Unraveling the mechanisms involved in zearalenone-mediated toxicity in permanent fish cell cultures

ilable at ScienceDirect

Toxicon 88 (2014) 44e61

Contents lists ava

Toxicon

journal homepage: www.elsevier .com/locate/ toxicon

Unraveling the mechanisms involved in zearalenone-mediated toxicity in permanent fish cell cultures

Constanze Pietsch a, b, *, Jürg Noser c, Felix E. Wettstein d,Patricia Burkhardt-Holm b, e

a Zurich University of Applied Sciences (ZHAW), Institute of Natural Resource Sciences (IUNR), Gruental, P.O. Box, CH-8820 Waedenswil,Switzerlandb Programm Man e Society e Environment, Department of Environmental Sciences, University of Basel, Vesalgasse 1, CH-4051 Basel,Switzerlandc Kantonales Laboratorium Basel, Gr€aubernstrasse 12, CH-4410 Liestal, Switzerlandd Agroscope Reckenholz-T€anikon (ART), Research Station ART, Reckenholzstrasse 191, CH-8046 Zürich, Switzerlande Department of Biological Sciences, University of Alberta, CW 405 Biological Sciences Building, T6G 2E9, Edmonton, AB, Canada

a r t i c l e i n f o

Article history:Received 8 March 2014Received in revised form 30 May 2014Accepted 10 June 2014Available online 17 June 2014

Keywords:MycotoxinCytotoxicityOxidative stressGenotoxicityEstrogenic action

* Corresponding author. Zurich University of ApplInstitute of Natural Resource Sciences (IUNR), GruenWaedenswil, Switzerland. Tel.: þ41 58 934 5613.

E-mail address: [email protected] (C. P

http://dx.doi.org/10.1016/j.toxicon.2014.06.0050041-0101/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

The world-wide occurrence of zearalenone (ZEN) as a contaminant in feed for farm ani-mals and fish requires the evaluation of toxicity mechanisms of action of ZEN. The presentstudy investigates possible metabolization of ZEN in fish cell lines suggesting that mainlyglucuronidation takes place. It demonstrates that concentrations up to 20,000 ng ml�1 ZENare capable of influencing cell viability in permanent fish cell cultures in a doseeresponsemanner with different response patterns between the five tested cell lines, whereby ly-sosomes appeared to be the main target of ZEN. ZEN toxicity is often discussed in thecontext of oxidative stress. Our study shows a biphasic response of the cell lines whenreactive oxygen species (ROS) production is monitored. Damage in cells was observed bymeasuring lipid peroxidation, DNA strand breaks, and alterations of intracellular gluta-thione levels. Metabolization of ZEN, especially at concentrations above 7500 ng ml�1 ZEN,does not prevent cytotoxicity. ZEN as an estrogenic compound may involve processesmediated by binding to estrogen receptors (ER). Since one cell line showed no detectableexpression of ER, an ER-mediated pathway seems to be unlikely in these cells. This con-firms a lysosomal pathway as a main target of ZEN in fish cells.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Mycotoxins are widely distributed in our environment.Most mycotoxins are relatively stable in aquatic ecosys-tems, and thus, relevant concentrations can be found insurface waters (Lagan�a et al., 2001; Spengler et al., 2001;

ied Sciences (ZHAW),tal, P.O. Box, CH-8820

ietsch).

Bucheli et al., 2008; Gromadzka et al., 2009; Kolpin et al.,2010). Mycotoxins in feed have been shown to possessmultifold detrimental effects on aquatic organisms (Postonet al., 1982; Santacroce et al., 2008; D€oll et al., 2011; Hooftet al., 2011; Sanden et al., 2012). Moreover, contaminationof animal feed with mycotoxins has been found to occur(Roberts and Patterson, 1975; Bryden et al., 1980;Abdelhamid, 1990; Juszkiewicz and Piskorska-Pliszczynska, 1992; Placinta et al., 1999; Ranjan and Sinha,1991). Consequently, mycotoxins have become a world-wide interest.

C. Pietsch et al. / Toxicon 88 (2014) 44e61 45

Zearalenone (ZEN), a resorcylic acid lactone, is animportant mycotoxin produced by fungi of the genusFusarium. Fusarium-derived mycotoxins have been identi-fied as aquatic micropollutants in surface waters (Bucheliet al., 2008; Hoerger et al., 2009). It has been reportedthat mycotoxins contribute to the estrogenic environ-mental exposure in aquatic systems (Hartmann et al.,2008). Especially ZEN has been shown to act as a typicalestrogenic compound showing high estrogenic potenciesvia estrogen receptors of rainbow trout (Bucheli et al.,2005) and effects on reproduction of zebrafish (Daniorerio) and early life stages of fathead minnow (Pimephalespromelas) have been described (Johns et al., 2009; Schwartzet al., 2010). Although metabolization of ZEN in fish hasrarely been described (Lagan�a et al., 2004), several me-tabolites of ZEN, namely a-ZEL and b-ZEL, have also estro-genic effects in fish (Matthews et al., 2000). Furthermetabolization of toxic compounds in phase II of thebiotransformation pathway in vertebrates often includesglucuronidation via UDP-glucuronosyltransferases (UGTs)(Meech and Mackenzie, 1997; Grancharov et al., 2001).Glucuronidation, which was observed for ZEN as well(Warth et al., 2013), is the combining with glucuronic acidthat leads to the formation of more hydrophilic derivatives.

The toxic effects of ZEN in mammals and fish have beendiscovered. Many studies assessed toxic effects of ZEN andits metabolites in mammalian cell lines and ZEN has beenfound to be hepatotoxic (Maaroufi et al., 1996; Obremskiet al., 1999; Conkov�a et al., 2001), immunotoxic (Marinet al., 1996; Berek et al., 2001), and genotoxic (Ouaneset al., 2003; Abid-Essefi et al., 2004; Lioi et al., 2004; Gaoet al., 2013). In fish, the mechanisms of toxicity of ZEN havenot yet been demonstrated but effects on reproduction,development, and iron metabolism were observed (Johnset al., 2009; Schwartz et al., 2010; Wo�zny et al., 2012;Bakos et al., 2013). However, a connection of the estro-genic potential and toxic effects of ZEN has not yet beeninvestigated in fish. Generally, binding of estrogenic com-pounds to estrogen receptors that function as ligand-activated transcription factors leads to their conforma-tional changes resulting in DNA binding and regulation ofgene transcription (MacGregor and Jordan, 1998). Thus,treatment with estrogenic compounds regulates theexpression of many genes including those which encodeproducts involved in cell survival and apoptosis (Tanzer andJones, 1997; Singer et al., 1998; Linford et al., 2001).

From a study using mammalian cell lines it has beenconcluded that it is unlikely that the toxicity of this sub-stance is solely due to its estrogenic potential because othermechanisms, such as oxidative stress, could be involved(Abid-Essefi et al., 2004). Generation of reactive oxygenspecies (ROS) is involved in important physiological func-tions of all aerobic organisms. However, an excess of ROSformation can lead to disturbances of metabolic pathways,depletions of cellular antioxidants, and damage to macro-molecules, including DNA, proteins and lipids. This oxida-tive injury can result in a reduction of differentiation andproliferation of cells or can lead to apoptosis (Abid-Essefiet al., 2004; Hassen et al., 2007; Bouaziz et al., 2008).However, in fish, the generation of oxidative stress due toexposure to ZEN has not yet investigated. Thus, possible

oxidative stress due to exposure to ZENwas analyzed in thepresent study using fish cell lines to show the involvementof this pathway in the toxicity of this mycotoxin evenoutside mammalian cell systems.

In aquatic toxicology, permanent fish cell cultures canbe used as a tool to generate initial information on themetabolism and effects of pollutants (Babich et al., 1986;Segner, 2004). Well-targeted experiments may indicatewhether metabolization of xenobiotics involves mecha-nisms of activation or deactivation and whether damage tobiological targets such as lipids, DNA, and proteins can beexpected. In the present study, five permanent fish celllines were used for the first time to evaluate the compar-ative cytotoxic effects of ZEN in vitro and to delineate itsdoseeresponse relationship. This was necessary to showcell line-specific differences in toxicity of ZEN. The con-centrations of ZEN have been chosen according to specifi-cations in the literature on high ZEN (up to 1500 mg kg�1) ingrains and feedstuff in Europe (Coppock et al., 1990;Veldman et al., 1992; Charmley et al., 1994; Whitlow andHagler, 2005; Mankevi�cien _e et al., 2007; Driehuis et al.,2008). Mycotoxin contents in fish feed have rarely beeninvestigated but recent data show that ZEN is highlyprevalent in ingredients and complete feedstuffs in aqua-culture (Måge et al., 2009; Santos et al., 2010; Pietsch et al.,2013). Thus, the aim of this study was to investigate thesensitivity of five fish cell lines, including cell lines derivedfrom different tissues of rainbow trout, carp and salmon,towards this fungal metabolite with respect to organ- andspecies-specific differences. Therefore, the acute toxic ef-fects and possible oxidative damage on these cell lineswereassessed as a function of ZEN concentration by usingcolorimetric and fluorescent dyes. Possible cellular damagewas further evaluated by assessing lipid peroxidation andDNA damages, and levels of reduced glutathione, as one ofthe most important antioxidants involved in phase II re-actions. The involvement of direct estrogenic pathways inZEN action was assessed by investigating the presence orabsence of estrogen receptors alpha or beta in each of thecell lines by means of semi-quantitative RT-PCR.

2. Materials and methods

2.1. Chemicals

All chemicals were obtained from SigmaeAldrich(Buchs, Switzerland) unless indicated otherwise. Zear-alenone (ZEN; purity � 99%; lot number 020M4056) wasdissolved in ethanol. Solvent concentrations in the expo-sition media did not exceed 0.1% (v/v). Stock solutions weremaintained in darkness at 4 �C.

2.2. Cell culture and exposure

All fish cell cultures and exposures were conducted at19 �C. The cells derived from rainbow trout liver (RTL-W1cells) and gills (RTgill-W1 cells), and salmon head kidney(SHK-1) cells were maintained in L-15 medium at 19 �Cwithout addition of CO2. The cells from epithelial gonadalcell line of rainbow trout (RT EQ clone 8; strain No. 994)were maintained in Dulbecco's modified Eagle medium

C. Pietsch et al. / Toxicon 88 (2014) 44e6146

diluted 1:1 (v/v) with Ham F12 medium (LuBioScience,Lucerne, Switzerland). Carp brain (CCB) cells were culturedin Eagle minimum essential medium containing Earle'ssalts. The cell culture media contained 10% fetal bovineserum (FBS) for subculturing with the exception of themedium for RTL-W1 cells which contained 5% FBS. All cellswere cultured, harvested, and seeded into 96-well micro-titre plates (TPP Techno Plastic Products AG, Trasadingen,Switzerland) at a concentration of 30,000 cells per well ashas been described by Pietsch et al. (2011). For exposure tochemicals cells were washed with Earle's medium and thereagent was added in serum-free Earle's medium for 24 h.All experiments were run in 3 independent replicates.

