Evaluation of a carp primary hepatocyte culture system for screening chemicals for oestrogenic activity

Aquatic Toxicology 94 (2009) 195–203

Contents lists available at ScienceDirect

Aquatic Toxicology journal homepage: www.elsevier.com/locate/aquatox

Evaluation of a carp primary hepatocyte culture system for screening chemicals for oestrogenic activity L.K. Bickley a,∗ , A. Lange a , M.J. Winter b , C.R. Tyler a a b

School of Biosciences, University of Exeter, Hatherly Laboratories, Prince of Wales Road, Exeter, Devon, EX4 4PS, United Kingdom Brixham Environmental Laboratory, AstraZeneca Limited, Freshwater Quarry, Brixham, Devon, TQ5 8BA, United Kingdom

a r t i c l e

i n f o

Article history: Received 13 January 2009 Received in revised form 1 July 2009 Accepted 4 July 2009 Keywords: In vitro culture Primary hepatocytes Common carp (Cyprinus carpio) 17␤-Oestradiol (E2)

a b s t r a c t The presence of endocrine disrupting chemicals (EDCs) in the environment has driven the development of screening and testing assays to both identify chemicals with hormonal activity and evaluate their potential to cause adverse effects. As the number of animals used for research and regulatory purposes rises, and set against a desire to reduce animal testing, there is increased emphasis on the development and application of in vitro techniques to evaluate chemical risks to the environment. Induction of vitellogenin (VTG) in isolated fish liver cells has been used successfully to identify a wide range of EDCs, including both natural and synthetic oestrogens and a variety of other xenoestrogens. However, the vitellogenic response reported for hepatocytes in culture has been shown to vary widely, making comparisons between studies difficult. The work presented in this paper explored the variability of the vitellogenic response in primary cultures of common carp (Cyprinus carpio) hepatocytes following exposure to the model oestrogenic compound, 17␤-oestradiol (E2). As expected, variability in the vitellogenic response was observed, both in terms of the sensitivity and magnitude of VTG induction, for hepatocytes isolated from different fish. An apparent difference was observed in the response of isolated hepatocytes based on the sex of the donor fish; maximum levels of E2-stimulated VTG synthesis in hepatocytes derived from females appeared higher (1962 ng mL−1 ± 487 [n = 9] compared with 1194 ng mL−1 ± 223 for hepatocytes from males [n = 9]) and EC50 values lower (1.61 ± 0.4 ␮M E2 for females and 2.12 ± 0.2 ␮M E2 for males). However, these differences were not statistically significant, likely in part due to the variation observed in the vitellogenic response. In particular, hepatocytes derived from female fish showed more variation than their male counterparts (the co-efficient of variation for females was 77% compared to 28% for males). Despite the variation observed in the vitellogenic response between different cultures, data from the different donor fish could be compared by standardising responses relative to the maximum VTG induction in each culture following exposure to E2. Adopting this approach in the future will allow for data from different hepatocyte cultures and from donor fish of different sexes, age and stage of maturity to be compared with greater consistency. Measurement of vtg mRNA expression was relatively more sensitive to the oestrogenic effects of E2 exposure than measurement of VTG protein (the LOEC at the transcriptome level was 10-fold lower [0.01 ␮M E2] than at the protein level [0.1 ␮M E2]) and changes in vtg mRNA expression showed less variation between individual hepatocyte isolations. Measurement of vtg mRNA in the hepatocyte culture system therefore may offer the most sensitive and consistent option for the screening of chemicals with oestrogenic activity in fish primary hepatocyte cultures. © 2009 Elsevier B.V. All rights reserved.

1. Introduction It is now well established that there are a wide variety of chemicals in the aquatic environment that have the capacity to act upon

Abbreviations: E2, 17␤-oestradiol; EDC, endocrine disrupting chemical; ELISA, enzyme-linked immunosorbent assay; ER, oestrogen receptor; EC50 , 50% effect concentration; FBS, foetal bovine serum; IMS, industrial methylated spirits; LDH, lactate dehydrogenase; LOEC, lowest observed effect concentration; VTG, vitellogenin. ∗ Corresponding author. Tel.: +44 (0)1392 263438; fax: +44 (0)1392 263700. E-mail address: [email protected] (L.K. Bickley). 0166-445X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2009.07.006

endocrine systems in wildlife. These chemicals, known as endocrine disrupting chemicals (EDCs), can adversely affect normal physiological functions, most notably those relating to reproduction and development (Crain and Guillette, 1997; Tyler et al., 1998; Vos et al., 2000). Concerns about the presence of EDCs in the environment have led to the development of screening and testing assays that are able to identify such substances and/or evaluate their potential to induce adverse effects. With the numbers of animals used for research and regulatory studies rising and set against a desire to reduce animal testing, there is increased emphasis on the development and application of in vitro techniques to evaluate chemical risks to the environment (Ankley et al., 1998;

