www.elsevier.com/locate/chemosphere

Chemosphere 68 (2007) 662–670

Nonylphenol algal bioaccumulation and its effect throughthe trophic chain

Gabriel Correa-Reyes a, Marıa Teresa Viana a, Facundo J. Marquez-Rocha b,Alexei F. Licea b, Elizabeth Ponce b, Rafael Vazquez-Duhalt c,*

a Instituto de Investigaciones Oceanologicas, Universidad Autonoma de Baja California (UABC), P.O. Box 453, Ensenada BC 22860, Mexicob Centro de Investigaciones Cientıficas y de Educacion Superior, Km 107 carretera Tij-Eda, 22860 Ensenada BC, Mexico

c Instituto de Biotecnologıa, Universidad Nacional Autonoma de Mexico, Av. Universidad 2001, Chamilpa, Cuernavaca Morelos, Mexico

Received 19 October 2006; received in revised form 12 February 2007; accepted 15 February 2007Available online 29 March 2007

Abstract

Nonylphenol is a metabolic intermediate from the microbial transformation of detergents used worldwide. While nonylphenol showssome acute toxicity, it is also able to mimic important hormones resulting in the disruption of several processes by interfering with thesignals that control the overall physiology of the organism. The effect of the pollutant nonylphenol (NP) through the trophic chain wasstudied. Microalgae Isochrysis galbana was able to bioconcentrate NP 6940 times, where 77% of initial NP (100 lg l�1) is accumulatedintracellularly after 1-h incubation. Crustacean Artemia fransiscana showed 25% higher growth when fed with NP-rich algae. However,Artemia metabolized almost all NP ingested and only traces of NP could be found in the organism, eliminating future NP effects. Zebra-fish (Brachydanio rerio) were affected by the presence of 171 lg g�1 of NP in the diet, showing higher levels of the hormone vitellogeninand lower levels of cytochrome P450 activity. These results showed that organisms placed in the first level of trophic chain are able tosignificantly bioconcentrate the pollutant and endocrine disruptor NP. These grassed organisms affect the growth of crustacean. More-over, the organisms placed on the top of some trophic chains, such as fish, could be affected by the presence of NP in their food, in boththe hormone levels and metabolic enzymes. This work shows that the environmental presence of NP should be considered as a risk forthe organisms living in an ecosystem.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Endocrine disruptor; Cytochrome P450; Growth effect; Nonylphenol; Vitellogenin

1. Introduction

Nonylphenol (NP) is a metabolic intermediate from themicrobial transformation of used worldwide detergents.The environmental fate of alkylphenols and their ethoxy-lates has been reviewed (Ying et al., 2002, Vazquez-Duhaltet al., 2006), and significant concentrations of NP arefound in air, waters, soils and sediments (up to 81 ng m�3

in atmosphere, up to 644 lg l�1 in superficial waters, and

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

doi:10.1016/j.chemosphere.2007.02.030

* Corresponding author. Tel.: +52 777 329 1655; fax: +52 777 317 2388.E-mail address: vazqduh@ibt.unam.mx (R. Vazquez-Duhalt).

up to 13700 lg kg�1 in soils). While, NP shows someacute toxicity, it is also able to mimic important hormonesresulting in the disruption of several processes by interfer-ing with the signals that control the overall physiology ofthe organism (Vazquez-Duhalt et al., 2006).

Chronic and acute toxicities of NP on aquatic organismshave been recently reviewed by Servos (1999), Lussier et al.(1999) and Staples et al. (2004). From the abundantdata analyzed by Servos (1999) with different organismsand using different test methods, NP showed to be toxic(LC50) to fish at concentrations ranging from 17 to 3000lg l�1. Invertebrates are also sensitive to NP in a rangeof LC50 of 21–3000 lg l�1, while algae showed LC50 valuesbetween 27 and 2500 lg l�1. Evidence supporting the

G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670 663

hazardous nature of NP is quite strong, with at least 53chronic values representing fish, invertebrates and algae,for both freshwater and marine species (Vazquez-Duhaltet al., 2006). Alkylphenols are lipophilic (Nimrod and Ben-son, 1996) and thus are accumulated in a wide range ofmarine and aquatic life including algae, crustacean, mol-lusks, and fish (Eklund et al., 1990; Marcomini et al.,1990; Ahel et al., 1993; Lewis and Lech, 1996; Stapleset al., 1998; Heinis et al., 1999). This is especially importantwhen these organisms are part of the lower trophic level.NP bioaccumulation by marine amphipods and its possibleimpact to higher trophic levels have been suggested (Fayet al., 2000; Hecht et al., 2004). Algae had a larger capacityfor bioaccumulation of NP showing bioconcentration fac-tors reaching up to 10000 (Ahel et al., 1993). The estimatedNP bioconcentration factors in fish tissues ranged from 24to 1250 (Servos, 1999).

The most important effect of NP in the environmentseems to be its capacity to act as endocrine disruptor.Nonylphenol is one of the most studied estrogen mimicsthat appear to interact with development in several organ-isms (Vazquez-Duhalt et al., 2006). A large list of effects ondifferent organisms have been reported including develop-mental abnormalities, changes in the sex ratio towardsfemales, increase in the incidence of hermaphroditism,reduction of fecundity, reduction of gonadal developmentand reproductive function, decrease in egg productionand a reduction in egg viability, changes in testosteroneconversion rate, vitellogenin induction, and many others(Vazquez-Duhalt et al., 2006). In Daphnia magna, NPwas considered as responsible of the metabolic eliminationof testosterone resulting in increased level of androgenwithin the organism, accompanied a decrease reproductionof the daphnids (Baldwin et al., 1997). The percentage ofmales of 60 days post-hatch zebrafish exposed to100 mg l�1 of NP changed from 45% in control experi-ments to 10% NP exposed organisms (Hill and Janz, 2003).

With solid evidence on the effects of high levels of NPon endocrine disruption, the only unanswered question isthe risk of ambient NP exposure. The evaluation of riskto compounds causing endocrine effects may prove to beone of the greatest challenges that the risk assessmentand regulatory organizations have ever faced (Fenner-Crisp, 1997). As mentioned above, there has been only limi-ted direct use of NP, and most of the dispersed NP in theenvironment originates from the environmentally-mediatedtransformation of detergents. Increasingly, researchers willstrive to include effects on entire ecosystems, and long-term, multigenerational effects on fertility, reproductivequality, and hormonal functions. Nevertheless, no studyon the effect of NP throughout a trophic chain is availablein the literature.

