and reflect the specific mode of actionof these compounds in fish (Snegaroff and Bach, 1989;Babin et al., 2005). For a proper risk evaluation of specificacting compounds, it is recommended to complement base-line cytotoxicity tests by in vitro assays investigating effectson cell-specific functions.

5. Conclusions

Characterization of the ecotoxicological hazard throughaquatic contaminants requires biological test systemswhich offer sufficient specific sensitivity for relevant modesof toxic action. For a given chemical and test assay, thebiological response depends on both compound propertiesand characteristics of the test system and endpoint. Theendpoint mortality in the acute fish test is an integrativeresponse of the organism to the chemical stressor with itsvarious modes of toxic actions in fish. A set of in vitro testsinvestigating effects on different endpoints in various cellculture models might be a suitable replacement for acutein vivo toxicity tests with fish. The challenge for selectingfunctional endpoints is to identify those that do have pre-dictive value for the expression of toxicity.

Acknowledgements

We are grateful to Dr. K. Fent for the supply of the fishcell line. We sincerely thank Dr. Beate Escher, Dr. KristinSchirmer, and Prof. Dr. Helmut Segner and two anony-mous reviewers for valuable comments and discussion.

References

Ankley, G.T., Jensen, K.M., Durhan, E.J., Makynen, E.A., Butterworth,B.C., Kahl, M.D., Villeneuve, D.L., Linnum, A., Gray, L.E., Cardon,M., Wilson, V.S., 2005. Effects of two fungicides with multiple modesof action on reproductive endocrine function in the fathead minnow(Pimephales promelas). Toxicol. Sci. 86, 300–308.

Babich, H., Borenfreund, E., 1991. Cytotoxicity and genotoxicity assayswit hcultured fish cells: a review. Toxicol. In Vitro 5, 91–100.

Babin, M., Casado, S., Chana, A., Herradon, B., Segner, H., Tarazona,J.V., Navas, J.M., 2005. Cytochrome P4501A induction caused by theimidazole derivative prochloraz in a rainbow trout cell line. Toxicol. InVitro 19, 899–902.

Bisso, M., Hontela, A., 2002. Cytotoxic and endocrine-disruptingpotential of atrazine, diazinon, endosulfan, and mancozeb in adreno-

cortical steroidogenic cells of rainbow trout exposed in vitro. Toxicol.Appl. Pharmacol. 180, 110–117.

Bols, N.C., Dayeh, V.R., Lee, L.E.J., Schirmer, K., 2005. Use of fish celllines in the toxicology and ecotoxicology of fish. In: Mohn, T.W.,Mommsen, T.P. (Eds.), . In: Biochemistry and Molecular Biology ofFishes, vol. 6. Elsevier Science, Amsterdam, Netherland.

Borenfreund, E., Puerner, J.A., 1985. A simple quantitative procedureusing monolayer cultures for cytotoxicity assays (HTD/NR-90). J.Tissue Culture Methods 9 (1), 7–9.

Bruschweiler, B.J., Wurgler, F.E., Fent, K., 1995. Cytotoxicity in-vitro oforganotin compounds to fish hepatoma cells PLHC-1 (Poeciliopsis

lucida). Aquat. Toxicol. 32, 143–160.Castano, A., Bols, N.C., Braunbeck, T., Dierick, P., Halder, M., Isomaa,

B., Kawahara, K., Lee, L.E.J., Mothersill, C., Part, P., Repetto, G.,Sintes, J.R., Rufli, H., Smith, R., Wood, C., Segner, H., 2003. The useof fish cells in ecotoxicology. The report and recommendations ofECVAM workshop 47. ALTA 31, pp. 317–351.

Castano, A., Cantarino, M.J., Castillo, P., Tarazona, J.V., 1996. Corre-lations between the RTG-2 cytotoxicity test EC50 and in vivo LC50rainbow trout bioassay. Chemosphere 32, 2141–2157.

Chang, L.W., Toth, G.P., Gordon, D.A., Graham, D.W., Meier, J.R.,Knapp, C.W., deNoyelles Jr., F.J., Campbell, S., Lattier, DL., 2005.Responses of molecular indicators of exposure in mesocosms: commoncarp (cyprinus carpio) exposed to the herbicides alachlor and atrazine.Environ. Toxicol. Chem. 24, 190–197.

Dayeh, V.R., Chow, S.L., Schirmer, K., Lynn, D.H., Bols, N.C., 2004.Evaluating the toxicity of Triton X-100 to protozoan, fish, andmammalian cells using fluorescent dyes as indicators of cell viabilityEcotoxicol. Environ. Safe. 57, 375–382.