196

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

ICCVAM, 2002; Strategy for a Future Chemicals Policy, 2001) and fish cell based assays, both primary and immortalised, are gaining ˜ et al., 2003; Navas increased favour and wider application (Castano and Segner, 2006; Schirmer, 2006). In vitro cell culture systems allow for precise control of the physical and chemical environment, which is often not possible in vivo. Cell culture systems also provide the opportunity to study toxic mechanisms in isolation from the multiple physiological systems that regulate normal in vivo activity. Eliminating the interactive systemic effects within complex whole animal responses can simplify data interpretation, and more easily reveal specific toxic mechanisms. Furthermore, in vitro models such as cell culture systems have the potential to significantly reduce the number of animals sacrificed for (eco)toxicological research. This said, however, the fact that they are not integrated biological systems in which all biochemical and molecular interactions are represented, means that they must be rigorously established and validated to ensure appropriate comparisons with in vivo situations. The liver plays a fundamental role in maintaining the internal homeostasis of vertebrates and plays a major role in the regulation of many metabolic and physiological processes. It is also central to many aspects of xenobiotic metabolism and is often a major target for chemical toxins. Consequently, isolated hepatocytes have been widely used in pharmacology and environmental toxicology. In fish and other oviparous vertebrates the liver is also the site of oestrogen-dependent synthesis of the egg yolk precursor protein, vitellogenin (VTG), one of the most widely utilised biomarkers for screening oestrogen exposure in aquatic organisms (Fenske et al., 2001; Mommsen and Walsh, 1988; Nilsen et al., 2004; Silversand et al., 1993; Tyler et al., 1996). VTG is inducible in both male and female fish, and most evidence supports induction operating through activation of the oestrogen-receptor alpha (ER␣) (Katsu et al., 2007). Isolated fish hepatocytes that are maintained in primary culture have been shown to retain their capacity to produce VTG following in vitro exposure to oestrogenic chemicals (Jobling and Sumpter, 1993; Navas and Segner, 2006; Pelissero et al., 1993; Rankouhi et al., 2002, 2004; Smeets et al., 1999b; Tollefsen et al., 2003) and to retain at least some of their in vivo metabolic capabilities (Cravedi and Zalko, 2005; Pedersen and Hill, 2000; Segner and Cravedi, 2001). Thus, induction of VTG in isolated hepatocytes has the potential to detect the effects of indirect oestrogens, for example those that require metabolic activation. VTG induction in primary cultures of fish hepatocytes has been exploited successfully to detect a wide range of EDCs including: natural and synthetic oestrogens (e.g. 17␤-oestradiol [E2] and ethinyloestradiol); and xenobiotics such as pesticides, polychlorinated biphenyls, alkylphenol (octyl- and nonyl-phenol), phthalates and bisphenol A (see Navas and Segner, 2006 for review). However, there is very pronounced variability in the reported levels of oestrogen-stimulated VTG induction, both in terms of the amount of VTG produced, and the effect concentrations for specific chemicals. For example, following in vitro exposure to E2 reported 50% effect concentrations (EC50 values) for primary cultures of fish hepatocytes vary between 10−12 and 10−6 M (Anderson et al., 1996; Flouriot et al., 1993; Gagné et al., 1999; Islinger et al., 1999; Jobling and Sumpter, 1993; Kim and Takemura, 2003; Kwon et al., 1993; Latonnelle et al., 2000; Navas and Segner, 2000; Pelissero et al., 1993; Rankouhi et al., 2004; Smeets et al., 1999a, 1999b; Tollefsen et al., 2003; Vaillant et al., 1988). Some of this variability may be accounted for by differences in the experimental techniques employed: both to isolate and culture hepatocytes and also to measure VTG. It is also reasonable to expect that the oestrogen sensitivity of cultured hepatocytes may vary between species, as highlighted in an interspecies comparative analysis by Rankouhi et al. (2004) and as reported to occur in vivo (Purdom et al., 1994; Routledge et al., 1998; Tyler

et al., 2009). Whilst these factors can be standardised for future studies, there are other factors likely to influence the sensitivity of the oestrogenic response that remain less well understood, including sex, stage of sexual development and seasonality. If primary cultures of fish hepatocytes are to be used in the routine screening of chemicals for (anti-)oestrogenic activity, it is essential that the in vitro vitellogenic response is characterised with the aim of ensuring such responses are both consistent and repeatable. The work presented in this paper explores the variability of the vitellogenic response of primary cultures of common carp (Cyprinus carpio) hepatocytes following exposure to the model oestrogenic compound, E2. The vitellogenic response from 18 individual hepatocyte isolates was characterised and compared, taking into account various physiological factors, including: the time of year of isolation; the sex, developmental status and size of each fish; and plasma VTG levels before cell isolation and basal VTG levels of hepatocytes following isolation. 2. Material and methods 2.1. Animals Common carp (C. carpio; obtained from Frontline Fish, Devon, UK) were maintained under a 12 h light–dark photoperiod in aerated tanks equipped with a flow-through water system, at a temperature of 15 ± 1 ◦ C. Fish were fed, ad libitum, with commercial pellets (Nishikoi Staple Food, Nishikoi Aquaculture, UK). Feeding was withheld 1 day prior to cell isolation. 2.2. Cell isolation All chemicals and reagents were obtained from Sigma–Aldrich, Poole, UK, unless stated otherwise. The isolation of carp hepatocytes was carried out according to Segner et al. (1993) with some modifications. All animal-use procedures were carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Prior to isolation, fish were temporarily anaesthetised by immersion in benzocaine (ethyl 4-aminobenzoate; 100 g L−1 ethanol, dissolved to ∼500 mg L−1 in tank water), and 200 ␮L blood were collected by caudal vessel puncture and dispensed into chilled, pre-heparinised syringes. Blood samples were centrifuged at 10 000 × g for 5 min and the plasma removed and stored at −20 ◦ C until analysis for VTG. After blood sampling, fish were injected with 1000 units of heparin (heparin sodium, 5000 Units mL−1 saline, CP Pharmaceuticals, UK) and then returned to freshwater for 10 min. Fish were terminally anaesthetised before commencing cell isolation. A two-step collagenase perfusion technique was used, performed completely in situ. All solutions were adjusted to pH 7.5, and 0.22 ␮M sterile filtered. Pre- and post-perfusion solutions were perfused at a flow rate of 15 mL min−1 , and the perfusion solution (containing 0.02% collagenase D [w/v], Roche Diagnostics, Switzerland) was perfused at 10 mL min−1 . The liver was dissected from the body cavity and placed into Ca2+ - and Mg2+ -free HEPES-buffered Hanks salt solution. Following this, the liver was finely chopped and the resultant cell suspension flushed through nylon screens of 250, 100 and 50 ␮m mesh size (Normesh Ltd., Oldham, UK). Hepatocytes were then separated by low speed centrifugation at 44 × g at 4 ◦ C for 5 min. The cell pellet was washed three times and re-suspended in culture medium, before cell number was counted using a haemocytometer. Cell viability was estimated immediately following isolation using the trypan blue exclusion assay. All cell isolates showed initial viability >85%. Before discarding the body of the donor fish, gonads were removed and processed for histological analysis to stage their sexual development.