2.3. Measurement of cell viability and ROS production

After exposure to different ZEN concentrations mem-brane integrity was assessed using theMTTassay accordingto Mosmann (1983), the neutral red (NR) test (Babich andBorenfreund, 1991) and the propidium iodide method.The concentration values resulting in 50% reduction of cellsurvival (EC50) were calculated for these viability assays byregression of the percentage of the response to ZEN againstthe natural logarithm of the compound concentration. Theresulting regression equations were used to calculate theconcentration of compound that would result in 50%reduction of cell viability. ROS production was measuredusing the fluorescent dye H2DCF-DA (LuBioScience,Lucerne, Switzerland). In addition, nitroblue tetrazoliumchloride (NBT) tests were conducted. All assays have beendescribed in detail by Pietsch et al. (2011). Each test wasrepeated three times and each microtitre plate had its owncontrol treatment containing the solvent ethanol. Treat-ments on each plate were based on the mean of 8 replicatetreatments per ZEN concentration or control incubation.

2.4. Lipid peroxidation assay

Cell culture media containing 180,000 cells from eachcell line were distributed in each well of separate 24-wellplates (BD Falcon™ 24-well Multiwell Plate, Becton Dick-inson AG, Allschwil, Switzerland) leading to comparablecell densities to those that have been used in the 96-wellplate assays. Cells were allowed to adhere over night,washed with Earle's medium and treated with ZEN for 24 hat different concentrations in the same medium. A positivecontrol was treated with 1.5 mM H2O2 in parallel. There-after, cells were washed with phosphate buffered saline(PBS), followed by a washing step with 200 ml versene(LuBioScience, Lucerne, Switzerland) and harvested bytrypsinization. Trypsinization was terminated by additionof PBS containing 5% fetal calf serum (FCS). Cells werecentrifuged for 3 min at 10,000 g at 4 �C (Centrifuge 5415R,Eppendorf, Basel, Switzerland). The cell pellet was re-suspended in PBS and homogenized using an UltraTurraxT8 (IKA Werke, Staufen, Germany) for 45 s at maximumspeed. TBARs assay was conducted according to Rau et al.(2004) with the following modifications. A volume of200 ml cell homogenate were mixed with 400 ml TBARssolution (Holt et al., 1986) containing 3.75 mg ml�1 thio-barbituric acid (TBA), 20% (w/v) trichloric acid (TCA),

9.1 ml ml�1 hydrochloric acid (37%), 0.06% (w/v) butylatedhydroxytoluene (BHT), and 866.9 ml destilled water. Sub-sequently, all samples were heated at 70 �C for 90 min,cooled to room temperature, and centrifuged at 16,000 gfor 10 min at room temperature. Standards containingmalondialdehyde (MDA) concentrations ranging from 0 to3200 nMMDAwere prepared similarly. Optical densities ofall samples were read at 532 nm (Infinite M200, TecanGroup Ltd., M€annedorf, Switzerland). In parallel, aliquots ofthe cell homogenates were also used for protein de-terminations using the bicinchoninic acid (BCA) assay(Sigma) according to the manufacturer's protocol.

2.5. Comet assay

Genotoxicity was analyzed as described by McKelvey-Martin et al. (1993) with the following modifications.RTL-W1 cells were cultured in 24-well plates using180,000 cells per well. After adherence of cells wells treatedwith the different ZEN concentrations in duplicates andafter 24 h of incubation cells were harvested as describedabove. Two slides (Superfrost Plus, Gerhard Menzel GmbH,Braunschweig, Germany) coated with 90 ml of 1.0% normalmelting agarose (w/v; SeaKem LE Agarose, Biozym Scien-tific GmbH, Hessisch Oldendorf, Germany; prepared in PBS)were mounted with 0.7% low melting agarose (w/v; Sea-Plaque GTG Agarose, Biozym Scientific GmbH, HessischOldendorf, Germany; prepared in PBS) containing the har-vested cells from onewell at 37 �C. These slides were placedon cold metal plates for 5 min and dried at 37 �C for further5 min. Subsequently, additional 90 ml of 0.7% low meltingagarose were spread on the slides as a protective layer,cooled for 5 min and dried at 37 �C for further 5 min. Lysisof cells was accomplished by incubation of slides in amixture of 100 mM ethylene-diamine-tetraacetic acid(EDTA) disodium salt hydrate, 2.5 M NaCl, 1% (v/v) Triton X-100, 10% (v/v) dimethyl sulfoxide (DMSO), and 1% N-laur-oylsarcosine sodium salt at pH 10.0 for 3 h at 4 �C in thedark. Afterward, slides were electrophoresed at 25 V and0.3 A for 25 min in ice-cold alkaline buffer containing12 g L�1 NaOH, and 0.37 g L�1 EDTA using a horizontalelectrophoresis tank (Maxigel ECO 2, Apelex, Evry Cedex,France) followed by neutralization of slides by using400 mM TriseHCl (pH 7.4) for 2 min. Until staining with1.56 mg ml�1 ethidium bromide slides were kept at 4 �C inthe dark. Individual comets were viewed using a Nikonepifluorescence microscope (Nikon Eclipse E400 equippedwith a Nikon Digital Camera DXM1200F; Nikon AG, Egg,Switzerland) at a 200� magnification. Image analysis wasdone using the software TriTek CometScore™ v 1.5 (TriTek,Sumerduck, USA) and DNA damages were quantified asincreases in tail moments (the product of the DNA per-centage in the tail and the tail length) which are indicatorsof damage that is proportional to the number of strandbreaks in cells.

2.6. Measurement of reduced and oxidized glutathione

Cells from each cell line were cultured in 24-well plates,treated and harvested as described above. The resultingcell pellets were suspended in homogenization buffer

C. Pietsch et al. / Toxicon 88 (2014) 44e61 47

containing 0.1 M potassium phosphate (pH 6.8), 154 mMKCl, and 5 mM diethylenetriaminepentaacetic acid (DTPA).These cell suspensions were homogenized for 15 s on ice(UltraTurrax T8, IKA Werke, Staufen, Germany) and celldebris was removed by centrifugation at 16,000 g for 3 min(Centrifuge 5415R, Eppendorf, Basel, Switzerland). Sampleswere then mixed with the same volume of ice-cold TCAbuffer [40 mM HCl, 10 mM DTPA, 20 mM ascorbic acid, 10%(v/v) trichloroacetic acid (TCA)]. These mixtures were pu-rified by centrifugation using Ultrafree® Centrifugal Filters(Millipore™, distributed by Carl Roth AG, Karlsruhe, Ger-many) and stored at �80 �C until analyses. Reducedglutathione (GSH) was analyzed according to Senft et al.(2000) with the modifications for a microplate assay asdescribed by Siraki et al. (2004). Therefore, 90 ml of the cellsamples were mixed with 40 ml 1 M potassium phosphatebuffer (pH 6.8). In parallel, additional wells were treatedwith 40 ml of 7.5 mM N-ethylmaleimide (NEM) in 1 M po-tassium phosphate buffer. After 15min of incubation,100 mlo-phthalaldehyde (OPA, 37.5 mM in 0.1 M potassiumphosphate buffer, pH 6.8) were added. Fluorescence emis-sion at 430 nm was read after 30 min of incubation in thedark using excitation at 345 nm. GSH standards were pre-pared from a 1 mM stock solution in redox-quenchingbuffer (RQB) containing 20 mM HCl, 5 mM DTPA and10 mM ascorbic acid, pH 6.8. Traces of glutathione disulfide(GSSG) were removed by addition of 25 mg zinc powderfollowed by centrifugation at 16,000 g for 3 min (Centrifuge5804 R, Eppendorf, Basel, Switzerland). Serial dilutionsfrom this GSH stock solution were made using RQB con-taining 5% (v/v) TCA. In parallel, GSSGwas analyzed from allsamples by incubation of 90 ml of the cell extracts with 40 mlof 7.5 mM N-ethylmaleimide (NEM). Subsequently, 10 mldithionite (DT) or RQB (as a blank) were added to the wellsand the plates were incubated for 60 min at room tem-perature in the dark. Thereafter, 100 ml OPA solution wereadded and fluorescence values weremeasured as describedabove. GSSG dilutions in RQB containing 5% (v/v) TCAwereused for the establishment of the standard curves.

2.7. Analyses of ZEN in cell culture media and cells

Since ZEN is mainly metabolized in liver (Kiessling andPettersson, 1978; Lagan�a et al., 2004) initial investigationswere conducted using the RTL-W1 cell line derived fromhepatic tissue of rainbow trout. In order to analyze distri-bution of ZEN between media solutions and cells in aninitial study, adherent RTL-W1 cells from cell culture flaskswere washed with Earle's medium, exposed to 300 ng ml�1

ZEN for 24 h, and harvested by trypzinization. Aftercentrifugation cell pellets were directly extracted by addi-tion of 1 ml acetonitrile for 1 min, followed by centrifuga-tion and transfer of 500 ml of the supernatants to HPLC vials.Volumes of 25 ml of the internal deuterated ZEN (D6-ZEN)standard were added to cell extracts. All samples wereevaporated to complete dryness under a gentle air streamat 40 �C. Subsequently, 1 ml of a water-acetonitrile-mixture(80/20; v/v) was added to the media samples. After that allsamples were mixed for 10 s. For determination of zear-alenone (ZEN) and its metabolites zearalanone (ZAN),alpha-zearalanol (a-ZAL), beta-zearalanol (b-ZAL), and

zearalenol (ZEL), which play a role duringmetabolization ofZEN in higher vertebrates (Schaut et al., 2008; Abid-Essefiet al., 2009), a HPLC-negative electrospray ionization(-ESI)-MS/MS analytical method was used as described byHartmann et al. (2007). Quantification of ZEN was enabledby using the internal standard method with water/aceto-nitrile solutions (80/20; v/v), containing increasingamounts of ZEN and constant amounts of D6-ZEN. ZEN(�99% purity), ZAN (�98% purity), alpha-ZAL (�97% pu-rity), beta-ZAL (�98% purity), alpha-ZEL (�98%) and beta-ZEL (�95% purity) were purchased from Sigma (Buchs,Switzerland). The deuterated internal standards D4-alpha-ZAL, D4-beta-ZAL, D4-alpha-ZEL (�95% purity) and D4-beta-ZEL (�95% purity) were purchased from RIVM (Bilt-hoven, The Netherlands). D6-ZEN was prepared by base-catalyzed hydrogenedeuterium exchange on native ZENas described in Miles et al. (1996). The purity of the internalstandard for ZEN (D6-ZEN) was tested by scanning themasses of ZEN and Dn-ZEN (n¼ 1e9). The relative amountswere 0.16%, 0.10%, 0.83%, 6.16%, 25.07%, 38.32%, 23.13%,5.50%, 0.70% and 0.03% for ZEN, D1-ZEN, D2-ZEN, D3-ZEN,D4-ZEN, D5-ZEN, D6-ZEN, D7-ZEN, D8-ZEN, and D9-ZEN,respectively. Ratios of the different grades of deuteratedZEN's did not change over time. Therefore, D6-ZEN wasfound to be a suitable internal standard and the otherdeuterated ZEN products did not influence quantification.