This work is focused on the study of the effect of nonyl-phenol throughout the food chain, starting with the bioac-cumulation of NP by microalgae Isochrysis galbana. Thecrustacean Artemia franciscana and the zebrafish (Brac-

hydanio rerio) were used for sequential trophic levels.

2. Materials and methods

2.1. Algal strain and culture conditions

I. galbana (strain 1323) was obtained from the Prova-soli-Guillard National Center for Culture of MarinePhytoplankton (USA). The strain was maintained in250 ml Erlenmeyer flasks containing 150 ml of f/2medium (Guillard and Ryther, 1962) at 18 (±2) �C andunder 150 lE m�2 s�1 with fluorescent cold-white lamps(Sylvania F40CW). Technical grade nonylphenol (NP)was purchased from Sigma–Aldrich group (St. Louis,USA).

The acute toxicity of nonylphenol on the alga I. galbanawas determined in cell suspensions (2.5 · 106 cells per ml).Algal suspension was incubated in triplicate groups for 1 hwith different NP concentrations (from 0 to 1000 lg l�1)and then the oxygen production was determined with a cal-ibrated oxygen electrode. Cell volume was estimated fromlinear dimensions measured in an inverted microscope ata magnification of 400 · (N = 30) and using a prolatespheroid equation.

The NP bioaccumulation in I. galbana cultures wasdetermined during incubation with 100 lg l�1 of NPin f/2 medium at 18 (±2) �C and under 150 lE m�2 s�1

with fluorescent cold-white lamps. After different incu-bation times, cells were harvested by centrifugationand then the pellet washed with 10% methanol. TheNP contained in the 10% methanolic washing solutionwas considered as surface adsorbed NP. The cell pelletwas then extracted three times with dichloromethane–methanol (1:2 v/v) solution to obtain intracellularNP. Both, intracellular and surface adsorbed NP werequantified by HPLC (Agilent 1100 series) equippedwith C18 reverse phase column (150 · 3.9 mm, WatersCorporation) and a fluorescence detector. Standard curveand measurements were performed with an excitationwavelength of 220 nm and detected for an emissionwavelength of 305 nm. Results obtained here were com-pared to that obtained with a control group withoutnonylphenol.

Algal biomass as food for Artemia was produced semi-continuously in 20 l plastic bottles containing 15 l of f/2medium with the addition of sodium silicate. The seawaterwas filtered sequentially through 20, 10 and 1 lm mem-branes, then UV-irradiated and chemically disinfected withsodium hypochlorite (5.6 mg Cl l�1) and neutralized withsodium thiosulphite according Hemerick (1973). The cul-ture pH was maintained between 7.5 and 8.2 by addingCO2 into aeration system, and under 450 lE m�2 s�1 withfluorescent cold-white lamps (Sylvania F40CW). The cellcounting was performed by using 0.1 mm deep hematocy-tometer. For NP containing algal biomass as Artemia’sfood, 100 lg l�1 of NP was added to algae cultures at thestationary growth phase and incubated for 2 h. After incu-bation time only traces of NP (<0.5 lg l�1) could bedetected in solution.

664 G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670

2.2. Artemia bioassays

Nonylphenol effect on Artemia franciscana was deter-mined using algae, that had accumulated NP (after 2 hexposure to 100 lg l�1 of NP), as food. Artemia franciscana

eggs were decapsulated with a commercial treatment solu-tion containing sodium hypochlorite, and then incubated inseawater at room temperature. Nauplii larvae were col-lected during the first 6 h post-hatching to assure the samecohort. Then, the nauplii larvae were randomly distributedthrough six experimental units, two treatments in tripli-cates (with and without nonylphenol).

Each experimental unit consisted of 6 l circular plasticbuckets containing 4 l of filtered and UV irradiated seawa-ter with a batch close system (one change every second day)and provided with an air stone to assure a satisfactory oxy-gen level. Temperature was maintained at 18 ± 1 �Cthroughout the experiment by submerging the buckets ina bed with a flow-through system connected to a heatpump from an insulated 700 l reservoir. A photoperiod of12L/12D was maintained throughout the experiment. Then1000 nauplii larvae per l were randomly distributed percontainer.

The microalgae (I. galbana) was grown at constant roomtemperature (18 ± 1 �C) with a pH at range from 7.5 to 8.2,and continuous light intensity of 150 lE m�2 s�1 providedby cool white fluorescent lamps. Cell density was calculatedby direct count using a 0.1 mm hematocytometer readusing a compound Microscope Olympus model BH2. Toavoid possible cell aggregations, ultrasound (1–3 min)was applied to the sample prior cell density evaluation.Every second day during 60 days total experimentation,an aliquot with five days old microalgae was provided inexcess to the experimental units to assure ad libitum feed-ing. As mentioned above, for the NP-treated algae,100 lg l�1 of NP was added 2 h prior to feeding Artemia,whereas no treated I. galbana was added to control ex-periments.

Every week, 60 Artemia from each experimental unitwere collected, where 30 were used for biometry and theother 30 for NP accumulation analysis. Then, growthwas estimated throughout photographs taken from theinverted microscope ZEISS model Stemi 2000-C. The dig-ital pictures were analyzed by a ‘‘Motic image softwarever. 1.3’’ to calculate growth. At the end of the trial allsamples where observed to determine sex and size. TheNP content in Artemia was determined, as for algae, afterextraction with dichloromethane–methanol mixture andby HPLC equipped with a fluorescence detector.