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

2.3. Cell culture A final cell density of hepatocytes was adjusted to 1.5 × 106 cells mL−1 in M199 culture medium (supplemented with 3.5 mM HEPES, 4.1 mM NaHCO3 , 3.4 mM CaCl2 , 2 mM lglutamine, 10 U mL−1 penicillin, 10 ␮g mL−1 streptomycin and 10% foetal bovine serum [FBS]). Cells were seeded into 24-well Falcon PrimariaTM culture plates (Becton Dickinson, Oxford, UK) with 400 ␮L cell suspension per well (6 × 105 cells well−1 ). Cells were cultured in a humidified atmosphere at 20 ◦ C. Once seeded, cells were allowed to attach to culture plates for 24 h without exposure to chemicals. At 24 h, 50% of the culture medium (200 ␮L) was removed and replaced with fresh medium without FBS and with the appropriate concentration of E2 (98% purity; lot 103k1117) dissolved in ethanol (final concentration 0.01%). Concentrations of test chemical were based on nominal values (0.01, 0.1, 1.0, 10 ␮M E2). Each experiment included medium and solvent controls and each treatment was replicated in at least 4 wells. Hepatocytes isolated from each fish were exposed for 4 days with medium changes every 24 h. In order to limit cell detachment during medium changes, only 50% of the culture medium was replaced (200 ␮L).

197

2005). Annealing temperature (Ta ) was 61.5 ◦ C and PCR efficiency (E) was 1.89. Ribosomal protein l8 (rpl8) was used for efficiency-corrected relative quantitation because its expression in hepatocytes did not change following E2 exposure (NCBI GenBank accession number AY919670; 5 -CTCCGTCTTCAAAGCCCATGT-3 /5 TCCTTCACGATCCCCTTGATG-3 ; Ta = 60 ◦ C and E = 2.10). 2.7. Gonad histology Histological analysis of the gonads was performed to enable determination of sex and stage of sexual development for each donor fish. After excision, gonads were fixed in Bouin’s solution (Fisher Scientific, Leicestershire, UK) for 8 h and subsequently washed twice in 70% industrial methylated spirits (IMS). Samples were progressively dehydrated in IMS up to 100% and embedded in paraffin wax using an automated tissue processor. Serial sections were cut at 5 ␮m, floated in a water-bath (45 ◦ C) and collected on glass microscope slides. Tissue sections were stained with haematoxylin and eosin, the slides mounted using HistomountTM (National Diagnostics, USA) and the sections analysed to determine the stage of germ cell development by light microscopy.

2.4. Cell viability assessments Cell viability was determined via lactate dehydrogenase (LDH) leakage from hepatocytes into the culture medium. LDH was measured using the CytoTox 96® non-radioactive cytotoxicity assay (Promega, Southampton, UK) according to manufacturer’s instructions. Absorbance was recorded at 490 nm using a SPECTRAmaxTM 340 PC microplate spectrophotometer. 2.5. VTG measurement VTG in plasma and hepatocyte culture medium supernatant samples was measured using the carp-VTG enzyme-linked immunosorbent assay (ELISA), as described by Tyler et al. (1999). Cell culture medium was diluted between 10 and 10 000-fold prior to analysis. There was no interference of the cell culture supernatant in the assay at any of the dilutions used (data from a two-fold serial dilution of cell culture medium remained parallel with the standard curve: data not shown). The detection limit of the assay was 2 ng L−1 VTG. Inter- and intra-assay variability were as reported in Tyler et al. (1999). 2.6. vtg mRNA analysis For measurements of vtg mRNA, after 96 h exposure, cell culture supernatant was removed, Tri-Reagent® RNA isolation reagent (Sigma–Aldrich, Poole, UK) added and cells were lysed. Cell lysates were snap frozen in liquid nitrogen and stored at −80 ◦ C until use. Total RNA was extracted from hepatocytes using Tri-Reagent according to manufacturer’s instructions. Total RNA concentration was estimated from absorbance at 260 nm (A260 nm ; Gene Quant, Amersham Pharmacia Biotech, UK) and RNA quality verified by electrophoresis on ethidium bromide-stained 1.5% agarose gels and by A260 nm /A280 nm ratios >1.8. RNA was DNase-treated and reversetranscribed to cDNA as described by Filby and Tyler (2005). Primers specific for common carp vtg (NCBI GenBank accession number AF414432) were designed with Beacon Designer 3.0 software (Premier Biosoft International) and purchased from MWG-Biotech (Eurofins, Germany) (vtg: 5 -TGAACAGTGAGAAAGAGATTGAAC3 /5 -ATTGATGGGAATGGCGTAGG-3 ). Real-time quantitative PCR was performed in triplicate for each sample with the iCycler iQ Real-time Detection System (Bio-Rad Laboratories Inc., CA) using methods described previously (Filby and Tyler,