Instrument linearity was determined with ZEN calibra-tion solutions prepared in Earle's medium. The procedureto determine method/instrument precision and methoddetection/quantification limit is described in Bucheli et al.(2008). The ion suppression was determined by themethod of standard addition (Freitas et al., 2004). An ionsuppression for ZEN of 17% was calculated (which wascompensated for by normalization to D6-ZEN). The limit ofdetection (LOD) obtained from the signal to noise ratio wasbetween 1.5 and 5 ng ml�1 ZEN, whereas a limit of quan-titation (LOQ) was between 4.5 and 8 ng ml�1 ZEN,depending on the analyzed matrix. Quality control pa-rameters were determined in Earle's medium. Involvementof phase II reactions in metabolization of ZEN in fish cellswas investigated in single exposures exemplarily after in-cubation of only RTL-W1 cells with 300 ng ml�1 ZEN inEarle's medium for 24 h. De-glucuronidation was achievedusing b-glucuronidase from Helix pomatia according toFrandsen (2007) with the following modifications. Sodiumacetate solution was added to the remaining cell mediasamples from all cell cultures to a final concentration of100 mM, the pH was adjusted to 5.5, and 8.8 U ml�1 b-glucuronidase in 0.2% NaCl were supplemented. Subse-quently, samples were mixed and incubated for 24 h at37 �C. The de-glucuronidation was stopped by addition ofethanol, followed by centrifugation and storage of the su-pernatant at �20 �C. Since the b-glucuronidase from mol-lusks also shows sulfatase activity the enzymatic de-conjugation was also conducted at pH 6.2 to favor thisenzymatic reaction. In parallel, acid hydrolysis was con-ducted as described by Frandsen (2007).

In a separate experiment high performance liquidchromatography with diode array detection (HPLC-DAD)was used to quantify ZEN metabolization in all cell lines. Toassess the amount of free ZEN and its conjugates with

C. Pietsch et al. / Toxicon 88 (2014) 44e6148

glucuronic acid produced by each cell line, cells weretreated as follows. A volume of 5 ml cell suspension wascultured at initial concentrations of 225,000 cells per ml incell culture flasks (25 cm2) over night. Subsequently, cellswere washed with serum-free Earle's medium and ZENsolutions with concentrations ranging from 0 to20,000 ng ml�1 ZEN in Earle's medium were applied induplicate for 24 h. To confirm nominal exposure concen-trations and ZEN stability analytically mycotoxin concen-trations in exposure media were verified by samplingfreshly prepared solutions before exposure and mediadirectly from the cell culture flasks after exposure of cells.Non-adherent cells were removed from medium samplesby centrifugation at 300 g for 3 min and added to thesamples from adherent cells harvested directly after sam-pling of the cell media to assess ZEN concentrations in cells.One part of media samples was mixed with the same vol-ume acetonitril and stored at �20 �C before analyses. Fromeach cell culture flask cells were also harvested by trypsi-nization which was stopped by addition of PBS containing5% FCS. Cells were then centrifuged and resuspended inPBS. One part of this suspension was supplemented withacetonitril, incubated for 15 min with the same volume ofacetonitril, centrifuged at 300 g for 3 min and stored at�20 �C for later analyses. The other part of the cell sus-pension was incubated with b-glucuronidase for 24 h asdescribed above before addition of acetonitrile as has beendescribed for the preparation of cell media samples.

Analyses were performed on a HPLC-DAD (high per-formance liquid chromatography with diode array detec-tion) by fluorescence detection on a HP 1046A using aLiChrospher 100 RP8 column (250 mm) with a cutoff filterat 407 nm. The mobile phase consisted of acetonitrile andnanopurewater (40:60, v:v), reaching a ratio of 60:40, (v:v)after 20 min, followed by a ratio of 95:5 (v:v) for further5 min at a flow rate of 1 ml min�1. The excitation wave-length of the fluorescence detector was set at 274 nm,while the emission wavelength was 466 nm. The limit ofdetection was 25 ng ml�1 and the mean recovery wasapproximately 88%.

2.8. RNA isolation, PCR and sequencing of ERa and ERb fromcell cultures and liver tissue of fish

Cells were grown to confluence in cell culture flasks(TPP Techno Plastic Products AG, Trasadingen,Switzerland), and harvested by trypsinization. Thereafter,cells were washed with 1 ml PBS and frozen at �80 �C in300 ml TRIzol (Invitrogen AG, Basel, Switzerland). RNA wasisolated from cell culture samples and snap-frozen liversamples of carp and rainbow trout according to B€ogi et al.(2002) using a phenol-chloroform method including aDNase treatment according to the manufacturer's protocol(DNA-free™ kit, Ambion®, distributed by Life TechnologiesEurope B.V., Zug, Switzerland). Total RNA was reversetranscribed into cDNA using the High-Capacity cDNAReverse Transcription Kit (Applied Biosystems distributedby Life Technologies Europe B.V., Zug, Switzerland) ac-cording to the manufacturer's directions. Two microgramsof the resulting cDNA were diluted in nuclease-free waterand used for semiquantitative PCR using the FastStart Taq

DNA Polymerase dNTPack (Roche Applied Science, Basel,Switzerland). Oligonucleotid primers for PCR measure-ments were designed based on known sequences for fishERa and ERb (Supplementary Table) and were obtainedfrom Microsynth (Balgach, Switzerland). All primers werevalidated for PCR by sequencing after gel extraction of PCRproducts according to the manufacturer's protocol (QIA-quick gel extraction kit, Qiagene, USA) followed by directsequencing with the BigDye sequencing chemistry(Applied Biosystems), analysis on an ABI 3130xl geneticanalyzer (Applied Biosystems) and subsequent comparisonby Basic Local Alignment Search Tool (BLAST) using thedatabases at the National Center for Biotechnical Informa-tion (NCBI) at the National Library of Medicine. Polymerasechain reaction were performed in Taq polymerase bufferusing 1 U Taq polymerase, MgCl2 (1.5 mM), dNTPs(200 mM), and primers (200 nM) in a total volume of 20 ml.PCR conditions included 4 min at 94 �C, followed by 40 s at94 �C, 40 s at annealing temperature and 40 s at 72 �C, and afinal extension for 10 min at 72 �C using a PCR system(Eppendorf Mastercycler EP Gradient S 96, Eppendorf,Basel, Switzerland). PCR reactions were optimized withrespect to the selection of the annealing temperature. ThePCR products were transferred to 3% agarose gel containing0.005% ethidium bromide (stock solution 10 mg ml�1, v/v;Carl Roth AG, Karlsruhe, Germany) and after horizontal gelelectrophoresis in TAE buffer (242 g TRIS-HCl, 37.2 g so-dium EDTA, 57.1 ml glacial acetic acid in 1 L distilled water,pH 7.5) for 35 min at 100 mV PCR products were visualizedusing AlphaImager®EP (Alpha Innotech distributed byBiozym Scientific GmbH, Hessisch Oldendorf, Germany).

2.9. Statistial analyses

Data of ROS measurements, cell viability, lipid peroxi-dation, glutathione levels and comet assay determinationsare presented as the mean ± standard error of the mean(SEM) from 3 independent cell cultures. Effects of treat-ments were determined by comparison of treatmentgroups to controls using Friedman test followed by Wil-coxon test (SPSS 16.0 for Windows). A p value of <0.05 wasaccepted as being statistically significant.

3. Results

3.1. ZEN concentrations in exposure media and cell extracts

The experiments with RTL-W1 cells showed that ZEN islargely not detectable as free ZEN by LC-MS/MS in exposuremedia after exposure to 300 ng ml�1 ZEN for 24 h (Table 1).Metabolites of ZEN, namely zearalanone (ZAN), alpha-zearalanol (a-ZAL), beta-zearalanol (b-ZAL), and zear-alenol (ZEL) could not be identified in media after cellexposure. Enzymatic de-conjugation increased the detect-able amount of free ZEN in exposure media to 76e82% ofthe amount that was initially applied to the cells, depend-ing on the pH that was used for the de-conjugation. Thissuggests that ZEN had been taken up by these cells and wasreleased to the surrounding medium as ZEN-glucuronide.Acid hydrolysis as described by Frandsen (2007) was notvaluable to assess glucuronidation of ZEN since it only

Table 1ZEN in cell media from RTL-W1 cells for 24 h with 300 ng ml�1 ZEN.Samples were treated with b-glucuronidase prior to analyses by LC-MS/MS.

ZEN ng ml�1

ZEN in Earle's medium without cell exposure 272ZEN after 24 h cell exposure 20ZEN after 24 h cell exposure treated with

b-glucuronidase pH 5.5223

ZEN after 24 h cell exposure treated withb-glucuronidase pH 6.2

208

ZEN after 24 h cell exposure treated by acidhydrolysis

29

C. Pietsch et al. / Toxicon 88 (2014) 44e61 49

marginally increased the amount of free ZEN in thesamples.

Analyses of ZEN in cell culture media by HPLC-DAD fromall cell lines after 24 h of incubation with this mycotoxin(Fig. 1) showed that in all cell media exposure to ZEN for24 h reduced the detectable amount of free ZEN comparedto the initial media concentrations. Especially, initial con-centrations of less than 625 ng ml�1 ZEN were no longerdetectable after 24 h of exposure to cells. The fate of ZEN athigher concentrations, however, differed between the celllines. When applied to the gill cell line an average of 24% ofthe initial applied ZEN concentration could be recovered inthe medium after 24 h, whereas the other cell lines derivedfrom rainbow trout (RTL-W1 and RT EQ clone 8 cells)showed decreasing ZEN concentrations in the mediumaveraging between 35 and 38% of the initial exposure me-dium concentration. In the medium of SHK-1 cells and inCCB cells averages of approximately 50% of the initiallyapplied ZEN concentrations could be detected after 24 hand especially at ZEN concentrations above 10,000 ng ml�1

the capacity to remove ZEN from the exposure mediumdecreased considerably in these cell lines.

0

25

50

75

100

125

150

ZEN

[ng

ml-1

]

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

5

nominal ZEN con

RTgill-W1 ETR1W-LTR

Fig. 1. ZEN concentrations in medium and cell extracts after 24 h of exposure in theCCB cells) expressed as percentages of the exposure media before incubation with

ZEN analyses from cell extracts showed different results(Fig. 1). At exposure medium concentrations of 313 ng ml�1

ZEN could rarely be detected in any of the cell lines. Inaddition, at higher exposure medium concentrations, onlylow ZEN amounts were accumulated in RT EQ clone 8 cellsand CCB cells (showing average values of 20% and 27%,respectively). SHK-1 cells showed an average ZEN value of48% at exposure concentrations higher than 313 ng ml�1

ZEN. The RTL-W1 and RTgill-W1 cells showed a linear trendbetween the internal cell concentrations and the initialmycotoxin concentrations in the exposure media that wereapplied. Especially at exposure medium concentrationsabove 7,500 ng ml�1 ZEN the mycotoxin concentrations inthese cell extracts amounted between 82 and 85% of theinitial concentration of the surrounding medium indicatingthat ZEN is taken up and accumulated to a certain extent inthese cell lines.

De-glucuronidation for 24 h improved the ZEN detec-tion by HPLC-DAD in the exposure media with initial con-centrations of less than 625 ng ml�1 ZEN (Fig. 2) whichsuggests that at low ZEN concentrations the mycotoxin isconjugated in the cells and released to the surroundingmedium rapidly. Increased ZEN concentrations were alsoobserved in media from RTL-W1 and RTgill-W1 cells afterincubationwith the highest ZEN concentration followed byde-glucuronidation (þ53% and þ184%, respectively)compared to samples without enzymatical de-conjugation.In cell extracts less ZEN was detectable in samples treatedwith b-glucuronidase compared to untreated cell extractswhen low ZEN concentrations were used. Less ZEN wasespecially found after de-glucuronidation of cell extractsfrom RTL-W1 and RTgill-W1 cells treated with ZEN. How-ever, concentrations of 5,000 ng ml�1 or higher oftenresulted in higher detection of free ZEN in extracts from theother cell lines when glucuronidase treatment was appliedcompared to extracts without de-glucuronidation. This

1025

2500

5000

7500

1000

015

000

2000

0

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

centrations [ng ml-1]

medium cells

BCC1-KHS8enolcQ

permanent fish cells cultures (RTL-W1, RTgill-W1, RT EQ clone 8, SHK-1, andcells.