2.3. Zebrafish (Brachydanio rerio) bioassay

Zebrafish were bought at the Scientific Hatcheries, Hun-tington Beach CA, USA. A total of 1000 zebrafish withan average initial body length 8.03 (±0.92) mm wereselected for use in the growth trial. Each experimental unitconsisted of a 20 l plastic bucket that was part of a

flow-through system (250 ml min�1). Each unit was sup-plied with an air stone to maintain satisfactory levels ofoxygen. Water temperature was maintained at 26 (±1)�Cthrough the feeding trial by a heat pump connected to aninsulated 700 l reservoir, from which water was distributedto the experimental units. Two treatments run in triplicatewere performed, where a photoperiod of 12L/12D wasmaintained through the experiment, and diets were fed inexcess three times a day during 90 days. A total of 150 fishwere randomly distributed into each experimental unit.Two experiment series with zebrafish were done. First, Art-

emia fed with NP-treated algae was used to feed zebrafish.In the second experiment, a commercial balanced diet wasenriched with NP Low NP concentration in food wasselected because fish experiment is a long term experiment(90 days) with the aim to reach fish sexual maturation andavoid lethal toxicity. The NP concentration in the diet wasverified by a GC analysis giving a concentration value of171 (±11) lg g�1 of NP. This enriched diet was used to feedzebrafish. Both experiments were performed under thesame conditions as described before and controls withoutNP were performed.

At the end of the experiments all fish was measured forweight and length. Sex ratio was also determined. Fishsamples were then either dissected to be used fresh or fro-zen at �80 �C to further analysis.

2.4. Superoxide dismutase activity (SOD)

The activity of SOD was determined by the oxidationinhibition of pyrogallol (Zhang et al., 2004) with a modifiedmethod reported by Ahmed et al. (2003). Homogenizedand centrifuged crude extract from whole organisms,Artemia or fish, were used. The final reaction mixture con-tained phosphate buffer 50 mM, pH 8.24, pyrogallol0.1 mM (stock solution 20 mM pyrogallol in 10 mMHCl) and 0.1 mg protein. The reaction mixture was incu-bated at 30 �C for 1 min and absorbance changes weremeasured at 325 nm. One unit of SOD activity was definedas 50% of oxidation inhibition with respect to control(Ahmed et al., 2003). Samples from each experimental unitin triplicate were pooled and worked individually.

2.5. Electrophoresis

SDS–PAGE electrophoresis was performed in 12.5%polyacrylamide gels, according to Laemmli (1970).Laemmli’s sample buffer was used and contained Tris–HCl 62.5 mM, pH 6.8, SDS 2.0%, 25% glycerol and0.01% blue bromophenol. After electrophoresis, the gelwas stained with Commasie blue. Zymograms were per-formed in no denaturizing conditions using a sample buffercontaining Tris–HCl 62.5 mM, pH 6.8, 40% glycerol and0.01% blue bromophenol, and no heating was provided.After run, the gel was soaked into a nitroblue tetrazolium(NBT) solution 2.5 mM with slow constant agitation for20 min. The NBT solution was replaced by a solution of

G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670 665

28 lM riboflavin and 8 mM EDTA under constant agita-tion. The gel was then placed under intense white light untilNBT was converted to formazan blue. The presence ofSOD prevent the reduction of NBT by the O2 scavenging,showing achromatic bands (Muniz et al., 2002).

2.6. Microsomal preparations

Both, Artemia treated with nonylphenol and theuntreated control were stored at �80 �C in triplicategroups until being used. Five organisms of each experimen-tal unit were pooled and homogenized in an extractionbuffer consisting in Tris–HCl 50 mM, pH 7.2, sucrose0.125 M, DTT 1 mM, EDTA 1 mM, and trypsin-inhibitortype II 1 mg ml�1. The homogenates were then centrifugedtwice; at 600g for 10 min and at 15 000g for 20 min at 4 �C.Then, supernatants were centrifuged at 105 000g for 90 min(Sorvall Ultra Pro-80) at 4 �C. The microsomal pellet wasresuspended in a buffer solution of 0.1 M potassium phos-phate pH 7.6 containing 250 mM sucrose. Microsomeswere resuspended in 1 ml buffer and stored at �80 �C.Total and microsomal protein contents were determinedusing BioRad protein assay (Bradford, 1976) and BSA asreference protein.

A total of 49 control fish (6.86 g) and 49 treatment fish(7.04 g) were used to obtain the microsomal fraction. Allfish were treated in their triplicate groups. The head, caudaland pectoral fins were eliminated. The remaining fish bodywas weighed, immediately was homogenized in extractionbuffer (0.125 M sucrose, 1 mM EDTA, 50 mM Tris–HCl(pH 7.2), 1 mM dithiothreitol and 1 mg ml�1 trypsin inhib-itor type II), followed by sequential centrifugation (30 minat 600g and 20 min at 15000g). The final centrifugation wasfor 90 min at 105000g. The microsomal pellet was resus-pended in 100 mM phosphate buffer (pH 7.6) containing250 mM sucrose and stored at �80 �C until analyses weretaken. All steps were performed at 4 �C.

2.7. Cytochrome P450 activity

The activity of cytochrome P450 in microsomal prepara-tions, from both Artemia and zebrafish, was estimated asEROD activity (7-ethoxyresorufin-O-deethylase) using amodified procedure by Burke and Mayer (1974). Blankand samples incubation mixtures consisted of Tris–HCl50 mM, pH 7.6, NADH 56 lM, 120–230 lg microsomalprotein and 0.6 lM of 7-ethoxyresorufin. The reactionwas carried out in 0.5 ml at 30 �C for 30 min. Reactionwas initiated with the addition of 7-ethoxyresorufin andstopped by addition of 500 ll of ice-cold methanol for15 min. Then the samples were centrifuged at 6000g for20 min at 4 �C. The reaction product was analyzed byHPLC chromatography using a column Zorbax 300SB-C18. Twenty microliter of the supernatant were loaded intoa HPLC (Agilent 1100) using a fluorescence detector(G1321A) at an excitation wavelength of 560 nm and emis-sion wavelength of 585 nm. The product was eluted with a

flow rate of 0.8 ml min�1, the first 4 min with solution20 mM phosphate buffer (pH 6.8)-methanol–acetonitrile(52:45:3, v/v) and then methanol was increased during16 min to a final proportion of buffer-methanol–acetonitrileof 41.6:56:2.4 v/v (Hanioka et al., 2000). Commercial res-orufin was used as standard for concentration or co-elution.

2.8. Vitellogenin analysis

Five fish from every group were washed with distillatedwater and placed in a previously sterilized homogenizercontaining 5 ml of a protein extraction buffer (40% acetoni-trile and 0.12% of trifluoroacetic acid). After homogeniza-tion, samples were centrifuged 15 min at 5000g at 4 �C.Supernatants were lyophilized and resuspended in 2 ml ofPBS. Samples were stored at �80 �C for further use.