2.8. Data analysis Throughout this paper, data are presented as mean ± standard error of the mean. All statistical analyses were carried out using SigmaStat 3.1 (Systat Software Inc., USA) unless stated otherwise. Data were assessed for normality and homogeneity of variances using the Kolmogorov–Smirnov and Levene’s Median tests, respectively. Non-normal data were log-transformed and these tests repeated. If assumptions of normality were met, data were analysed using Analysis Of Variance (ANOVA). Data were considered statistically significant at P < 0.05. If the null hypothesis was rejected, means were compared using the Holm–Sidak multiple comparison method. When normality assumptions were not met, Kruskal–Wallis one-way ANOVA was used to compare several medians and Dunn’s multiple comparison procedure applied if the null hypothesis was rejected. Independent non-paired t-tests were used to compare the means of two groups. Linear regression was performed using the least squares method. Concentration effects curves for VTG induction were constructed and EC50 values were calculated using GraphPad Prism 3 DEMO version (GraphPad Software Inc., San Diego, CA). Concentration–effect curves were constructed using non-linear regression. VTG induction data were first normalised relative to maximum VTG induction following exposure to E2, constraints were set between 100% (or maximum induction) and baseline (or control) levels of VTG, and concentration values (×data) were plotted on a log scale. Data in GraphPad Prism was fitted to the equation below, with a standard slope (Hillslope) of 1.0: Y = Baseline+

Maximum − Baseline 1 + 10((log EC50 −x)Hillslope)

where Baseline is the level of response in the absence of E2 (i.e. solvent control value), Maximum the maximum response following exposure to E2 (i.e. 100% VTG induction) and Hillslope the steepness of dose–response curve. The EC50 value refers to the concentration of E2 that induced half-maximal induction of VTG. The lowest observed effect concentration (LOEC) refers to the lowest E2 test concentration inducing a response that differed significantly from that of the solvent control treatment group.

198

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

Fig. 1. Changes in morphology of isolated hepatocytes in culture with increasing incubation time. Images were taken at (a) 0 h (immediately after hepatocytes were seeded into culture plates), (b) 24 h, (c) 32 h, (d) 72 h, (e) 96 h and (f) 120 h. Bars = 100 ␮m.

3. Results Hepatocytes maintained good viability (>85%) for the first 5 days in culture, and showed a series of morphological changes with increasing duration of incubation. Immediately upon isolation, hepatocytes were seen as single separated cells, which were rounded in appearance (Fig. 1a). A few hours after plating, hepatocytes began to attach to the surface of the culture plates and within 24 h, cells were clustering into groups forming strands of between 2 and 10 cells (Fig. 1b). Over time, larger clusters of hepatocytes formed and their rounded appearance was lost as the boundaries between cells became increasingly difficult to define (Fig. 1c–e). After approximately 4 days in culture, a monolayer network of hepatocytes had formed across the plate (Fig. 1f). Hepatocytes remained in this state for up to 8 days before cell viability began to decline significantly (Fig. 2a) and cell detachment from the culture plates occurred. To assess the oestrogenic response of hepatocytes in primary culture following exposure to E2 over 10 days, the amount of VTG secreted into the culture medium was measured every 24 h. Exposure to increasing concentrations of E2 resulted in an increase in the amount of VTG secreted into the medium in a concentrationrelated manner (Fig. 2b). There were no significant differences in VTG induction between culture medium and solvent control treatment groups throughout the exposure period. Without oestrogenic stimulation, average basal VTG levels in the culture medium remained low (<5 ng mL−1 ). Subsequent histological examination

revealed cells in this case were isolated from a male fish, with highly spermiating testes. E2-induced VTG synthesis peaked on exposure day 8, with up to a 1200-fold increase in VTG above the solvent control treatment group. After 8 days of exposure, VTG levels decreased in the E2-exposed treatment groups, corresponding with a decline in cell viability (Fig. 2a), and this was associated with an observed increase in cell detachment from the culture plates and changes in the morphological characteristics of the cells (hepatocytes appeared smaller and denser, and showed reduced cell to cell contact; data not shown). The VTG induction response showed variation between replicate wells of the same treatment (e.g. for cells exposed to 10 ␮M E2, co-efficient of variation values ranged between 2% and 39% across exposure days 1–9 and 72% on exposure day 10). However, there was no correlation between the amount of variation between replicate wells and the length of exposure time (linear regression, F1,9 = 0.9, R2 = 0.11, P = 0.68). Table 1 gives information on the general oestrogenic response of hepatocytes, from each individual isolation performed; including plasma VTG levels of donor fish, baseline VTG levels in medium control treatment groups, maximum VTG induction in hepatocytes following in vitro E2 exposure, and corresponding EC50 values. It also shows the sex of the donor fish and the season in which the isolation was performed. The average number of cells isolated was 6.7 ± 0.07 × 105 hepatocytes g−1 fish (n = 18) and the average viability of cells immediately following isolation was 86 ± 1.3% as estimated by the trypan blue exclusion assay. The number of cells

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

199

Table 1 Information on the physiology of each donor fish from which hepatocytes were isolated and the subsequent oestrogenic response of those hepatocytes in culture. Sex of donor fish

Month of cell isolation

Plasma VTG (ng/mL)