Dekeyser, M.A., 2005. Acaricide mode of action. Pest. Manag. Sci. 61,103–110.

ECB, 2003. Technical Guidance Document on Risk Assessment inSupport of Commission Directive 93/67/EEC on Risk Assessmentfor New Notified Substances, Commission Regulation (EC) No. 1488/94 on Risk Assessment for Existing Substances, Directive 98/8/EC ofthe European Parliament and of the Council concerning the placing ofbiocidal products on the market. Environmental Risk Assessment(Part II). Institute for Health and Consumer Protection, EuropeanChemicals Bureau.

EPA (Environmental Protection Agency), 2000. Pesticide Fact Sheet,Acibenzolar-S-Methyl, Office of Prevention, Pesticides and ToxicSubstances, 7501C.

Escher, B.I., Hermens, J.L.M., 2002. Modes of action in ecotoxicology:their role in body burdens, species sensitivity, QSARs, and mixtureeffects. Environ. Sci. Technol. 36, 4201–4217.

Escher, B.I., Schwarzenbach, R.P., 1996. Partitioning of substitutedphenols in liposome–water, biomembrane–water, and octanol–watersystems. Environ. Sci. Technol. 30, 260–270.

Escher, B.I., Schwarzenbach, R.P., 2002. Mechanistic studies on baselinetoxicity and uncoupling of organic compounds as a basis for modelingeffective membrane concentrations in aquatic organisms. Aquat. Sci.64, 20–35.

European Commission Health and Consumer protection Directorate-General, 1991. Guidance document on aquatic ecotoxicology, directive91/414/EEC.

Fent, K., Hunn, J., 1996. Cytotoxicity of organic environmental chemicalsto fish liver cells (PLHC-1). Mar. Environ. Res. 42, 1–4.

Finney, D.J., 1971. Probit Analysis, third ed. Cambridge University Press,London, pp. 655–660.

Fisch, R.D., Strehlau, G.A., 1998a. A simplified approach to calibrateconfidence sets in multilinear and nonlinear regression. InternalReport, Novarits Crop Protection, Basel, Switzerland.

Fisch, R.D., Strehlau, G.A., 1998b. ECOS Statistische Analyse vonOekotoxikologie-Daten, Version 1.1, Dokumentation und Manual.Novartis Services AG, Basel.

Fornelio, C.A., Caabeiro, R.F., Gonzalez, J.A., 1987. The mode of actionof some benzimidazole drugs on Trichinella spiralis. Parasitology 95,61–70.

K. Knauer et al. / Chemosphere 68 (2007) 1435–1441 1441

Forney, D.R., Davis, D.E., 1981. Effects of low concentrations ofherbicides on submersed aquatic plants. Weed Sci. 29, 677–685.

Gratzner, H.G., 1982. Monoclonal antibody to 5-bromo- and 5-iodode-oxyuridine: a new reagent for detection of DNA replication. Science218, 474–475.

Heringa, M.B.M., Screus, R.H.M.M., Busser, F., Van der Saag, P.T., Vander Burg, B., Hermens, J.L.M., 2004. Toward more useful in vitrotoxicity data with measured free concentrations. Environ. Sci. Technol.38, 6263–6270.

Jespers, A.B.K., Dewaard, M.A., 1995. Effect of fenpiclonil on phosphor-ylation of glucose in fusarium–sulphureum. Pesticide Science 44, 167–175.

Konemann, H., 1981. Quantitative structure–activity relationships in fishtoxicity studies, Part 1: Relationship for 50 industrial pollutants.Toxicology 19, 209–221.

Maeder, V., Escher, B.I., Scheringer, M., Hungerbuhler, K., 2004. Toxicratio as an indicator of the intrinsic toxicity in the assessment ofpersistent, bioaccumulative, and toxic chemicals. Environ. Sci. Tech-nol. 38, 3659–3666.

McCarty, L.S., Hodson, P.V., Craig, G.R., Kaiser, K.L.E., 1985. The useof quantitative structure activity relationships to predict the acute andchronic toxicities of organic-chemicals to fish. Environ. Toxicol.Chem. 4, 595–606.

McCarty, L.S., Mackay, D., Smith, A.D., Ozburn, G.W., Dixon, D.G.,1991. Interpreting aquatic toxicity QSARs – the significance oftoxicant body residues at the pharmacological end-point. Sci. TotalEnviron. 109, 515–525.

McCarty, L.S., Mackay, D., Smith, A.D., Ozburn, G.W., Dixon, D.G.,1993. Residue-based interpretation of toxicity and bioconcentrationqsars from aquatic bioassays – polar narcotic organics. Ecotoxicol.Environ. Safe. 25, 253–270.