ELISA analyses were carried out on coated plates withthe extracted protein at 5 lg ml�1 in PBS, 2 h at 37 �C.Wells were then blocked with 200 ll of 3% BSA (over nightat 4 �C). BSA was then eliminated and 50 ll of a 1:1000anti-vitellogenin antibody linked to HRP was added, 1 hat 37 �C. Wells were washed with PBS/Tween 20 and devel-oped with ABTS as substrate. Optical density was mea-sured at 405 nm. Results of vitellogenin present are thenreported as percentage of control experiments by densitom-etry between bands obtained in the same gel.

2.9. Statistical analysis

All experiments were performed in triplicate groups byrandomize experimental design. Further analyses were alsokept in triplicate by pooling organisms from the sameexperimental unit. To determine whether growth, expressedas final body weight (log transformed) and body length,were significantly different among Artemia fed with NP-treated and untreated I. galbana, a two-way analysis ofvariance, two diets and three periods of 28 day each, wasconducted. Diet and interaction were estimated usingorthogonal contrasts. All other indexes for NP bioconcen-tration, cytochrome P450 activity, and vitellogenin levelwere analyzed using a one-way analysis of variance modelat a P < 0.05 (SAS Institute Inc.).

3. Results

Acute toxicity of nonylphenol on I. galbana was testedusing the algal photosynthesis. No significant effect on oxy-gen production was detected on I. galbana at NP concen-trations up to 100 lg l�1, while at 1 mg l�1 photosynthesiswas completely inhibited as shown in Fig. 1.

Algae incubation in the presence of 100 lg l�1 of NPshowed a significant bioconcentration (Fig. 2). After 1-hincubation with NP, the algal cells accumulate 917(±40) lg g�1 dry biomass of NP and almost no free NPwas left in the aqueous medium. The intracellular accu-mulated NP represented 77.1 (±3.6)% of the total NP,while the surface adsorbed NP was 5.4 (±0.5)%. The cell

0

40

80

120

1 10 100 1000

Nonylphenol (μg/l)

Oxy

gen

prod

uctio

n (%

of c

ontro

l)

Fig. 1. Nonylphenol effect on the photosynthesis of Isochrysis galbana.Oxygen production was measured after 1 hr incubation with different NPconcentrations and reported as percentage of non treated cells. Cellsuspension (57 ml) containing 2.5 · 106 cells ml�1 were bubbled with N2

under dark until no oxygen was detected by the oxygen electrode. Then,the suspension was lighted under 1050 lE seg�1 m�2 and the oxygenproduction followed for 15 min. Average values and standard deviationwere obtained from three independent replicates (P < 0.001).

0

200

400

600

800

1000

1200

1400

0 4 8 10Time (hours)

Non

ylph

enol

(μg

g-1

dry

biom

ass)

Adsorbed

Accumulated

2 6

Fig. 2. Nonylphenol accumulation by I. galbana. One hundred lg of NPwere added to algal cultures containing 1.836 · 106 cells per ml, equivalentto 90 mg dry weight per ml. The surface adsorbed and the intracellularaccumulated nonylphenol were quantified as percentage of NP added.Average values and standard deviation were obtained from three in-dependent replicates (P < 0.001).

0

1

2

3

4

5

6

7

8

10 20 30 40 50 60

NP treated cultures

Contr

0

1

2

3

4

5

6

7

8

10 20 30 40 50 60

Time (days)

Leng

th (m

m)

Control cultures

Fig. 3. Size of Artemia fransiscana fed with Isochrysis galbana grown inthe presence (d) or in the absence (s) of 100 lg l�1 of nonylphenol.Average values and standard deviation were obtained from threeindependent replicates (P < 0.05).

666 G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670

concentration was 1 835613 (±201917) cells ml�1, equiva-lent to 49.18 (±3.55) lg of dry biomass per 106 cells. Thus,the calculated NP concentration in cells after 1 h incuba-tion was 45.1 ng per 106 cells. Considering the cellularvolume of cultivated I. galbana (65 lm3), the averageNP concentration inside cells could be estimated at693.8 lg ml�1. The bioconcentration factor was determinedas the ratio between the NP concentrations in algal cellsdivided by the initial NP concentration in water. Thus,the bioconcentration factor of I. galbana for NP underthe present conditions was 6940 times.

The crustacean Artemia is a valuable test organism forecotoxicology studies and research (Nunes et al., 2006).

Wide geographic distribution, relatively simple laboratoryculture, resistance to manipulation, short life-cycle, andan important amount of information available are amongthe characteristic that turn this organism suitable for eco-toxicology research.

The effect of NP through the trophic chain was testedfeeding Artemia franciscana with I. galbana previouslyincubated with NP. Controls were also performed usingalgae without NP. After 56 days, Artemia franciscana fedwith NP-treated I. galbana showed a 25% significant highergrowth than that observed in the control group fed withnon-treated algae (Fig. 3). This higher growth strongly sug-gests a possible endocrine alteration in the crustacean dueto the NP content in algae.

Nonylphenol accumulation by Artemia after 56 days fedwith NP-treated algae has been determined. Only traces ofnonylphenol could be found in Artemia, indicating anefficient NP metabolization. In order to know the possiblerole of cytochrome P450 enzymatic system on the NPmetabolization, the level of cytochrome P450 activity wasmeasured in both, treated and untreated Artemia organ-isms. Cytochrome P450 activity showed significant differ-ences between the treatments, the NP-treatment showeda microsomal EROD activity of 17.4 (±1.5) pmol mg�1

microsomal prot., while the control cultures showed only6.7 (±0.5) pmol mg�1 microsomal prot. On the otherhand, superoxide dismutase (SOD) activity in treatedand untreated Artemia cultures fail to show significantdifferences, (SOD activity of 0.52 (±0.29) and 0.50(±0.07) U mg�1 of protein, respectively).

Consistent to the non accumulation of NP by Artemia,freeze dried Artemia cultures from both, treated anduntreated NP-culture of I. galbana fed to zebrafish failedto show any significant differences in growth, reproductionalteration, cytochrome P450 activity, SOD activity andvitellogenin levels.