Male Male Female Female Female Female Female Male Female Male Female Male Male Male Male Female Male Female

January January January February February April April June June July July October October October November November November December

18 21 634 475 826 4482 1867 13 3520 33 36 171 22 40 454 163 8 2023

Baseline VTG (ng/mL)

7 3 5 4 3 7 6 9 2 4 4 6 2 6 2 2 33

isolated was significantly greater in fish of a larger size (linear regression, F1,16 = 17.5, R2 = 0.57, P < 0.01), which is likely related to the larger mass of the liver. However, liver weight was not measured due to the in situ perfusion technique employed for the cell isola-

Fig. 2. (a) Viability and (b) induction of VTG in primary cultures of carp hepatocytes following exposure to E2 over 10 days. Each symbol or column represents mean ± SE, n = 4 replicates from one cell isolate. * denotes significantly different to solvent control treatment group within each day of exposure and “a” denotes significant differences to viability on day 1 (two-way ANOVA, with Holm–Sidak multiple comparison procedure, P < 0.05). Hepatocytes isolated from a male donor fish.

Max VTG induction (ng/mL)

1511 1417 3692 40 1653 3876 855 223 1962 1417 142 1739 1400 851 2081 2021 104 3415

EC50 (96 h) ␮M

Std err log EC50

2.6 2.4 1.1 3.7 0.5 0.6 3.2 2.0 0.4 2.2 2.7 1.9 1.7 2.6 0.8 1.1 2.7 1.3

0.1 0.4 0.1 0.1 0.1 0.2 0.1 0.1 0.1 0.3 0.1 0.3 0.1 0.4 0.2 0.2 0.4 0.1

tion. No significant relationships were found between initial cell viability (data not shown) and subsequent vitellogenic response, or between plasma VTG levels in the donor fish, basal VTG levels of hepatocytes in control treatment groups and subsequent vitellogenic response. There were no significant effects in the variation observed in the vitellogenic response relating to either the season in which hepatocytes were isolated or the size (length and weight; data not shown) of the donor fish. Fig. 3 shows the calculated concentration-related response curves for all individual cell isolates, before (Fig. 3a) and after (Fig. 3b) data were normalised by expressing VTG levels as a percentage relative to maximum induction at 10 ␮M E2. This figure highlights the variability in response between hepatocytes isolated from different fish. Once normalised, it was considerably easier to compare the vitellogenic response of individual cell isolates, and apparent differences in response between male- and femalederived hepatocytes were observed. However, hepatocytes derived from both male and female fish showed an average LOEC of 1 ␮M (Fig. 4a and b). Furthermore, there was no statistically significant sex specific difference in the maximum levels of E2-stimulated VTG synthesis (1962 ng mL−1 ± 487 for hepatocytes derived from females [n = 9] compared with 1194 ng mL−1 ± 223 for hepatocytes from males [n = 9]; Fig. 4d). The EC50 value for hepatocytes derived from female fish was lower (1.61 ± 0.4 ␮M E2) compared with hepatocytes derived from male fish (2.12 ± 0.2 ␮M E2; Fig. 4c); however, due to the variability between the individual cultures, these differences were not statistically significant. Variation in the vitellogenic response between individual cell isolates was higher in hepatocytes isolated from female fish compared with hepatocytes derived from male fish (the co-efficient of variation, when considering EC50 values, was 77% for females compared to 28% for males). Histological examination of the gonads revealed that all the male fish studied were mature and spermiating and all females had not yet entered vitellogenesis, with primary oocytes as the most advanced gonadal cell type. Three hepatocyte cultures previously analysed for VTG protein were also used to analyse vtg mRNA expression. These hepatocytes were all isolated from mature male fish with an average weight and standard length of 110 ± 5 g and 18 ± 0.2 cm, respectively. The VTG protein response to 96 h E2 exposure specifically in these hepatocyte cultures is shown in Fig. 5a. The average LOEC was 0.1 ␮M and the average maximum VTG induction at 10 ␮M E2 was 1959 ± 1038 ng mL−1 (an approximately 500-fold induction compared to VTG levels in the solvent control treatment group). Fig. 5b shows the relative expression of vtg mRNA in the same

200

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

Fig. 3. (a) Absolute and (b) normalised (relative to maximum induction at 10 ␮M E2) E2-stimulated VTG induction in primary carp hepatocytes. Data shown are from hepatocytes isolated from 18 fish, of which 9 were female (shown in red), and 9 were male (shown in blue). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

primary cultures of carp hepatocytes at the same time point. As expected, vtg analysis showed concentration-related increases in relative expression consistent with the response observed from VTG protein analysis. However, the average LOEC was ≤0.01 ␮M E2, with maximum relative vtg expression (at 10 ␮M) over 80 000 times higher than that observed for the solvent control. Furthermore, less variation was seen in the vtg mRNA response compared to the VTG protein response (the co-efficient of variation was 40% for vtg mRNA induction at 10 ␮M E2 compared to 92% for VTG protein induction at the same concentration). 4. Discussion Carp hepatocytes in culture during the 4-day exposure period maintained good cell viability and the morphological changes occurring over this time were in accordance with that reported for

previous studies on fish hepatocytes (Blair et al., 1990; Klaunig et al., 1985; Segner et al., 1994; Segner, 1998). Significant decreases in viability were not seen across any treatment groups until exposure day 9, when there was a progressive increase in the number of hepatocytes detached from the culture plates, as has been reported for the few other studies that have undertaken longer term hepatocyte cultures. Deterioration of cellular integrity in rainbow trout (Oncorhynchus mykiss) hepatocytes in culture has been shown previously to occur after 5 days in culture and is related to senescence (Braunbeck and Storch, 1992). These authors showed that between days 4 and 8 in culture, there were changes in the cytology of the hepatocytes that included, an increase in heterochromatin content; elongation and increased heterogeneity of the mitochondria; progressive fractionation and vesiculation of the rough endoplasmic reticulum; an increase in the number of lysosomes, myelinated bodies and autophagic vacuoles; and a successive decrease in cellular