0

25

50

75

100

125

150

ZEN

[ng

ml-1

]

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

EtO

H31

362

510

2525

0050

0075

0010

000

1500

020

000

medium cells

nominal ZEN concentrations [ng ml-1]

RTgill-W1 BCC1-KHS8enolcQETR1W-LTR

Fig. 2. ZEN concentrations in medium (A) and cell extracts (B) after 24 h of exposure in the permanent fish cells cultures (RTL-W1, RTgill-W1, RT EQ clone 8, SHK-1, and CCB cells) followed by b-glucuronidase treatment for 24 h expressed as percentages of the exposure media before ZEN incubation of cells.

C. Pietsch et al. / Toxicon 88 (2014) 44e6150

de-glucuronidation led to 30e33% higher ZEN levelsdetected in cell extracts when RT EQ clone 8, SHK-1 or CCBcells were treated with 20,000 ng ml�1 ZEN for 24 h.

3.2. Influence on cell viability

Cell viability studies demonstrated that after 24 h ofincubation with the mycotoxin ZEN a biphasic responsewas observed in RTL-W1, SHK-1 and in CCB cells in NRassays, whereas the same assay in RTgill-W1 and RT EQclone 8 cells showed only a decreased NR uptake at con-centrations above 625 ng ml�1 ZEN and 1250 ng ml�1 ZEN,respectively (Fig. 3).

Cell viability was also measured using the MTT assaydetermining membrane integrity and mitochondrial activity(Fig. 4). This assay revealed a biphasic response in all celllines which was especially pronounced in CCB cells (Fig. 4E).

Membrane integrity of cells was also investigated usingthe PI assay showing strongly increased membrane dam-ages at concentrations higher than 7,500 ng ml�1 ZEN in allcell lines (Fig. 5). This assay also revealed that the loss ofmembrane integrity was very pronounced in RT EQ clone8 cells which showed increased damages to membranes atall ZEN concentrations used. Looking at the EC50 values (theconcentration values resulting in 50% reduction of cellresponse in control incubations) for these tests, theassessment of membrane integrity using these assaysyielded different results (Table 2). The NR assays yieldedlower EC50 values than the MTT assays. EC50 values for thePI assays were found to be quite similar for all fish cell linesand showed higher values than for the NR assay.

3.3. Influence on ROS production

The results from the NBT assay showed a biphasicresponse to ZEN were obtained using for most cell lines

(Fig. 6). Similarly, a biphasic effect of ZEN on ROS produc-tion in all cell cultures was also measurend by the fluo-rescent dye H2DCF-DA (Fig. 7). The ROS production aftertreatment of cells with low concentration of ZEN rangingfrom 313 to 625 ng ml�1 ZEN was found to be significantlyreduced with the exception of SHK-1 cells. All cell linesshowed increased ROS production at moderate ZEN con-centrations and again decreased oxidative stress at highZEN concentrations.

3.4. Lipid peroxidation

The measurement of MDA as an indicator of lipid per-oxidation in cell extracts showed that in RTL-W1 cellstreatment with ZEN led to reduced values of lipid peroxi-dation. In RTgill-W1 cells and SHK-1 cells a similar reduc-tion of lipid damages was observed at ZEN concentrationsbetween 1,250 and 15,000 ng ml�1 ZEN but not at20,000 ng ml�1 ZEN. The lack of significance at exposureconcentrations of 20,000 ng ml�1 ZEN in these two celllines is probably influenced by low protein concentrationsin the samples which were used for MDA calculationscaused by high cytotoxic effects of the mycotoxin. Incontrast, RT EQ clone 8 cells showed reducedMDAvalues atZEN concentrations between 1,250 and 5,000 ng ml�1 ZEN,but increased lipid peroxidation at 10,000 and20,000 ng ml�1 ZEN (Fig. 8c). Similarly, ZEN-treated CCBcells showed lower MDA values at 313 ng ml�1 ZEN andincreased lipid peroxidation at mycotoxin concentrationsabove 5,000 ng ml�1 ZEN (Fig. 8E).

3.5. DNA damage

ZEN-treated RTL-W1 cells were heavily damaged at allZEN concentrations compared to untreated control cellswhen analyzed using the alkaline comet assay (Fig. 9). At

0

20

40

60

80

100

120

140

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

D

*****

****

**

***

*****

***

ZEN (ng/ml)

0

20

40

60

80

100

120

140Vi

talit

y(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

A***

***

**

***

***

***

***

ZEN (ng/ml)

0

20

40

60

80

100

120

140

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

ZEN (ng/ml)

B

*****

*

***

*** ******

0

20

40

60

80

100

120

140

Vita

lity

(%co

ntr o

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

C

***

***

***

*****

ZEN (ng/ml)

0

20

40

60

80

100

120

140

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

*****

****

*** **

****

** E

ZEN (ng/ml)

Fig. 3. Membrane integrity and lysosomal activity determined by uptake of neutral red uptake by permanent fish cells cultures (A ¼ RTL-W1, B ¼ RTgill-W1,C ¼ RT EQ clone 8, D ¼ SHK-1, E ¼ CCB) after incubation with ZEN for 24 h, mean ± SE, difference to solvent controls: * ¼ P < 0.05; ** ¼ P < 0.01, *** ¼ P < 0.001.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C. Pietsch et al. / Toxicon 88 (2014) 44e61 51

the highest ZEN concentrations the tail moments areincreased 23-folde26-fold compared to cells treated withthe solvent only. The 50% effect concentration of ZEN forDNA damages in RTL-W1 cells was found to be2009 ng ml�1 ZEN.

3.6. Glutathione contents

The values for reduced glutathione (GSH) weredecreased by treatment of RTL-W1 cells with ZEN con-centrations lower than 1250 ng ml�1 ZEN (Fig. 10A). Athigh ZEN concentrations both, GSH and GSSG, increased inthese cells. A decrease of GSH concentrations in cellstreated with high ZEN concentrations was observed inRTgill-W1 and CCB cells, although GSSG levels were not

significantly influenced. The effects of ZEN on glutathionelevels were not as pronounced in RT EQ clone 8 andSHK-1 cells.

3.7. PCR and sequencing of ERa and ERb from cell cultures

Both, estrogen receptor alpha (ERa) and the estrogenreceptor beta isoforms (ERb1 and ERb2), have been iden-tified in normal liver tissue of carp but not in the samplefrom the CCB cell culture (see pictures in the Supplement).In contrast, ERb has been found in trout liver tissue and inRT EQ clone 8 cells and in SHK cells while it was absent inRTL-W1 cells and RTgill-W1 cells. ERa1 and ERa2 wereidentified in all samples of rainbow trout origin but not inthe SHK-1 cell line.

0

50

100

150

200

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

D

***

*****

** **

***

**

ZEN (ng/ml)

0

50

100

150

200

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

***

***

***

**

*** *

E

ZEN (ng/ml)

0

50

100

150

200Vi

talit

y(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

A

*****

*

***

***

***

** * *

ZEN (ng/ml)

0

50

100

150

200

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

****

***

***

***

***

****

B

ZEN (ng/ml)

0

50

100

150

200

Vita

lity

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

C

***

***

***

***

***

*

ZEN (ng/ml)

Fig. 4. MTT conversion by different cells lines (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8, D ¼ SHK-1, E ¼ CCB) after incubation with ZEN for 24 h, mean ± SE;difference to solvent controls: * ¼ P < 0.05, ** ¼ P < 0.01, *** ¼ P < 0.001.

C. Pietsch et al. / Toxicon 88 (2014) 44e6152

4. Discussion

Small amounts of ZEN can rapidly be metabolized bypermanent fish cell lines. ZEN was found to be metabolizedto zearalanone (ZAN), alpha-zearalanol (a-ZAL), beta-zearalanol (b-ZAL), and zearalenol (ZEL) in a mammaliancell line (Schaut et al., 2008; Videmann et al., 2008). How-ever, these metabolites of ZEN could not be detected in RTL-W1 cell cultures after exposure to ZEN for 24 h. This in-dicates that metabolization of ZEN to these derivatives hasnot taken place in these cells after this exposure period. Theopposite was found in vivowhereby ZENwasmetabolized toa-ZEL and b-ZEL in trout liver (Lagan�a et al., 2004) and to a-ZEL in carp (Pietsch, unpublished data) which implies thatthe ability to hydroxylate ZEN via 3a- and 3b-hydroxysteroiddehydrogenases (Olsen et al., 1981) has been lost in the RTL-W1 cell line that has been used in the present study.

Wewere able to show that glucuronidationmediated byUDP-glucuronosyltransferases (UGTs) plays a role inmetabolization of ZEN in fish cells which has also beenshown for mammalian systems (Pfeiffer et al., 2010). Mul-tiple UGTs have been identified in fishes which emphasizesthe importance of this biotransformation pathway(Pesonen and Andersson, 1987; George and Taylor, 2002).However, our investigations indicate that glucuronidationis probably not the only biotransformation process thattakes place in ZEN-treated fish cells since de-glucuronidation often led to less detectable free ZEN thanexpected. This is especially evident in the three cell linesderived from rainbow trout where the sum of detectableZEN in media and cell extracts results in less ZEN thaninitially applied. The decrease in detection of ZEN is morepronounced in cell extracts which suggests that biotrans-formation enzymes further metabolized the ZEN that was

0

500

1'000

1'500

2'000

PIflu

ores

cenc

e(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

***

***

**

A

ZEN (ng/ml)

0

500

1'000

1'500

2'000

PIflu

ores

cenc

e(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

*** **

**** B

ZEN (ng/ml)

0

500

1'000

1'500

2'000

PIflu

ores

cenc

e(%

cont

r ol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

C

***

***

***

***

***

***

***

** **

ZEN (ng/ml)

0

500

1'000

1'500

2'000

PIflu

ores

cenc

e(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

*****

****

**

D

ZEN (ng/ml)

0

500

1'000

1'500

2'000

PIflu

ores

cenc

e(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

E

***

***

***

***

***

******

ZEN (ng/ml)

Fig. 5. Influence of ZEN on membrane integrity measured as PI fluorescence in permanent fish cell cultures (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8,D ¼ SHK-1, E ¼ CCB) for 24 h, mean ± SE; difference to solvent controls: * ¼ P < 0.05, ** ¼ P < 0.01, *** ¼ P < 0.001.

C. Pietsch et al. / Toxicon 88 (2014) 44e61 53

de-glucuronidated to other metabolites within 24 h. Theinvolvement of undetected ZEN metabolites is also indi-cated by the fact that the analysis of media and cell extractsdoes not lead to 100% recovery of the initial applied con-centrations. This is supported by the observation that ab-sorption of ZEN to culture flasks can be excluded because

Table 2EC50 values of cell viability assays (NR, MTT and PI) in fish cell culturesafter exposure to ZEN for 24 h.