With the aim to estimate the effect of NP through thediet on zebrafish, balanced diet enriched with NP

G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670 667

(171 lg g�1) was used to feed zebrafish. No differences onsize and weight were found. The final fish length for NP-treated organisms was 21.85 (±2.44) mm, while for controlfishes was 20.27 (±2.52) mm. On the other hand, the finalweight of NP-treated fishes was 0.12 (±0.04) g and of con-trols was 0.10 (±0.04) g. However, significant differences invitellogenin levels estimated by immunological techniqueswere found between the fish fed with NP-enriched dietand controls cultures fed with commercial diet withoutNP. The vitellogenin level was 335 (±41)% higher thatthose of control experiments, which were considered as100 (±30)%.

Moreover, the cytochrome P450 activity in microsomalpreparations from NP-treated fish showed significantlylower EROD activity (69.8 ± 2.6 pmol min�1 mg�1 micro-somal prot.) than the control organisms (88.2 ± 2.7 pmolmin�1 mg�1 microsomal prot.) (p < 0.05; T-alpha =2.7936, T-value = 8.4882).

4. Discussion

The NP effect on photosynthesis inhibition in I. galbana

(Fig. 1) is in agreement to earlier reports (Graff et al., 2003)for freshwater algae, where a concentration of 500 lg l�1

NP was able to reduce algal growth (EC50). In the presentwork no significant differences on photosynthesis wereobserved at 100 lg l�1 of NP, and a complete photosynthe-sis inhibition was observed at 1 mg l�1. Measurements ofalgal NP accumulation showed that the NP uptake is arapid process, after 1-h incubation the algal cells accumu-late 917 lg g�1 dry biomass of NP. Most of the NP wasaccumulated inside the cells (77%), whereas only 5% wasadsorbed by the cell walls. Considering the cellular volumeof cultivated I. galbana, the average NP concentrationinside cells could be estimated at 694 lg ml�1, representingan effective bioconcentration factor for I. galbana of 6940under our experimental conditions. These values are closeto those reported for the macrophytic algae, Cladophora

glomerata, for which a bioconcentration factor of NP of10000 has been reported (Ahel et al., 1993). The highand fast bioconcentration of this lipophilic pollutant byalgae suggests that the main NP uptake mechanism forhigher organisms is by ingestion, and therefore algal accu-mulation seems to be the first step for NP to get throughthe food chain.

Unlike Artemia species from Europe, Africa and Asia,which show parthenogenetic reproduction, American Art-emia species show sexual reproduction (Browne, 1992;Van Stappen, 2002). This reproduction characteristicmakes Artemia franciscana suitable for endocrine disrup-tion testing. The growth increment (25%) showed by Art-

emia franciscana fed with NP-treated I. galbana during 56days (Fig. 3) was unexpected, because most reports onNP effect on growth results in a significant reduction beforedeath. In spite of those results obtained mainly with verte-brates, there are few reports that show a similar result. Asoil nematode Caenorhabditis elegans in the presence of

60–200 lg l�1 of NP (Hoss et al., 2002) resulted in a growthrate increase. As for Artemia, the NP effect on nematodegrowth seems to be a consequence of an alteration of endo-crine system of the organism. Endocrine disruption andother metabolic effects caused by the presence of NP havebeen reported. Theoretically, the endocrine disrupteraffects any biological function, and therefore can be linkedto a broad spectrum of adverse outcomes. First of all, mostmodifications are depending to a dose response and there-fore a wide variety of effects should be expected. Forinstance, in the present work no effect on reproductioncould be detected in Artemia whereas a significant increasein growth was registered at the used concentration. More-over, other type of endocrine disruptors like chlorinatedphenol compounds are related to thyroid disrupting effectwith general consequences over the metabolism (Yamauchiet al., 2003). The invertebrate physiology is somehow dif-ferent that vertebrate ones, and so far some informationis still lacking on hormonal regulation of crustaceangrowth and reproduction (Huberman, 2000). NP-uptakereached by feeding Artemia with NP-enriched algae sug-gests a hormesis, where toxicants at low concentrationsmay have a positive impact; however a negative effect athigher concentrations should not be ruled out.

No NP accumulation was found in Artemia fed withnonylphenol-treated algae, indicating an efficient NPmetabolization. In addition, the cytochrome P450 activityor EROD activity in Artemia fransciscana showed to be dif-ferent between treatments fed with NP treated anduntreated algae. The higher cytochrome P450 activity inNP-treated experiments could indicate its participation inthe NP metabolization. This could explain the lower sensi-bility of Artemia to pollutants when compared to othercrustacean species (Nunes et al., 2006). The in vitro meta-bolism of NP in rat and fish hepatic microsomes has beenexamined (Meldahl et al., 1996; Lee et al., 1998) and theNP metabolism was NADPH dependent and inhibited bythe cytochrome P450 inhibitors, even if the EROD activitywas not significantly altered by NP exposure. However, asingle exposure of NP was able to significantly increasethe liver P450 content in juvenile sea bass, Dicentrarchus

labrax (Teles et al., 2004). On the other hand and as men-tioned above, NP is weakly estrogenic to rodents and tosome species of fish. All evidence to date has indicated thatthe estrogenic effects of NP are due to the interaction of NPwith the estrogen receptor (Vazquez-Duhalt et al., 2006);nevertheless, an alternative mechanism for NP-inducedestrogenicity could be via induction by the cytochromeP450 enzymes.

In the first experiments, zebrafish was fed with Artemia

being previously fed with NP-treated I. galbana and no dif-ferences were found when compared with control experi-ments. This result is consistent with the Artemia’sefficiency to metabolize NP and thus their inability to accu-mulate NP. In order to follow the effect of NP through thetrophic chain, a second experiment series were performed,where the zebrafish was fed the NP-enriched commercial