Fig. 4. Induction of VTG in primary cultures of carp hepatocytes following exposure to E2 for 4 days. (a) Average VTG induction in female-derived hepatocytes, (b) average VTG induction in male derived hepatocytes, (c) average EC50 values and (d) average maximum VTG induction in female and male derived hepatocytes. Each box indicates interquartile range, median (solid line), mean (dotted line) and SE. * denotes statistically significant differences between solvent control and E2-exposed treatment groups, n = 9 (4 replicates per treatment group for each cell isolate). EC50 values or maximum VTG induction levels were not significantly different between male- and female-derived hepatocytes (non-paired t-test, P > 0.05).

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

201

Fig. 5. The vitellogenic response of hepatocytes in primary culture following exposure to E2 for 4 days in (a) average VTG protein induction and (b) average relative expression of vtg mRNA. Each column represents mean ± SE, n = 3, with at least 4 replicates per treatment per fish. Hepatocytes derived from mature male fish. Different letters denote significant differences between treatment groups (Kruskal–Wallis one-way ANOVA, with Dunn’s multiple comparison procedure, P < 0.05).

glycogen deposits. Measuring LDH leakage into the culture medium in our study constituted an indirect assessment of cell viability only, however, it correlated well with the microscopic cytological assessments of hepatocytes, as well as their functional capability (VTG production in response to oestrogen exposure). Concentration-related vitellogenic response to E2 demonstrated the isolated hepatocytes were functional and retained their ability to respond to oestrogenic stimulation over a 10-day period of exposure. The response threshold (0.01 ␮M E2 [10−8 M]) corresponded well with previous studies on VTG induction in isolated fish hepatocytes, where reported LOECs vary from 10−12 to 10−6 M E2 (Jobling and Sumpter, 1993; Pelissero et al., 1993; Rankouhi et al., 2002, 2004; Smeets et al., 1999b; Tollefsen et al., 2003). However, the vitellogenic response in isolated carp hepatocytes appears less sensitive to E2 stimulation compared with that observed in carp in vivo (10−10 E2; Gimeno et al., 1998). Comparisons of chemical potency between in vivo exposures with in vitro exposures, however, are difficult, as factors such as bioconcentration and metabolism need to be considered. Our initial studies monitoring VTG production and hepatocyte viability in culture (over a period of 10 days) led us to choose a 5-day culture period for all subsequent hepatocyte culture assessments, as this represented a compromise between an enhanced capacity to synthesise VTG with time, up to 8 days (when maximal VTG induction occurred) and a significant decline in cell viability between days 6 and 10. VTG induction in intact fish is now a well-characterised method for chemical screening and environmental monitoring (Burki et al., 2006; Purdom et al., 1994; Thorpe et al., 2003). However, induction of VTG in isolated hepatocytes is less well-studied and major differences in the vitellogenic response thresholds across and between different fish species has been reported. In carp, minimum concentrations of E2 reported to elicit a vitellogenic response vary between 10−9 and 10−6 M (Rankouhi et al., 2004; Smeets et al., 1999a), whereas in rainbow trout reported concentrations vary between 10−11 and 10−7 M E2 (Islinger et al., 1999; Jobling and Sumpter, 1993; Okoumassoun et al., 2002; Pelissero et al., 1993; Tremblay and Van der Kraak, 1998). Other species that have been used to assess the vitellogenic response in isolated hepatocytes include the bream (Abramis brama) and Atlantic salmon (Salmo salar) where minimal concentration of E2 eliciting VTG production were 10−6 and 10−12 M, respectively (Rankouhi et al., 2004; Tollefsen et al., 2003). In an effort to investigate the nature of this variability, we compared the oestrogenic responses of hepatocytes isolated from 18 carp, taking into account sex, age, size of fish, and time of year that the isolations were undertaken. In doing so, apparent differences were observed in the response of hepatocytes isolated from male fish compared to those isolated from female fish. VTG synthesis depends on the expression of oestrogen receptors and hepatocytes of female