NR50 ng ml�1 MTT50 ng ml�1 PI50 ng ml-

RTL-W1 8,973 17,903 12,104RTgill-W1 2,837 >20,000 11,693RT EQ clone 8 4,819 >20,000 11,828SHK-1 4,197 >20,000 11,296CCB 3,828 5,046 12,386

the incubation of ZEN in culture flasks without cells for 24 hdid not markedly influence ZEN recovery in the exposuremedia (Table 1). Thus, it is assumed that other metabolitesof ZENmay have been formed that could not be detected byour HPLC-DAD method. Further well-targeted experimentsshould indicate whether these possible metabolites mayinclude catechol derivatives of ZEN that are formed by ar-omatic hydroxylation and which are known for theiroxidative DNA damage (Fleck et al., 2012). This may be amajor explanation for the cytotoxic effects that have beenfound for ZEN in the present study. Cytotoxic effects of ZENwhich were comparable to those observed in mammaliancell lines (Abid-Essefi et al., 2004; Kouadio et al., 2005)could be confirmed in the present study using fish cell linesbymeans of three different cell viability assays in our study.The results obtained in this study using in vitro modelsshow that fish cells are sensitive to fungal metabolites such

0

50

100

150

200

250R

OS

prod

uctio

n(%

cont

r ol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

*****

*

**

A

ZEN (ng/ml)

0

50

100

150

200

250

RO

Spr

oduc

tion

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

B

***

***

**

**

**

ZEN (ng/ml)

0

50

100

150

200

250

RO

Spr

oduc

tion

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

*****

*

** ***

C

ZEN (ng/ml)

0

50

100

150

200

250

RO

Spr

oduc

tion

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

D

***

*

*

ZEN (ng/ml)

0

50

100

150

200

250

RO

Spr

oduc

tion

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

**

***

***

***

E

ZEN (ng/ml)

Fig. 6. Influence on the ROS production as determined by NBT assay in several permanent fish cells (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8, D ¼ SHK-1,E ¼ CCB) due to exposure to ZEN for 24 h, mean ± SE; difference to solvent controls: * ¼ P < 0.05, ** ¼ P < 0.01, *** ¼ P < 0.001.

C. Pietsch et al. / Toxicon 88 (2014) 44e6154

as ZEN in a concentration-dependent manner which is thefirst report of cytotoxic effects of this fungal metabolite onpermanent fish cell cultures. The cytotoxicity assessmentusing neutral red (NR) correlatedwell with results obtainedwith the tetrazolium salt reduction (MTT) techniquealthough the latter method showed higher EC50 values.Still, the EC50 values calculated for the MTT assay for theRTL-W1 (17,903 ng ml�1 ZEN) and the CCB cells(5,046 ng ml�1 ZEN) are lower than most values reportedfor mammalian cell lines which certainly is also due to thefact that 10% FBS is used for ZEN exposure in mostmammalian systems while the exposure media in thepresent study did not contain FBS. For example, the cyto-toxicity of ZEN assessed by this assay after 24 h exposure ofhuman HepG2 cells to ZEN in medium containing 10% FBS

ranged from 12,735 to 57,306 ng ml�1 (Hassen et al., 2007;Ayed-Boussema et al., 2008). Cetin and Bullerman (2005)also used cell culture media containing 10% FBS and re-ported IC50 values (the concentration values resulting in50% inhibition of cell response) of more than90,000 ng ml�1 ZEN for different mammalian cell lines(CHOeK1, Caco-2, V79 and HepG2) after 48 and 72 h ofexposure to this mycotoxin, whereas C5eO cells showed avalue of 24,000 ng ml�1 ZEN at 72 h of exposure to ZEN. Incontrast, Kouadio et al. (2005) showed lower IC50 valuesafter 72 h of exposure to concentrations ranging from 4,775to 7,959 ng ml�1 ZEN for similar treated Caco-2 cells grownin media containing 10% FBS. Moreover, an insect cellline incubated with ZEN showed an EC50 of 5,571 ng ml�1

for the cytotoxicity as measured by the MTT assay

0

50

100

150

200

250

300

fluor

esce

nce

unit

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

D

** ** *** *

* ***

***

ZEN (ng/ml)

0

50

100

150

200

250

300flu

ores

cenc

eun

its(%

cont

rol)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

*****

******

**

A

ZEN (ng/ml)

0

50

100

150

200

250

300

fluor

esce

nce

units

(%co

ntr o

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

* *

***

**

***

*

B

ZEN (ng/ml)

0

50

100

150

200

250

300

fluor

esce

nce

units

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

C

**

***

ZEN (ng/ml)

0

50

100

150

200

250

300

fluor

esce

nce

units

(%co

ntro

l)

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

E

*****

****

***

***

***

****

ZEN (ng/ml)

Fig. 7. ROS production by different cells lines (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8, D ¼ SHK-1, E ¼ CCB) after exposure to ZEN for 24 h measured byfluorescence of oxidized DCF followed by emission at 535 nm after excitation at 485 nm, mean ± SE, difference to solvent controls: * ¼ P < 0.05, ** ¼ P < 0.01,*** ¼ P < 0.001.

C. Pietsch et al. / Toxicon 88 (2014) 44e61 55

(Fornelli et al., 2004). However, since the absence of FBSdoes not negatively influence cell viability of fish cells suchas the RTgill-W1 cells (Michel et al., 2014) and resulted ineven better differentiation of rainbow trout cells and sus-tained growth in vitro for over 15 days (Mothersill et al.,1995), this exposure system is appropriate for cellviability testing. Thus, from the MTT values for our celllines, which indicate influences on mitochondrial func-tionality, it can be concluded that fish cell lines are moresensitive or at least similar in sensitivity to ZEN comparedto mammalian or insect cell lines. The EC50 values obtainedfor the NR assay (ranging from 2,837e8,973 ng ml�1 ZEN)showed lower values than for the MTT assays. These valuesare considerably lower than the IC50 of 34,625 ng ml�1 ZEN

reported for CHOeK1 cells treated for 24 h with ZEN in thepresence of 10% FBS (Ferrer et al., 2009). Decreased uptakeof NR due to ZEN exposure has also been shown in Caco-2 cells in a concentration-dependent manner revealing anIC50 of 4,775 ng ml�1 (Kouadio et al., 2005). Thus, it can beassumed that despite the fact that serum absence influ-enced the sensitivity of the fish cells, these are very sensi-tive to ZEN.

Although ZEN does not seem to specifically target theprotein synthesis machinery such as has been reported forthe mycotoxin deoxynivalenol (Kouadio et al., 2005),apoptotic effects of ZEN involving amitochondrial signalingpathway have been confirmed in human HepG2 cells(Ayed-Boussema et al., 2008). A loss of membrane integrity

0

1'000

2'000

3'000

4'000ng

MD

Am

gpr

otei

n-1

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

A

***

****

** ** ***

**

ZEN (ng/ml)

0

1'000

2'000

3'000

4'000

ngM

DA

mg

prot

ein-1

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

* ** *

**

B

ZEN (ng/ml)

0

1'000

2'000

3'000

4'000

ngM

DA

mg

prot

ein-1

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

C

** ****

**

*

ZEN (ng/ml)

0

1'000

2'000

3'000

4'000

ngM

DA

mg

prot

ein-1

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

** ** ** ** *

D

ZEN (ng/ml)

0

1'000

2'000

3'000

4'000

ngM

DA

mg

prot

ein-1

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

E

****

**

**

ZEN (ng/ml)

Fig. 8. Lipid peroxidation expressed as ng malondialdehyde (MDA) per mg protein in different cells lines (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8,D ¼ SHK-1, E ¼ CCB) after exposure to ZEN for 24 h, mean ± SE, difference to solvent controls: * ¼ P < 0.05, ** ¼ P < 0.01, *** ¼ P < 0.001.

0

30

60

90

120

tail

mom

ent

EtOH 313 625 1250 2500 5000 7500 10000 15000 20000

***

*** **

*

***

***

***

*** **

*

***

ZEN (ng/ml)

Fig. 9. Genotoxic effects of ZEN measured as tail moments in the alkalinecomet assay in RTL-W1 cells after exposure to different concentrations ofZEN for 24 h, mean ± SE, difference to solvent controls: *** ¼ P < 0.001.

C. Pietsch et al. / Toxicon 88 (2014) 44e6156

indicating apoptotic events in cells was also observed in allcell lines by means of the propidium iodide (PI) assays. Theobservations of destabilization of lysosomes determined bythe NR assay together with damages to membranes andmitochondria revealed by the PI assay and the MTT teststrongly indicate that apoptosis takes place in these cells.The PI assay showing rather similar EC50 values for all celllines was more sensitive than the MTT test. Thus, it can beconcluded that at exposure concentrations higher thanapproximately 11,000 ng ml�1 ZEN membrane integrity ofall cell types is impaired. Induction of apoptosis has beenshown in several mammalian cell lines by treatment withZEN (Ouanes et al., 2003; Abid-Essefi et al., 2004). Since theEC50 values of the NR assay are much lower than for theother assays it may be assumed that ZEN exerts specificeffects on lysosomes and/or endosomes before more gen-eral toxic effects on cell membranes occur. A central role of

Fig. 10. Reduced (GSH) and oxidized glutathione (GSSG) in different cells lines (A ¼ RTL-W1, B ¼ RTgill-W1, C ¼ RT EQ clone 8, D ¼ SHK-1, E ¼ CCB) after exposureto ZEN for 24 h, mean ± SD, difference to solvent controls: * ¼ P < 0.05.

C. Pietsch et al. / Toxicon 88 (2014) 44e61 57

a lysosomal pathway in the cytotoxicity of ZEN was alsofound in a human embryonic kidney cell line (Gao et al.,2013). However, the involvement of specific reactionscausing toxicity of ZEN should be the subject of furtherresearch. At least a major role of estrogenic modes of actionin toxicity of ZEN which has occasionally been assumed(Wasowicz et al., 2005) can be excluded in the fish cell linessince cell lines such as the CCB cells without identifiableexpression of nuclear estrogen receptors still show highsensitivity to this mycotoxin. Similar to our results,mammalian cell lines show similar cytotoxicity to ZENalthough they differentially express estrogen receptors(Abid-Essefi et al., 2004). This suggests that an estrogenreceptor-mediated action in cells is rather unlikely to be theprime target of ZEN in these cell lines. However, since es-trogens are also capable of initiating second messengersignaling events, including mobilization of intracellularcalcium (Morley et al., 1992), release of cyclic AMP (Razandi

et al., 1999), and activation of themitogen-activated proteinkinases (Watters et al., 1997; Singh et al., 1999), furtherresearch is needed to investigate these possible mecha-nisms in fish cells.

Furthermore, it is possible that ZEN toxicity involvesoxidative stress as a potential mode of action. Cytotoxiceffects in combination with ROS produced due to ZENtreatment of mammalian cells (Abid-Essefi et al., 2004;Hassen et al., 2007) have also been observed in our fishcell lines. Loss of mitochondrial membrane potential due torapid ROS production and damage at the mitochondrialpathway by ZEN has also been observed in CHOeK1 cells(Ayed-Boussema et al., 2008; Ferrer et al., 2009). Since ROSproductionmay influencemembrane function fluorescencedyes such as H2DCF-DA may yield more reliable resultsthan absorbance measurements using NBT. However, cellsunder way in apoptosis may still maintain their metabolicactivity to a certain extent and are counted as living cells in

C. Pietsch et al. / Toxicon 88 (2014) 44e6158

theMTTmethod (Bouaziz et al., 2008). Therefore, also otherend points of cell function have been included in our ex-periments. Overproduction of ROS is known to causedepletion of cellular antioxidants and oxidative injury incells including DNA damage, oxidation of proteins and lipidperoxidation (Mates, 2000; Hassen et al., 2007). Lipid-peroxidation due to ZEN treatment in vitro and in vivo hasbeen reported (Kouadio et al., 2005; Hassen et al., 2007;Abbes et al., 2007; Ferrer et al., 2009). However, theoccurence of lipid peroxidations in the fish cell lines wasoften not as pronounced as the effects observed in otherassays. Therefore, it can be concluded that oxidative stressin these cells may not progress through membranedamages.