668 G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670

balanced diet. Difference in vitellogenin levels wereobserved form this second series of experiments. Abundantinformation is available on the estrogenic effects of NP onseveral fish, mainly freshwater species, both in vivo andin vitro. One of the major target organs for estrogens isthe liver where, in all egg-laying vertebrates, the inductionof vitellogenin is driven specifically by the female sexualsteroid 17-b-estradiol. Estrogen receptors have shown alower affinity to NP than to 17-b-estradiol using the recep-tor assay for [3H]-17-b-estradiol binding in the liver cytosol(Lutz and Kloas, 1999). Nevertheless, NP is able to bindthe 17-b-estradiol receptor. The estrogenic effect of NPon zebrafish has been demonstrated after three weeks expo-sure in which an increase of vitellogenin levels and a reduc-tion of the ovarian somatic index (OSI) were detected onfemale fish (Van den Belt et al., 2004). Zebrafish exposedto NP showed a concentration-dependent suppression ofgametogenesis, in both female and male organisms (Weberet al., 2003). In addition, embryos of zebrafish after 10-days incubation in the presence of NP showed abnormali-ties such as edema, microcephalia, movement alterationsand high mortality (Willey and Krone, 2001). Thus, vitello-genin mRNA has been used for the evaluation of endocrinedisruption by NP in primary cultures of rainbow trout(Islinger et al., 1999; Pait and Nelson, 2003). Other fish spe-cies, such as Atlantic salmon have shown NP effects onvitellogenin synthesis in both juvenile and male (Yadetieet al., 1999). Kinetics of hepatic vitellogenin mRNA regu-lation, plasma vitellogenin accumulation, and vitellogeninclearance have been determined during and after exposureto NP in sheepshead minnows (Cyprinodon variegatus),mummichog (Fundulus heteroclitus), barbs (Barbus gra-

ellsi), medaka (Oryzia latipes), and in the common carp(Cyprinus carpio) (Hemmer et al., 2002; Garcıa-Reyeroet al., 2004).

On the other hand, the unexpected lower EROD activityin the NP-treated zebrafish group seems to be contradic-tory to the possible role of cytochrome P450 in the NPdetoxification metabolism (Thibaut et al., 2002). Howeversimilar results are also reported in literature, where theexpression of cytochrome P450 was lower in the Atlanticcod and turbot in the presence of NP than in the controlcultures without NP (Hasselberg et al., 2005; Martin-Skil-ton et al., 2006). The detoxification pathway in the blackgoby (Gobius niger) was investigated by Maradonna et al.(2004) and the most relevant finding of this paper wasthe evidence that cytochrome P4501A1 inhibition by estro-gen and estrogen-like compounds is mediated through theactivation of the aryl hydrocarbon receptor repressor.Some studies have demonstrated variations of gene expres-sion patterns for key brain proteins; steroidogenic proteinsand xenobiotic- and steroid-metabolizing P450 isoenzymeof NP-exposed (5–50 lg of NP per l) juvenile Atlantic sal-mon (Salmo salar) (Arukwe, 2005; Meucci and Arukwe,2006a,b). The brain P450 aromatase gene was up-regulatedin the brain after NP-exposure in the hermaphroditic fishRivulus marmoratus, while the ovarian P450 aromatase

gene was significantly down-regulated in the gonad (Leeet al., 2006). In contrast, NP does not have appreciableeffects on hepatic CYP1A and GST activities in male gold-fish at concentrations as low as 100 lg l�1 (Ishibashi et al.,2004). In addition, experiments done with both male andfemale Medaka over 10 days, showed a fairly consistentincrease in the expression levels of CYP19 gene expressionafter exposure (Min et al., 2003). These differences inresults are certainly due, in part, to the diversity of cyto-chrome P450 enzymes that could be involved in both NPdetoxification metabolism and endocrine fine tuning. Stillefforts should be carried out to elucidate the role of differ-ent cytochrome P450 on the NP metabolism.

From this work, we can conclude so far that organismsplaced in the first level of trophic chain are able to signifi-cantly bioconcentrate the pollutant and endocrine disrup-tor NP. These grassed organisms affect the growth ofcrustacean. At intermediate trophic levels organisms couldbe able to transform NP inactivating it for future effects.Our results show, quite clearly, that Artemia are able tometabolize NP and, therefore, reduce any risk of contami-nant exposure to higher trophic levels, i.e. zebrafish. Thiscould be an exception case, because it has been reportedan effective NP bioconcentration in several organismsplaced at second and higher levels of food chain (Servos,1999). It was only when commercial food was contami-nated with NP that effects (i.e. vitellogenin induction) wereobserved in the zebrafish. This suggests then that in someaquatic food chains the higher trophic levels are less likelyto be affected if their diet is live food. Nevertheless, theseresults on one specific food chain does not exclude thatthe organism placed on the top of some trophic chainscould be affected by the presence of NP in their food, inboth the hormone levels and metabolic enzymes. As men-tioned in the Introduction, in superficial waters is possibleto find NP concentration up to 6-times higher than thoseused in this work. Thus, this work shows that the environ-mental presence of NP should be considered as a risk forthe organisms present in an ecosystem.

Acknowledgements

We thank Dr. Carlos Vazques and Dr. Marıa del PilarSanchez for their contribution. We thank Raunel Tinocofor his technical assistance. This work was funded by theNational Council for Science and Technology of Mexico(Grant FOSEMARNAT-2004-01-42).

References

Ahel, M., McEvoy, J., Giger, W., 1993. Bioaccumulation of the lipophilicmetabolites of non-ionic surfactants in freshwater organisms. Environ.Poll. 79, 243–248.

Ahmed, H., Schott, E.J., Gauthier, J.D., Vasta, G.R., 2003. Superoxidedismutase from the oyster parasite Perkinsus marinus: purification,biochemical characterization, and development of a plate microassayfor activity. Anal. Biochem. 318, 132–141.

G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670 669

Arukwe, A., 2005. Modulation of brain steroidogenesis by affectingtranscriptional changes of steroidogenic acute regulatory (StAR)protein and cholesterol side chain cleavage (P450scc) in juvenileAtlantic salmon (Salmo salar) is a novel aspect of nonylphenol toxicity.Environ. Sci. Technol. 39, 9791–9798.

Baldwin, W.S., Graham, S.E., Shea, D., LeBlanc, G.A., 1997. Metabolicandrogenization of femele Daphia magna by the xenoestrogen 4-nonylphenol. Environ. Toxicol. Chem. 16, 1905–1911.

Burke, M.D., Mayer, R.T., 1974. Ethoxyresorufin: direct fluorometricassay of a microsomal O-dealkylation which is preferentially inducibleby 3-methycholanthrene. Drug Metab. Dispos. 2, 583–588.

Bradford, M.M., 1976. A rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254.

Browne, R.A., 1992. Population genetics and ecology of Artemia: insightsinto parthenogenic reproduction. Trends Ecol. Evol. 7, 232–237.