origin have been shown to have a higher constitutive expression of these receptors (Filby and Tyler, 2005; Halm et al., 2004; Pakdel et al., 1997) and thus it might be expected that hepatocytes isolated from female fish may respond more quickly to stimulation or show greater sensitivity to oestrogen exposure. Indeed, Riley et al. (2004) found this to be the case in tilapia (Oreochromis mossambicus), where E2 at 10−7 M was sufficient to significantly induce VTG production in hepatocytes isolated from females, compared with a concentration of 10−4 M E2 required to do so in hepatocytes isolated from males. In our study on carp, hepatocytes isolated from females appeared to show, on average, higher maximum VTG induction levels and lower EC50 values compared with males, however, this apparent difference was not statistically significant (due to the considerable variation in the response for females). Smeets et al. (1999a), using genetically uniform strains of carp in an effort to limit inter-individual and inter-experimental variation, similarly found no difference in the reported EC50 levels between males and females but did however, find hepatocytes isolated from female fish produced significantly higher maximum levels of VTG. An interesting point to note from our work presented here is that hepatocytes isolated from female fish showed approximately three times more variation in their overall oestrogenic response compared with hepatocytes isolated from male fish. This may be related to variation in the reproductive cycle of female fish at the time of hepatocyte isolation or higher variation in the constitutive expression of ERs in female fish. Previous studies (Navas and Segner, 2006; Pelissero et al., 1993) have reported a predictable absence of basal production of VTG in hepatocytes isolated from male fish, whereas hepatocytes isolated from some female fish produce VTG without the addition of an oestrogenic stimulation. In the present work, however, there were no significant differences in the basal levels of VTG synthesised by hepatocytes isolated from either male or female fish. Nor were there any relationships between plasma VTG levels in donor fish or basal VTG levels of hepatocytes immediately following isolation, and the subsequent oestrogenic response of hepatocytes when stimulated by E2. The absence of VTG production without E2 stimulation in the present study is likely due to the stage of gonadal development of the donor fish: female fish were still at a relatively immature stage of sexual development, with ovaries containing primary oocytes only (i.e. they had not yet entered the vitellogenesis). In a study by Pelissero et al. (1993) hepatocytes isolated from male or immature female rainbow trout released no VTG protein into the culture medium whereas hepatocytes isolated from female fish at later (vitellogenic) stages of development synthesized and excreted considerably higher amounts of VTG without stimulation. In the final part of the work presented in this paper we compared the VTG protein induction response of hepatocytes with the response at the transcriptional level by measuring expression of

202

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203

vtg mRNA. Here we found, as expected, a concentration-related increase in vtg mRNA expression, consistent with the response measured for VTG protein synthesis. However, in the hepatocyte isolations analysed, measurement of vtg mRNA was relatively more sensitive, producing a value for the LOEC 10-fold lower (0.01 ␮M E2) than measuring VTG protein (0.1 ␮M E2). Furthermore, changes in vtg mRNA expression showed much less variation between individual hepatocyte isolations than measurement of the protein. Hepatic vtg mRNA induction in fish has been shown to be an effective signal of oestrogen exposure (Hemmer et al., 2002; Schmid et al., 2002). It is a very specific response and highly sensitive, particularly to steroidal oestrogens. Furthermore, mRNA responses can be detected after relatively short exposure periods, up to 24 h, providing potential as an early indicator of exposure. However, proteins may be arguably as important as RNA transcripts as biomarkers of biological function, when considering not all mRNA sequences are transcribed and many proteins undergo post-translational modification before becoming physiologically active. It may be possible to reduce the variability observed in VTG protein measurements in future studies by normalising data to amount of protein rather than supernatant volume, as described by Navas and Segner (2000). vtg mRNA expression has previously been measured via q-RT-PCR in rainbow trout hepatocytes, where vtg expression increased by over 2000-fold compared to controls following 24 h exposure to 10 nm ethinyloestradiol (EE2; Finne et al., 2007). 5. Conclusion Whilst factors such as sex and stage of sexual maturity of each fish may influence the oestrogenic response of hepatocytes, they do not account for all of the variation observed in the present study. If this system is to be used in the routine screening of chemicals for oestrogenic activity, as might it be applied within the context of REACH, it is important that the response of hepatocytes is both consistent and repeatable. The inherent variation seen between the cultures has a major bearing on the reproducibility of the VTG induction response. Despite this, we found that data from different cultures could be directly compared through appropriate standardisations. Once normalised, by expressing VTG induction as a percentage of maximum induction following exposure to 10 ␮M E2 for each culture, the hepatocyte VTG assay was more amenable to inter-assay comparisons. Overall, normalised concentration–response curves for the isolated hepatocytes following exposure to E2 were similar across independent exposure experiments. Using males, as opposed with females, as donor fish might also be recommended to help reduce inter-assay variation in the vitellogenic response for any oestrogen screening programme. Furthermore, analysis of the vitellogenic response at the level of the transcriptome appeared to offer a more sensitive and less variable method for the screening of chemicals with oestrogenic activity. If VTG induction by isolated hepatocytes is to be used for assessing the ability of other chemicals to stimulate an oestrogenic response, then including an E2 standard curve with every individual isolation experiment is essential to provide the required positive reference compound with which to assess the relative potencies of test chemicals. This study highlights some of the difficulties in the development of fish primary hepatocyte cultures for their potential use in regulatory screens for environmental oestrogens. It further identifies the need to understand the biological basis for the variability in the relative responses to oestrogen between different cell isolates. The work undertaken, however, also shows that by adopting a standardised approach for normalisation of the vitellogenic response, data between cultures for fish of different sexes, age and stage of maturity can be compared directly and with a high level of confidence. The primary hepatocyte culture technique, using either VTG