Moreover, damage to DNA by ROS can lead to genotox-icity (Epe, 1991). ZEN-induced DNA damages were verypronounced in RTL-W1 cells. This confirms the genotoxicpotential of ZEN which has already been shown inmammalian cell lines (Ouanes et al., 2003; Abid-Essefiet al., 2004; Lioi et al., 2004; Venkataramana et al., 2014).Moreover, increased DNA strand breaks in RTL-W1 cellswere found at even lower concentrations than reportedpreviously in a human cell line (Gao et al., 2013) and thefish cell line proved to be more sensitive to this mycotoxinthan Vero or Caco-2 cells by showing a lower EC50 value inthe alkaline comet assay than for the determination of DNAsynthesis inhibition in the mammalian cells (2009 ng ml�1

ZEN versus 2865 and 4139 ng ml�1 ZEN, respectively)(Abid-Essefi et al., 2004).

Furthermore the exposure to ZEN generated a consid-erable amount of reactive oxygen species in mammaliancells leading to apoptosis which could be counteracted bythe supplementation with antioxidants such as vitamin E(El Golli Bennour et al., 2008) and N-acetyl cysteine(Venkataramana et al., 2014). In our study, we showed thatglutathione as an antioxidant is influenced by ZEN expo-sure, although less reduced GSH levels not necessary led toincreased GSSG levels. This is probably due to the fact thatwhole-cell extracts do not allow investigating the exactregulation of GSSG levels in different cellular compart-ments (Morgan et al., 2013). However, increased ROS pro-duction was accompanied by decreased GSH levels at1250 ng ml�1 in RTL-W1 cells and at 5,000 ng ml�1 inRTgill-W1 cells. No coherence between ROS production andGSH levels was found in RT EQ clone 8 cells and CCB cells,and reduced ROS production showed also reduced GSHlevels in SHK-1 cells at the same time. Thus, cell-specificdifferences in ROS metabolization and GSH homeostasismust be assumed. These findings confirm that exposure toZEN leads to oxidative stress in cells. The increased GSH toGSSG ratio at concentrations above 2,500 ng ml�1 ZEN thatoccurred in some of the cell lines was probably due torebound effects that already have been observed in livertissue after treatment with several compounds which havebeen reported to deplete GSH (Plummer et al., 1981). Thiseffect commonly comes into action 12e48 h after initialstress probably due to activation of GSH-providing en-zymes (Deneke and Fanburg, 1989).

The antioxidant N-acetyl cysteine counteracted theZEN-mediated ROS generation and effects on mitochon-drial membrane potential in human neuroblastoma cells

(Venkataramana et al., 2014). However, the use of theantioxidant hydroxytyrosol revealed that ROS productionin human embryonic kidney cells due to ZEN exposurecould be reduced whereas the genotoxic potential was notinfluenced (Gao et al., 2013). This proved that there is nodirect causal connection between oxidative stress and DNAstrand breaks in these cells which might also apply to thepresent study on fish cell lines.

Moderate concentrations of ZEN induced an increase ofabsorption values in the NR and theMTT tests, andwewereable to show that ZEN has concentration-dependentbiphasic effects on cell viability. Biphasic effects in biolog-ical systems caused by xenobiotics are thought to beadaptive responses by which an initial disruption of ho-meostasis induces compensatory biological processes(Calabrese and Baldwin, 2002). Accordingly, biphasic ef-fects have been described for cell lines treated with themycotoxin deoxynivalenol or plant-derived toxins such asisoflavones (de Lemos, 2001; Chen and Donovan, 2004;Guo et al., 2004; Diesing et al., 2011; Pietsch et al., 2011)but not for exposure to ZEN. Increased values in the MTT orNR tests at low ZEN concentrations are certainly not due tocell proliferations since the absence of FBS in the exposuremedia prevents cell proliferation as has been shown for theRTgill-W1 cells (Michel et al., 2014). Thus, the biphasic ef-fects that have been observed in the cytotoxicity assays andthe ROS measurements in the present study may haveother reasons. On the one hand it has been proposed thatcarcinogens may induce repair mechanisms that not onlyreverse damages provoked by the exogenous chemical butalso background damages (Conolly and Lutz, 2004). Thus, atlow concentrations of a toxicant the induction of repairmechanisms may over-compensate the incremental le-sions. These findings have been demonstrated for DNAdamages (Conolly and Lutz, 2004). Thus, this hypothesismay also be true for the effects of ZEN on fish cells since forthis substance the production of DNA adducts and geno-toxic effects in mammals have already been shown (Pfohl-Leszkowicz et al., 1995; Ouanes et al., 2003; Abid-Essefiet al., 2004; Lioi et al., 2004) although the carcinogenicityis still a matter of controversy (Becci et al., 1982; NTP,1982).On the other hand minor DNA damages e.g. at low toxicantconcentrations are thought to affect the cell cycle to allow asufficient DNA repair before replication (Conolly and Lutz,2004) whereas high concentrations of a cytotoxic sub-stance may accelerate cell divisions resulting in a regen-erative hyperplasia. Effects of ZEN on cell cycle progressionhave been described in two mammalian cell lines (Abid-Essefi et al., 2004). At the moment there is no evidence ifthis hypothesis may be true for the present study using fishcell lines, but further research will be employed to describepossible effects on cell cycle progression in fish cells.

In conclusion, we were able to show that (I) ZEN isprincipally metabolized by glucuronidation in the testedfish cell lines, and (II) metabolization of ZEN does notprevent cytotoxicity whereby lysosomal functions areaffected first. In addition, the results show that (III) theestrogenic potential of ZEN is not determining the cytotoxiceffects in the cell lines via estrogen receptor a and b.Further, (IV) the cytotoxic and genotoxic potential of ZEN isaccompanied by a substantial degree of intracellular

C. Pietsch et al. / Toxicon 88 (2014) 44e61 59

oxidative stress generated by this substance, but recentresearch shows that lysosomes are the main targets of ZENtoxicity and oxidative stress is considered to be a by-product of disruption of cell homeostasis (Gao et al.,2013). Thus, the present study allows new insights to themechanisms of action of ZEN although whereby all theseeffects are induced remains still not completely under-stood. Especially, the impairment of lysosomes and furtherprogress of cell toxicity as well as the involvement ofmetabolization of ZEN to catechols and possible quinoneand semiquinone intermediates which probably promoteROS production in fish cells should be the subject of furtherresearch.

Acknowledgments

The authors like to thank Prof. Dr. Kristin Schirmer(EAWAG, Dübendorf, Switzerland), Dr. Dannevig (NationalVeterinary Institute, Oslo, Norway), Dr. Edwige Quillet(Jouy-en-Josas Cedex, France), and the Fish DiseaseResearch Unit at the School of Veterinary Medicine(Hannover, Germany) for providing the cell lines.

Conflict of interest

The authors declare that there are no conflicts ofinterest.

Transparency document

Transparency document related to this article can befound online at http://dx.doi.org/10.1016/j.toxicon.2014.06.005.

Appendix A. Supplementary data

Supplementary data related to this article can be foundat http://dx.doi.org/10.1016/j.toxicon.2014.06.005.

References

Abbes, S., Zouhour, O., Salah-Abbes, J.B., Abdel-Wahhab, M.A., Oueslati, R.,Bacha, H., 2007. Preventive role of aluminosilicate clay against in-duction of micronuclei and chromosome aberrations in bone-marrowcells of Balb/c mice treated with zearalenone. Mutat. Res. 631, 85e92.

Abdelhamid, A.M., 1990. Occurrence of some mycotoxins (aflatoxins,ochratoxin A, citrinin, zearalenone and vomitoxin) in various Egyp-tian feeds. Arch. Tierernahr. 40 (7), 647e664.

Abid-Essefi, S., Ouanes, Z., Hassen, W., Baudrimont, I., Creppy, E.E.,Bacha, H., 2004. Cytotoxicity, inhibition of DNA and protein synthesesand oxidative damage in cultured cells exposed to zearalenone.Toxicol. In Vitro 18, 467e474.

Abid-Essefi, S., Bouaziz, C., El Golli-Bennour, E., Ouanes, Z., Bacha, H.,2009. Comparative study of toxic effects of zearalenone and its twomajor metabolites alpha-zearalenol and beta-zearalenol on culturedhuman Caco-2 cells. J. Biochem. Mol. Toxicol. 23, 233e243.

Ayed-Boussema, I., Bouaziz, C., Rjiba, K., Valenti, K., Laporte, F., Bacha, H.,Hassen, W., 2008. The mycotoxin zearalenone induces apoptosis inhuman hepatocytes (HepG2) via p53-dependent mitochondrialsignaling pathway. Toxicol. In Vitro 22, 1671e1680.

Babich, H., Borenfreund, E., 1991. Cytotoxicity and genotoxicity assayswith cultured fish cells: a review. Toxicol. In Vitro 5, 91e100.

Babich, H., Shopsis, C., Borenfreund, E., 1986. In vitro cytotoxicity testing ofaquatic pollutants (cadmium, copper, zinc, nickel) using establishedfish cell lines. Ecotoxicol. Environ. Saf. 11, 91e99.

Bakos, K., Kov�acs, R., Staszny, �A., Sipos, D.K., Urb�anyi, B., Müller, F., Csenki, Z.,Kov�acs, B., 2013. Developmental toxicity and estrogenic potencyof zearalenone in zebrafish (Danio rerio). Aquat. Toxicol. 136e137,13e21.

Becci, P.J., Voss, K.A., Hess, F.G., Gallo, M.A., Parent, R.A., Stevens, K.R.,Taylor, J.M., 1982. Long-term carcinogenicity and toxicity study ofzearalenone in the rat. J. Appl. Toxicol. 2, 247e254.

Berek, L., Petri, I.B., Mesterhazy, A., Teren, J., Molnar, J., 2001. Effects ofmycotoxins on human immune functions in vitro. Toxicol. In Vitro 15,25e30.

B€ogi, C., Levy, G., Lutz, I., Kloas, W., 2002. Functional genomics and sexualdifferentiation in amphibians. Comp. Biochem. Physiol. B 133,559e570.

Bouaziz, C., el Dein, O.S., El Golli, E., Abid-Essefi, S., Brenner, C., Lemaire, C.,Bacha, H., 2008. Different apoptotic pathways induced by zear-alenone, T-2 toxin and ochratoxin A in human hepatoma cells. Toxi-cology 254, 19e28.

Bryden, W.L., Lloyd, A.B., Cumming, R.B., 1980. Aflatoxin contamination ofAustralian animal feeds and suspected cases of mycotoxicosis. Aust.Vet. J. 56 (4), 176e180.

Bucheli, T.D., Erbs, M., Hartmann, N., Vogelsang, S., Wettstein, F.E.,Forrer, H.R., 2005. Estrogenic mycotoxins in the environment. Mitt.Leb. Hyg. 96, 386e403.

Bucheli, T.D., Wettstein, F.E., Hartmann, N., Erbs, M., Vogelsang, S.,Forrer, H.-R., Schwarzenbach, R.P., 2008. Fusarium mycotoxins:overlooked aquatic micropollutants. J. Agric. Food Chem. 56,1029e1034.

Calabrese, E.J., Baldwin, L.A., 2002. Defining hormesis. Hum. Exp. Toxicol.21, 91e97.

Cetin, Y., Bullerman, L.B., 2005. Cytotoxicity of Fusarium mycotoxins tomammalian cell cultures as determined by the MTT bioassay. FoodChem. Toxicol. 43 (5), 755e764.