Mosmann, T., 1983. Rapid colorimetric assay for cellular growth andsurvival: application to proliferation and cytotoxicity assays. J.Immunol. Methods 65, 55–63.

Navas, J.M., Segner, H., 2006. Vitellogenin synthesis in primary culturesof fish liver cells as endpoint for in vitro screening of the (anti)estrogenic activity of chemical substances. Aquat. Toxicol. 80, 1–2.

OECD Guidelines for Testing of Chemicals, 1984. Effects on BioticSystems. Method 203. Fish, Prolonged Toxicity Test. Adopted April 4,1984.

OECD Guidelines for Testing of Chemicals, 1992. Effects on BioticSystems. Method 203. Fish, Acute Toxicity Test. Adopted July 17,1992.

OECD Guidelines for Testing of Chemicals, 1996. Effects on BioticSystems. Method 305. Bioconcentration: Flow Trought Fish Test.Adopted June 14, 1996.

Rievere, J.L., Leroux, P., Bach, J., Gredt, M., 1984. Effects of someergosterol biosynthesis inhibting fungicides on sterols and cytchrome

P-450 from the Japanese quail Coturnix coturnix. Pesticide Sci. 15,317–323.

Russom, C.L., Bradbury, S.P., Broderius, S.J., Hammermeister, D.E.,Drummond, R.A., 1997. Predicting modes of toxic action fromchemical structure: acute toxicity in the fathead minnow (Pimephales

promelas). Environ.Toxicol. Chem. 16, 948–967.Saad, B., Scholl, F.A., Thomas, H., Schawalder, H., Streit, V., Waechter,

F., Maier, P., 1993. Crude liver membrane frations and extracellularmatrix components as substrate regulate differentially the preservationand inducibillity of cytochrome P-450 isoenzymes in cultured rathepatocystes. Eur. J. Biochem. 213, 805–814.

Saito, H., Iwami, S., Shigeoka, T., 1991. In vitro cytotoxicity of 45pesticides to goldfish gf-scale (gfs) cells. Chemosphere 23, 525–537.

Saito, H., Koyasu, J., Yoshida, K., Shigeoka, T., Koike, S., 1993.Cytotoxicity of 109 chemicals to goldfish GFS cells and relationshipswith 1-octanol–water partition coefficients. Chemosphere 25, 1015–1028.

Sandbacka, M., Christianson, I., Isomaa, B., 2000. The acute toxicity ofsurfactants on fish cells, Daphnia magna and fish – a comparativestudy. Toxicol. In Virto 14 (1), 61–68.

SAS, 1990. SAS/STAT User’s Guide, Version 6. SAS Institute, Cary, NC,USA.

Schirmer, K., 2006. Proposal to improve vertebrate cell cultures toestablish them as substitutes for the regulatory testing of chemicals andeffluents using fish. Toxicology 224, 163–183.

Segner, H., 1998. Fish cell lines as a tool in aquatic toxicology. In:Braunbeck, T., Hinton, D.E., Streit, B. (Eds.), Fish Ecotoxicology.Birkhauser, Basel, Switzerland, pp. 1–38.

Segner, H., 2004. Cytotoxicity assays with fish cells as an alternative to theacute lethality test with fish. ATLA 32, 375–382.

Segner, H., Braunbeck, T., 2003. End points for in vitro toxicity testingwith fish cells. In: Mothersill, C., Austin, B. (Eds.), In-Vitro Methodsin Aquatic Toxicology. Springer Praxis, Chichester, UK, pp. 77–141.

Segner, H., Lenz, D., 1993. Cytotoxicity assays with the rainbow trout R1cell line. Toxicol. In Vitro 7, 537–540.

Snegaroff, J., Bach, J., 1989. Effects of the fungicide prochloraz onxenobiotic metabolism in rainbow trout: inhibition in vitro and timecourse of induction in vivo. Xenobiotica 19, 255–267.

Tomlin, C.D.S. (Ed.), 1997. Pesticide Manual, 11th ed. British CropProtection Council, UK.

Vaes, W.H.J., Ramos, E.U., Verhaar, H.J.M., Cramer, C.J., Hermens,J.L.M., 1998. Understanding and estimating membrane/water parti-tion coefficients: Approaches to derive quantitative structure propertyrelationships. Chem. Res. Toxicol. 11 (8), 847–854.

Verhaar, H.J.M., van Leeuwen, C.J., Hermens, J.L.M., 1992. Classifyingenvironmental-pollutants 1. Structure–activity-relationships for pre-diction of aquatic toxicity. Chemosphere 25, 471–491.