Eklund, R., Bergman, A., Granmo, A., Berggren, M., 1990. Bioaccumu-lation of 4-nonylphenol in marine animals: a re-evaluation. Environ.Poll. 64, 107–120.

Fay, A.A., Brownawell, B.J., Elskus, A.A., McElroy, A.E., 2000. Criticalbody residues in the marine amphipod Ampelisca abdita: sedimentexposures with nonionic organic contaminants. Environ. Toxicol.Chem. 19, 1028–1035.

Fenner-Crisp, P.A., 1997. Endocrine disruptor risk characterization: anEPA perspective. Regul. Toxicol. Pharmacol. 26, 70–73.

Garcıa-Reyero, N., Raldua, D., Quiros, L., Llaveria, G., Cerda, J.,Barcelo, D., Grimalt, J.O., Pina, B., 2004. Use of vitellogenin mRNAas a biomarker for endocrine disruption in feral and cultured fish.Anal. Bionanal. Chem. 378, 670–675.

Graff, L., Isnard, P., Cellier, P., Bastide, J., Cambon, J.P., Narbonne, J.F.,Budzinski, H., Vasseur, P., 2003. Toxicity of chemicals to microalgaein river and in standard waters. Environ. Toxicol. Chem. 22, 1368–1379.

Guillard, R.R., Ryther, J.H., 1962. Studies of marine planktonic diatoms.I. Cyclotella nana Hustedt, and Detonula confervacea (cleve) Gran.Can. J. Microbiol. 8, 229–239.

Hanioka, N., Tararako, N., Jinno, H., Arizono, K., Ando, M., 2000.Determination of cytochrome P450 1A activities in mammalian livermicrosomes by high-performance liquid chromatography with fluo-rescence detection. J. Chromatograp. B 744, 399–406.

Hasselberg, L., Grosvik, B.E., Goksoyr, A., Celander, M.C., 2005.Interactions between xenoestrogens and ketoconazole on hepaticCYP1A and CYP3A, in juvenile Atlantic cod (Gadus morhua). Comp.Hepatol. 4, 2.

Hecht, S.A., Gunnarsson, J.S., Boese, B.L., Lamberson, J.O., Schaffner,C., Giger, W., Jepson, P.C., 2004. Influences of sediment organicmatter quality on the bioaccumulation of 4-nonylphenol by estuarineamphipods. Environ. Toxicol. Chem. 23, 865–873.

Heinis, L.J., Knuth, L., Liber, K., Sheedy, B.R., Tunell, R.L., Ankley, G.T.,1999. Persistence and distribution of 4-nonylphenol following repeatedapplication to the littoral. Environ. Toxicol. Chem. 18, 363–375.

Hemerick, G., 1973. Mass culture 255–273. In: Stein, J.R. (Ed.),Handbook of Phycological Methods. Culture Methods and GrowthMeasurement. Cambridge University Press, New York, p. 448.

Hemmer, M.J., Bowman, C.J., Hemmer, B.L., Friedman, S.D., Marco-vich, D., Kroll, K.J., Denslow, N.D., 2002. Vitellogenin mRNAregulation and plasma clearance in male sheepshead minnows,(Cyprinodon variegatus) after cassation of exposure to 17b-estradioland p-nonylphenol. Aquatic Toxicol. 58, 99–112.

Hill, R.L., Janz, D.M., 2003. Developmental estrogenic exposure inzebrafish (Danio rerio): I. Effects on sex ratio and breeding success.Aquatic Toxicol. 63, 417–429.

Hoss, S., Juttner, I., Traunspurger, W., Pfister, G., Schramm, K.-W.,Steinberg, C.E.W., 2002. Enhanced growth and reproduction ofCaenorhabditis elegans (Nematoda) in the presence of 4-nonylphenol.Environ. Pollut. 120, 169–172.

Huberman, A., 2000. Shrimp endocrinology: a review. Aquaculture 191,191–208.

Ishibashi, H., Tachibana, K., Tsuchimoto, M., Soyano, K., Tatarazako,N., Matsumura, N., Tomiyasu, Y., Tominaga, N., Arizono, K., 2004.Effects of nonylphenol and phytoestrogen-enriched diet on plasmavitellogenin, steroid hormone, hepatic cytochrome P450 1A, andglutathione-S-transferase values in goldfish (Carassius auratus). Comp.Med. 54, 54–62.

Islinger, M., Pawlowski, S., Hollert, H., Volkl, A., Braunbeck, T., 1999.Measurement of vitellogenin-mRNA expression in primary cultures ofrainbow trout hepatocytes in a non-radioactive dot blot/RNAseprotection-assay. Sci. Total Environ. 233, 109–122.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature 227, 680–685.

Lee, P.C., Marquardt, M., Lech, J.J., 1998. Metabolism of nonylphenol byrat and human microsomes. Toxicol. Lett. 99, 117–126.

Lee, Y.M., Seo, J.S., Kim, I.C., Yoon, Y.D., Lee, J.S., 2006. Endocrinedisrupting chemicals (bisphenol A, 4-nonylphenol, 4-tert-octylphenol)modulate expression of two distinct cytochrome P450 aromatase genesdifferently in gender types of the hermaphroditic fish Rivulus marmo-

ratus. Biochem. Biophys. Res. Commun. 345, 894–903.Lewis, S.K., Lech, J.J., 1996. Uptake, disposition, and persistence of

nonylphenol from water in rainbow trout (Oncorhynchus mykiss).Xenobiotica 26, 813–819.

Lussier, S.M., Champlin, D., LiVolsi, J., Poucher, S., Pruella, R.J., 1999.Acute toxicity of para-nonylphenol to saltwater animals. Environ.Toxicol. Chem. 19, 617–621.

Lutz, I., Kloas, W., 1999. Amphibians as a model to study endocrinedisruptors: I. Environmental pollution and estrogen receptor binding.Sci. Total Environ. 225, 49–57.

Maradonna, F., Polzonetti, V., Bandiera, S.M., Migliarini, B., Carnevali,O., 2004. Modulation of the hepatic CYP1A1 system in the marine fishGobius niger, exposed to xenobiotic compounds. Environ. Sci. Tech-nol. 38, 6277–6282.

Marcomini, A., Pavoni, B., Sfriso, A., Orio, A.A., 1990. Persistentmetabolites of alkylphenol polyethoxylates in the marine environment.Marine Chem. 29, 307–323.