Charmley, L.L., Rosenberg, A., Trenholm, H.L., 1994. Factors responsible foreconomic losses due to Fusarium mycotoxin contamination of grains,foods and feedstuffs. In: Miller, J.D., Trenholm, H.L. (Eds.), Mycotoxinsin Grain-Compounds Other than Aflatoxin. Eagan Press, Minnesota,pp. 471e486.

Chen, A.-C., Donovan, S.M., 2004. Genistein at a concentration present insoy infant formula inhibits Caco-2BBe cell proliferation by causingG2/M cell cycle arrest. J. Nutr. 134, 1303e1308.

Conkov�a, E., Laciakov�a, A., P�astorov�a, B., Seidel, H., Kov�ac, G., 2001. Theeffect of zearalenone on some enzymatic parameters in rabbits.Toxicol. Lett. 121 (3), 145e149.

Conolly, R.B., Lutz, W.K., 2004. Nonmonotonic dose-response relation-ships. Mechnistic basis, kinetic modeling, and implications for riskassessment. Toxicol. Sci. 77, 151e157.

Coppock, R.W., Mostrom, M.S., Sparling, C.G., Jacobsen, B., Ross, S.C., 1990.Apparent zearalenone intoxication in a dairy herd from feedingspoiled acid treated corn. Vet. Hum. Toxicol. 32, 246e248.

de Lemos, M., 2001. Effects of soy phytoestrogens genistein and daidzeinon breast cancer growth. Ann. Pharmacother. 35, 1118e1121.

Deneke, S.M., Fanburg, B.L., 1989. Regulation of cellular glutathione. Am. J.Physiol. 257, 163e173.

Diesing, A.-K., Nossol, C., Panther, P., Walk, N., Post, A., Kluess, J.,Kreutzmann, P., D€anicke, S., Rothk€otter, H.-J., Kahlert, S., 2011.Mycotoxin deoxynivalenol (DON) mediates biphasic cellular responsein intestinal porcine epithelial cell lines IPEC-1 and IPEC-J2. Toxicol.Lett. 200, 8e18.

D€oll, S., Valenta, H., Baardsen, G., M€oller, P., Koppe, W., Stubhaug, I.,D€anicke, S., 2011. Effects of Increasing Concentrations of Deoxy-nivalenol, Zearalenone and Ochratoxin A in Diets for Atlantic Salmon(Salmo salar) on Performance, Health and Toxin Residues. Abstracts ofthe 33rdMycotoxinWorkshop. Freising, Germany, 30thMaye 1st June.

Driehuis, F., Spanjer, M.C., Scholten, J.M., te Giffel, M.C., 2008. Occurrenceof mycotoxins in feedstuffs of dairy cows and estimation of total di-etary intakes. J. Dairy Sci. 91 (11), 4261e4271.

El Golli Bennour, E., Bouaziz, C., Ladjimi, M., Renaud, F., Bacha, H., 2009.Comparative mechanisms of zearalenone and ochratoxin A toxicitieson cultured HepG2 cells: Is oxidative stress a common process? En-viron. Toxicol. 24, 538e548.

Epe, B., 1991. Genotoxicity of singlet oxygen. Chem. Biol. Interact. 80 (3),239e260.

Ferrer, E., Juan-Garcia, A., Font, G., Ruiz, M.J., 2009. Reactive oxygen spe-cies induced by beauvericin, patulin and zearalenone in CHO-K1 cells.Toxicol. In Vitro 23, 1504e1509.

Fleck, S., Hildebrand, A., Müller, E., Pfeiffer, E., Metzler, M., 2012. Geno-toxicity and inactivation of catechol metabolites of the mycotoxinzearalenone. Mycotox. Res. 28, 267e273.

C. Pietsch et al. / Toxicon 88 (2014) 44e6160

Fornelli, F., Minervcini, F., Logrieco, A., 2004. Cytotoxicity of fungal me-tabolites to lepidopteran (Spodoptera frugiperda) cell line (SF-9). J.Invertebr. Pathol. 85, 74e79.

Frandsen, H., 2007. Deconjugation of N-glucuronide conjugated metab-olites with hydrazine hydrate e biomarkers for exposure to the food-borne carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyri-dine (PhIP). Food Chem. Toxicol. 45, 863e870.

Freitas, L.G., G€otz, C.W., Ruff, M., Singer, H.P., Müller, S.R., 2004. Quanti-fication of the new triketone herbicides, sulcotrione and mesotrione,and other important herbicides and metabolites, at the ng/l level insurface waters using liquid chromatography-tandem mass spec-trometry. J. Chromatogr. A 1028, 277e286.

Gao, F., Jiang, L.-P., Chen, M., Geng, C.-Y., Yang, G., Ji, F., Zhong, L.-F., Liu, X.-F., 2013. Genotoxic effects induced by zearalenone in a human em-bryonic kidney cell line. Mutat. Res./Genetic Toxicol. Environ.Mutagen. 755, 6e10.

George, S.G., Taylor, B., 2002. Molecular evidence for multiple UDP-glucuronosyltransferase gene familes in fish. Mar. Environ. Res. 54,253e257.

Grancharov, K., Naydenova, Z., Lozeva, S., Golovinsky, E., 2001. Natural andsynthetic inhibitors of UDP-glucuronosyltransferase. Pharmacol. Ther.89, 171e186.

Gromadzka, K., Waskiewicz, A., Golinski, P., Swietlik, J., 2009. Occurrenceof estrogenic mycotoxin e zearalenone in aqueous environmentalsamples with various NOM content. Water Res. 43, 1051e1059.

Guo, J.M., Xiao, B.X., Liu, D.H., Grant, M., Zhang, S., Lai, Y.F., Guo, Y.B.,Liu, Q., 2004. Biphasic effect of daidzein on cell growth of humancolon cancer cells. Food Chem. Toxicol. 42, 1641e1646.

Hartmann, N., Erbs, M., Wettstein, F.E., Schwarzenbach, R.P., Bucheli, T.D.,2007. Quantification of estrogenic mycotoxins at the ng/L level inaqueous environmental samples using deuterated internal standards.J. Chromatogr. A 1138, 132e140.

Hartmann, N., Erbs, M., Forrer, H.-R., Vogelsang, S., Wettstein, F.E.,Schwarzenbach, R.P., Bucheli, T.D., 2008. Occurrence of zearalenoneon Fusarium graminearum infected wheat and maize field in croporgans, soil, and drainage water. Environ. Sci. Technol. 42,5455e5460.

Hassen, W., Ayed-Boussema, I., Azqueta, A., De Cerain Lopez, O.A.,Bacha, W., 2007. The role of oxidative stress in zearalenone-mediatedtoxicity in HepG2 cells: oxidative DNA damage, glutathione depletionand stress proteins induction. Toxicology 232, 294e302.

Hoerger, C., Schenzel, J., Strobel, B., Bucheli, T., 2009. Analysis of selectedphytotoxins and mycotoxins in environmental samples. Anal. Bioanal.Chem. 395, 1261e1289.

Holt, S., Gunderson, M., Joyce, K., Nayini, N.R., Eyster, G.F., Garitano, A.M.,Zonia, C., Krause, G.S., Aust, S.D., White, B.C., 1986. Myocardial tissueiron delocalization and evidence for lipid peroxidation after twohours of ischemia. Ann. Emerg. Med. 15, 1155e1159.

Hooft, J.M., Elmor, A.E.H.I., Encarnaç~ao, P., Bureau, D.P., 2011. Rainbow trout(Oncorhynchus mykiss) is extremely sensitive to the feed-borne Fusa-rium mycotoxin deoxynivalenol (DON). Aquaculture 311, 224e232.

Johns, S.M., Denslow, N.D., Kane, M.D., Watanabe, K.H., Orlando, E.F.,Sepulveda, M.S., 2009. Effects of estrogens and antiestrogens on geneexpression of fathead minnow (Pimephales promelas) early life stages.Environ. Toxicol. 26 (2), 195e206.

Juszkiewicz, T., Piskorska-Pliszczynska, J., 1992. Occurrence of mycotoxinsin animal feeds. J. Environ. Pathol. Toxicol. Oncol. 11 (4), 211e215.

Kiessling, K.H., Pettersson, H., 1978. Metabolism of zearalenone in ratliver. Acta Pharmacol. Toxicol. 43, 285e292.

Kolpin, D.W., Hoerger, C.C., Meyer, M.T., Wettstein, F.E., Hubbard, L.E.,Bucheli, T.D., 2010. Phytoestrogens and mycotoxins in Iowa streams:an examination of underinvestigated compounds in agricultural ba-sins. J. Environ. Qual. 39, 2089e2099.

Kouadio, J.H., Mobio, T.A., Baudrimont, I., Moukha, S., Dano, S.D.,Creppy, E.E., 2005. Comparative study of cytotoxicity and oxidativestress induced by deoxynivalenol, zearalenone or fumonisin B1 inhuman intestinal cell line Caco-2. Toxicology 213, 56e65.

Lagan�a, A., Fago, G., Marino, A., Santarelli, D., 2001. Development of ananalytical system for the simultaneous determination of anabolicmacrocyclic lactones in aquatic environmental samples. Rapid Com-mun. Mass Spectrom. 15, 304e310.

Lagan�a, A., Faberi, A., Fago, G., Marino, A., Pastorini, E., Samperi, R., 2004.Application of an innovative matrix solid-phase dispersionesolid-phase extractioneliquid chromatographyetandem mass spectrom-etry analytical methodology to the study of the metabolism of theestrogenic mycotoxin zearalenone in rainbow trout liver andmuscular tissue. Int. J. Environ. Anal. Chem. 84 (13), 1009e1016.

Linford, N.J., Yang, Y., Cook, D.G., Dorsa, D.M., 2001. Neuronal apoptosisresulting from high doses of isoflavone genistein: role for calcium and

P42/44 mitogen-activated protein kinase. J. Pharmacol. Exp. Ther.299, 1e9.

Lioi, M.B., Santoro, A., Barbieri, R., Salzano, S., Ursini, M.V., 2004.Ochratoxin and zearalenone: a comparative study on genotoxiceffects and cell death induced in bovine lymphocytes. Mutat. Res.557, 19e24.

Maaroufi, K., Chekir, L., Creppy, E.E., Ellouz, F., Bacha, H., 1996. Zearalenoneinduces modifications of haematological and biochemical parametersin rats. Toxicon 34, 535e540.

MacGregor, J.I., Jordan, V.C., 1998. Basic guide to the mechanisms ofantiestrogen action. Pharmacol. Rev. 50, 151e196.

Måge, A., Julshamn, K., Lunestad, B.T., 2009. Overvåkningsprogram forforvarer til fisk og andre akvatiske dyr e Årsrapport 2008 og 2009.NIFES, Bergen (in Norwegian only).

Mankevi�cien _e, A., Butkute, B., Dabkevi�cius, Z., Suproniene, S., 2007.Fusarium mycotoxins in lithuanian cereals from the 2004e05 har-vests. Ann. Agric. Environ. Med. 14, 103e107.

Marin, L., Murtha, J., Dong, W., Pestka, J.J., 1996. Effects of mycotoxins oncytokine production and proliferation in EL-4 thymoma cells. J. Tox-icol. Environ. Health 48, 379e396.

Mates, J.M., 2000. Effects of antioxidant enzymes in the molecular controlof reactive oxygen species toxicology. Toxicology 153, 83e104.