Martin-Skilton, R., Thibaut, R., Porte, C., 2006. Endocrine alteration injuvenile cod and turbot exposed to dispersed crude oil and alkylphe-nols. Aquatic Toxicol. 78, S57–S64.

Meldahl, A.C., Nithipatikom, K., Lech, J.J., 1996. Metabolism of several14C-nonylphenol isomers by rainbow trout (Oncorhynchus mykiss):in vivo and in vitro microsomal metabolites. Xenobiotica 26, 1167–1180.

Meucci, V., Arukwe, A., 2006a. The xenoestrogen 4-nonylphenol mod-ulates hepatic gene expression of pregnane X receptor, aryl hydrocar-bon receptor, CYP3A and CYP1A1 in juvenile Atlantic salmon (Salmo

salar). Comp. Biochem. Physiol. C Toxicol. Pharmacol. 142, 142–150.Meucci, V., Arukwe, A., 2006b. Transcriptional modulation of brain and

hepatic estrogen receptor and P450arom isotypes in juvenile Atlanticsalmon (Salmo salar) after waterborne exposure to the xenoestrogen,4-nonylphenol. Aquatic Toxicol. 77, 167–177.

Min, J., Lee, S.K., Gu, M.B., 2003. Effects of endocrine disruptingchemicals on distinct expression patterns of estrogen receptor, cyto-chrome P450 aromatase and p53 genes in oryzias latipes liver. J.Biochem. Mol. Toxicol. 17, 272–277.

Muniz, R.M., Cabral de Oliveira, M., Okamoto, K., Lopes, P.F.,Colepicolo, P., 2002. Effect of visible light on superoxide dismutase(SOD) activity in the red alga Gracilariopsis tenuifrons (Gracilariales,Rhodophyta). J. Appl. Phycol. 14, 151–157.

Nimrod, A.C., Benson, W.H., 1996. Environmental estrogenic effects ofalkylphenol ethoxylates. Crit. Rev. Toxicol. 26, 335–364.

Nunes, B.S., Carvalho, F.D., Guilhermino, L.M., Van Stappen, G., 2006.Use of the genus Artemia in ecotoxicity testing. Environ. Pollut. 144,453–462.

Pait, A.S., Nelson, J.O., 2003. Vitellogenesis in male Fundulus heteroclitus

(killifish) induced by selected estrogenic compounds. Aquatic Toxicol.64, 331–342.

Servos, M.R., 1999. Review of the aquatic toxicity, estrogenic responsesand bioaccumulation of alkylphenols and alkylphenols polyethoxy-lates. Water Qual. Res. J. Can. 34, 123–177.

670 G. Correa-Reyes et al. / Chemosphere 68 (2007) 662–670

Staples, C.A., Weeks, J., Hail, J.F., Naylor, C.G., 1998. Evaluation of theaquatic toxicity and bioaccumulation of C8- and C9-alkylphenolethoxylates. Environ. Toxicol. Chem. 17, 2470–2480.

Staples, C., Mihaich, E., Carbone, J., Woodbrun, K., Klecka, G., 2004. Aweight of evidence analysis of the chronic ecotoxicity of nonylphenolethoxylates, nonylphenol ether carboxylates, and nonylphenol. HumanEcol. Risk Assessm. 10, 999–1017.

Teles, M., Gravato, C., Pacheco, M., Santos, M.A., 2004. Juvenile seabass biotransformation, genotoxic and endocrine responses to beta-naphthoflavone, 4-nonylphenol and 17 beta-estradiol individual andcombined exposures. Chemosphere 57, 147–158.

Thibaut, R., Debrauwer, L., Perdu, E., Goksoyr, A., Cravedi, J.P.,Arukwe, A., 2002. Regio-specific hydroxylation of nonylphenoland the involvement of CYP2K- and CYP2M-like iso-enzymes inAtlantic salmon (Salmo salar). Aquatic Toxicol. 56 (Feb), 177–190.

Van den Belt, K., Berckmans, P., Vangenechten, C., Vergeyen, R.,Witters, H., 2004. Comparative study on the in vitro/in vivo estrogenicpotencies of 17b-estardiol, estrone, 17b-ethynyestradiol and nonylphe-nol. Aquatic Toxicol. 66, 183–195.

Van Stappen, G., 2002. Zoogeography. In: Abatzopoulos, Th.J., Beard-more, J.A., Clegg, J.S., Sorgeloos, P. (Eds.), Artemia: Basic andApplied Biology. Kluwer Academic Publishing, pp. 171–224.

Vazquez-Duhalt, R., Marquez-Rocha, F., Ponce, E., Licea, A.F., Viana,M.T., 2006. Nonylphenol, an integrated vision of a pollutant. Appl.Ecol. Environ. Res. 4, 1–25.

Weber, L.P., Hill, R.L., Janz, D.M., 2003. Developmental estrogenicexposure in zebrafish (Danio rerio): II Histological evaluation ofgametogenesis and organ toxicity. Aquatic Toxicol. 63, 431–446.

Willey, J., Krone, P.H., 2001. Effects of endosulfan and nonylphenol onthe primordial germ cell population in pre-larval zebrafish embryos.Aquatic Toxicol. 54, 113–123.

Yadetie, F., Arukwe, A., Gokøyr, A., Male, R., 1999. Induction of hepaticestrogen receptor in juvenile Atlantic salmon in vivo by the environ-mental estrogen, 4-nonylphenol. Sci. Total Environ. 233, 201–210.

Yamauchi, K., Ishihara, A., Fukazawa, H., Terae, Y., 2003. Competitiveinteractions of chlorinated phenol compounds with 3,3 0,5-triiodothy-ronine binding to transthyretin: detection of possible thyroid-disrupt-ing chemicals in environmental waste water. Toxicol. Appl.Pharmacol. 187, 110–117.

Ying, G.G., Williams, B., Kookana, R., 2002. Environmental fate ofalkylphenols and alkylphenol ethoxylates. A review. Environ. Int. 28,215–226.

Zhang, J., Shen, H., Wang, X., Wu, J., Xue, Y., 2004. Effects of chronicexposure of 2,4-dichlorophenol on the antioxidant system in liver offreshwater fish Carassius auratas. Chemosphere 55, 167–174.