Matthews, J., Celius, T., Halgren, R., Zacharewski, T., 2000. Differentialestrogen receptor binding of estrogenic substances: a species com-parison. J. Steroid Biochem. Mol. Biol. 74, 223e234.

McKelvey-Martin, V.J., Green, M.H.L., Schmezer, P., Pool-Zobel, B.L., DeM�eo, M.P., Collins, A., 1993. The single cell gel electrophoresis assay(comet assay): a European review. Mutat. Res. 288, 47e63.

Meech, R., Mackenzie, P.I., 1997. Structure and function of uridinediphosphate glucuronosyltransferases. Clin. Exp. Pharmacol. Physiol.24, 907e915.

Michel, C., Herzog, S., de Capitani, C., Burkhardt-Holm, P., Pietsch, C., 2014.Natural mineral particles are cytotoxic to rainbow trout gill epithelialcells in vitro. PLOS ONE. http://dx.doi.org/10.1371/journal.-pone.0100856 (submitted for publication).

Miles, C.O., Erasmuson, A.F., Wilkins, A.L., Towers, N.R., Smith, B.L.,Garthwaite, I., Scahill, B.G., Hansen, R.P., 1996. Ovine metabolism ofzearalenone to a-zearalanol (zeranol). J. Agric. Food Chem. 44,3244e3250.

Morgan, B., Ezerina, D., Amoako, T.N.E., Riemer, J., Seedorf, M., Dick, T.P.,2013. Multiple glutathione disulfide removal pathways mediatecytosolic redox homeostasis. Nat. Chem. Biol. 9, 119e125.

Morley, P., Whitfield, J.F., Vanderhyden, B.C., Tsang, B.K., Schwartz, J.L.,1992. A new, nongenomic estrogen action: the rapid release ofintracellular calcium. Endocrinology 131, 1305e1312.

Mosmann, T., 1983. Rapid colorimetric assay for cellular growth andsurvival: application to proliferation and cytotoxicity assays. J.Immunol. Methods 65 (1e2), 55e63.

Mothersill, C., Lyng, F., Lyons, M., Cottell, D., 1995. Growth and differen-tiation of epidermal cells from the rainbow trout established as ex-plants and maintained in various media. J. Fish. Biol. 46 (6),1011e1025.

NTP, National Toxicology Program USA, 1982. Technical Report on theCarcinogenesis Bioassay of Zearalenone (CAS No. 17924-92-4) in F344/N Rats and B6C3F1 Mice (Feed Study). NIH Publ. No. 83, 1791,Research Triangle Park, NC.

Obremski, K., Zielonka, L., Zaluska, G., Zwierzchowski, W., Pirus, K.,Gajecki, M., 1999. The influence of low doses of zearalenone on liverenzyme activities in gilts. In: Proceedings of the X. Conference.Microscopic Fungi e Plant Pathogens and their Metabolites. Pznan,p. 66.

Olsen, M., Pettersson, H., Kiessling, K.H., 1981. Reduction of zearalenone tozearalenol in female rat liver by 3 alpha-hydroxysteroid dehydroge-nase. Acta Pharmacol. Toxicol. 48, 157e161.

Ouanes, Z., Abid, S., Ayed, I., Anane, R., Mobio, T., Creppy, E., Bacha, H.,2003. Induction of micronuclei by zearalenone in Vero monkey kid-ney cells and in bone marrow cells of mice: protective effect of vit-amine E. Mutat. Res. 538, 63e70.

Pesonen, M., Andersson, T., 1987. Subcellular localization and propertiesof cytochrome P-450 and UDP glucuronosyltransferase in the rainbowtrout kidney. Biochem. Pharmacol. 36, 823e829.

Pfeiffer, E., Hildebrand, A., Mikula, H., Metzler, M., 2010. Glucuronidationof zearalenone, zeranol and four metabolites in vitro: formation ofglucuronides by various microsomes and human UDP-glucuronosyltransferase isoforms. Mol. Nutr. Food Res. 54,1468e1476.

Pfohl-Leszkowicz, A., Chekir-Ghedira, L., Bacha, H., 1995. Genotoxicity ofzearalenone, an estrogenic mycotoxin: DNA adduct formation in fe-male mouse tissues. Carcinogenesis 16 (10), 2315e2320.

C. Pietsch et al. / Toxicon 88 (2014) 44e61 61

Pietsch, C., Bucheli, T., Wettstein, F., Burkhardt-Holm, P., 2011. Frequentbiphasic cellular responses of permanent fish cell cultures to deoxy-nivalenol (DON). Toxicol. Appl. Pharmacol. 256, 24e34.

Pietsch, C., Kersten, S., Burkhardt-Holm, P., Valenta, H., D€anicke, S., 2013.Occurrence of deoxynivalenol and zearalenone in commercial fishfeed. Toxins 5, 184e192.

Placinta, C.M., D`Mello, J.P.F., Macdonald, A.M.C., 1999. A review ofworldwide contamination of cereal grains and animal feed withFusarium mycotoxins. Anim. Feed Sci. Technol. 78, 21e37.

Plummer, J.L., Smith, B.R., Sies, H., Bend, J.R., 1981. Chemical depletion ofglutathione in vivo. Methods Enzymol. 77, 50e59.

Poston, H.A., Coffin, J.L., Combs, G.F., 1982. Biological effects of dietary T-2toxin on rainbow trout, Salmo gairdneri. Aquat. Toxicol. 2, 79e88.

Ranjan, K.S., Sinha, A.K., 1991. Occurrence of mycotoxigenic fungi andmycotoxins in animal feed from bihar, India. J. Sci. Food Agric. 56 (1),39e47.

Rau, M.A., Whitaker, J., Freedman, J.H., Di Giulio, R.T., 2004. Differentialsusceptibility of fish and rat liver cells to oxidative stress and cyto-toxicity upon exposure to prooxidants. Comp. Biochem. Physiol. PartC Toxicol. Pharmacol. 137, 335e342.

Razandi, M., Pedram, A., Greene, G.L., Levin, E.R., 1999. Cell membrane andnuclear estrogen receptors (ERs) originate from a single transcript:studies of ERa and ERb expressed in Chinese hamster ovary cells. Mol.Endocrinol. 13, 307e319.

Roberts, B.A., Patterson, D.S., 1975. Detection of twelve mycotoxins inmixed animal feedstuffs, using a novel membrane cleanup procedure.J. Assoc. Off. Anal. Chem. 58 (6), 1178e1181.

Sanden, M., Jorgensen, S., Hemre, G.-I., Ornrud, R., Sissener, N.H., 2012.Zebrafish (Danio rerio) as a model for investigating dietary toxic ef-fects of deoxynivalenol contamination in aquaculture feeds. FoodChem. Toxicol. 50 (12), 4441e4448.

Santacroce, M., Conversano, M., Casalino, E., Lai, O., Zizzadoro, C.,Centoducati, G., Crescenzo, G., 2008. Aflatoxins in aquatic species:metabolism, toxicity and perspectives. Rev. Fish Biol. Fish. 18, 99e130.

Santos, G.A., Rodrigues, I., Naehrer, K., Encarnacao, P., 2010. Mycotoxins inaquaculture: occurrence in feed components and impact on animalperformance. Aquac. Eur. 35 (4), 6e10.

Schaut, A., De Saeger, S., Sergent, T., Schneider, Y.-J., Larondelle, Y.,Pussemier, L., Van Peteghem, C., 2008. Study of the gastrointestinalbiotransformation of zearalenone in a Caco-2 cell culture systemwithliquid chromatographic methods. J. Appl. Toxicol. 28, 966e973.

Schwartz, P., Thorpe, K.L., Bucheli, T.D., Wettstein, F.E., Burkhardt-Holm, P.,2010. Short-term exposure to the environmentally relevant estro-genic mycotoxin zearalenone impairs reproduction in fish. Sci. TotalEnviron. 409, 326e333.

Segner, H., 2004. Cytotoxicity assays with fish cells as an alternative to theacute lethality test with fish. ATLA 32, 375e382.

Senft, A.P., Dalton, T.P., Shertzer, H.G., 2000. Determining glutathione andglutathione disulfide using the fluorescence probe o-phthalaldehyde.Anal. Biochem. 280, 80e86.

Singer, C.A., Rogers, K.L., Dorsa, D.M., 1998. Modulation of Bcl-2 expres-sion: a potential component of estrogen protection in NT2 neurons.Neuroreport 9, 2565e2568.

Singh, M., Setalo, G., Guan, X., Warren, M., Toran, A.C., 1999. Estrogen-induced activation of mitogen-activated protein kinase in cerebralcortical explants: convergence of estrogen and neurotrophinsignaling pathways. J. Neurosci. 19, 1179e1188.

Siraki, A.G., Chan, T.S., O'Brien, P.J., 2004. Application of quantitativestructure-toxicity relationships for the comparison of the cytotoxicityof 14 p-benzoquinone congeners in primary cultured rat hepatocytesversus PC12 cells. Toxicol. Sci. 81, 148e159.

Spengler, P., Korner, W., Metzger, J.W., 2001. Substances with estrogenicactivity in effluents of sewage treatment plants in southwesternGermany. 1. Chemical analysis. Environ. Toxicol. Chem. 20, 2133e2141.

Tanzer, L., Jones, K.J., 1997. Gonadal steroid regulation of hamster facialnerve regeneration: effects of dihydrotestosterone and estradiol. Exp.Neurol. 146, 258e264.

Veldman, A., Borggreve, G.J., Mulders, E.J., Van de Lagemaat, D., 1992.Occurence of the mycotoxins ochratoxin A, zearalenone and deoxy-nivalenol in feed components. Food Add. Contam. 9 (6), 647e655.

Venkataramana, M., Chandra Nayaka, S., Anand, T., Rajesh, R., Aiyaz, M.,Divakara, S.T., Murali, H.S., Prakash, H.S., Lakshmana Rao, P.V., 2014.Zearalenone induced toxicity in SHSY-5Ycells: the role ofoxidative stressevidenced by N-acetyl cysteine. Food Chem. Toxicol. 65, 335e342.

Videmann, B., Mazallon, M., Tep, J., Lecoeur, S., 2008. Metabolism andtransfer of the mycotoxin zearalenone in human intestinal Caco-2cells. Food Chem. Toxicol. 46, 3279e3286.

Warth, B., Sulyok, M., Berthiller, F., Schuhmacher, R., Krska, R., 2013. Newinsights into the human metabolism of the Fusarium mycotoxinsdeoxynivalenol and zearalenone. Toxicol. Lett. 220, 88e94.

Wasowicz, K., Gajecka, M., Calka, J., Jakimiuk, E., Gajecki, M., 2005. In-fluence of chronic administration of zearalenone on the processes ofapoptosis in the porcine ovary. Vet. Med. Czech 50 (12), 531e536.

Watters, J.J., Campbell, J.S., Cunningham, M.J., Krebs, E.G., Dorsa, D.M.,1997. Rapid membrane effects of steroids in neuroblastoma cells:effects of estrogen on mitogen activated protein kinase signalingcascade and c-fos immediate early gene transcription. Endocrinology138, 4030e4033.

Whitlow, L.W., Hagler Jr., W.M., 2005. Mycotoxins in feeds. Feedstuffs 77(38), 69e79.

Wo�zny, M., Brzuzan, P., Gusiatin, M., Jakimiuk, E., Dobosz, S.,Ku�zmi�nski, H., 2012. Influence of zearalenone on selected biochemicalparameters in juvenile rainbow trout (Oncorhynchus mykiss). Pol. J.Vet. Sci. 15 (2), 221e225.