or vtg mRNA as an endpoint, offers a highly effective alternative for the screening of chemicals with oestrogenic activity. The advantages of this system over some of the other high throughput systems (e.g. ER reporter assays) include the fact that the hepatocytes retain their metabolic capability and so the ability to detect environmental oestrogens that require metabolic activation. Based on the requirements for technical expertise and cost, however, there is probably not much to choose between the primary hepatocyte system and some of the in vitro alternatives for screening environmental oestrogens. Acknowledgements LKB was funded through a British Biotechnology and Biosciences Research Council (BBSRC) Industrial Case-supported PhD studentship with AstraZeneca. AL was funded by the UK Natural Environment Research Council (Grant ref. NE/D002818/1 and within the Environmental Genomics Programme (NER/T/S/2002/00182)) to CRT. References Anderson, M.J., Miller, M.R., Hinton, D.E., 1996. In vitro modulation of 17beta-estradiol-induced vitellogenin synthesis: Effects of cytochrome P4501A1 inducing compounds on rainbow trout (Oncorhynchus mykiss) liver cells. Aquat. Toxicol. 34, 327–350. Animals (Scientific Procedures) Act, 1986. Legislation for the protection of animals used for experimental or other scientific purposes in the United Kingdom. www.archive.official-documents.co.uk/document/hoc/321/321-xa.htm. Date accessed 01/11/08. Ankley, G.T., Mihaich, E., Stahl, R., Tillitt, D., Colborn, T., McMaster, S., Miller, R., Bantle, J., Campbell, P., Denslow, N., Dickerson, R., Folmar, L., Fry, M., Giesy, J., Gray, L.E., Guiney, P., Hutchinson, T., Kennedy, S., Kramer, V., LeBlanc, G., Mayes, M., Nimrod, A., Patino, R., Peterson, R., Purdy, R., Ringer, R., Thomas, P., Touart, L., Van der Kraak, G., Zacharewski, T., 1998. Overview of a workshop on screening methods for detecting potential (anti-)estrogenic/androgenic chemicals in wildlife. Environ. Toxicol. Chem. 17, 68–87. Blair, J.B., Miller, M.R., Pack, D., Barnes, R., The, S.J., Hinton, D.E., 1990. Isolated trout liver cells—establishing short-term primary cultures exhibiting cell-to-cell interactions. In vitro Cell. Dev. Biol. 26, 237–249. Braunbeck, T., Storch, V., 1992. Senescence of hepatocytes isolated from rainbow trout (Oncorhynchus mykiss) in primary culture—an ultrastructural study. Protoplasma 170, 138–159. Burki, R., Vermeirssen, E.L.M., Korner, O., Joris, C., Burkhardt-Holm, P., Segner, H., 2006. Assessment of estrogenic exposure in brown trout (Salmo trutta) in a Swiss midland river: integrated analysis of passive samplers, wild and caged fish, and vitellogenin mRNA and protein. Environ. Toxicol. Chem. 25, 2077–2086. ˜ A., Bols, N., Braunbeck, T., Dierickx, P., Halder, M., Isomaa, B., Kawahara, Castano, K., Lee, L.E.J., Mothersill, C., Part, P., Repetto, G., Sintes, J.R., Rufli, H., Smith, R., Wood, C., Segner, H., 2003. The use of fish cells in ecotoxicology—the report and recommendations of ECVAM workshop 47. Altern. Lab. Anim. 31, 317–351. Crain, D.A., Guillette, L.J., 1997. Endocrine-disrupting contaminants and reproduction in vertebrate wildlife. Rev. Toxicol. 1, 47–70. Cravedi, J.P., Zalko, D., 2005. The potential use of rat and fish isolated hepatocytes in predicting the metabolic pathways of organic pollutants in vivo. Comp. Biochem. Physiol. A 126, 33–133. Fenske, M., van Aerle, R., Brack, S., Tyler, C.R., Segner, H., 2001. Development and validation of a homologous zebrafish (Danio rerio) vitellogenin enzyme-linked immunosorbent assay (ELISA) and its application for studies on estrogenic chemicals. Comp. Biochem. Physiol. C 129, 217–232. Filby, A.L., Tyler, C.R., 2005. Molecular characterization of estrogen receptors 1, 2a, and 2b and their tissue and ontogenic expression profiles in fathead minnow (Pimephales promelas). Biol. Reprod. 73, 648–662. Finne, E.F., Cooper, G.A., Koop, B.F., Hylland, K., Tollefsen, K.E., 2007. Toxicogenomic responses in rainbow trout (Oncorhynchus mykiss) hepatocytes exposed to model chemicals and a synthetic mixture. Aquat. Toxicol. 81, 293–303. Flouriot, G., Vaillant, C., Salbert, G., Pelissero, C., Guiraud, J.M., Valotaire, Y., 1993. Monolayer and aggregate cultures of rainbow trout hepatocytes—long-term and stable liver-specific expression in aggregates. J. Cell. Sci. 105, 407–416. Gagné, F., Pardos, M., Blaise, C., 1999. Estrogenic effects of organic environmental extracts with the trout hepatocyte vitellogenin assay. Bull. Environ. Contam. Toxicol. 62, 723–730. Gimeno, S., Komen, H., Jobling, S., Sumpter, J., Bowmer, T., 1998. Demasculinisation of sexually mature common carp, Cyprinus carpio, exposed to 4-tert-pentylphenol during spermatogenesis. Aquat. Toxicol. 43, 93–109. Halm, S., Martinez-Rodriguez, G., Rodriguez, L., Prat, F., Mylonas, C.C., Carrillo, M., Zanuy, S., 2004. Cloning, characterization, and expression of three oestrogen receptors (ER␣, ER␤1 and ER␤2) in the European sea bass, Dicentrarchus labrax. Mol. Cell. Endocrinol. 223, 63–75.

L.K. Bickley et al. / Aquatic Toxicology 94 (2009) 195–203 Hemmer, M.J., Bowman, C.J., Hemmer, B.L., Friedman, S.D., Marcovich, D., Kroll, K.J., Denslow, N.D., 2002. Vitellogenin mRNA regulation and plasma clearance in male sheepshead minnows, (Cyprinodon variegatus) after cessation of exposure to 17b-estradiol and p-nonylphenol. Aquat. Toxicol. 58, 99–112. ICCVAM: Interagency Coordinating Committee on the Validation of Alternative Methods, 2002. Expert panel evaluation of the validation status of in vitro test methods for detecting endocrine disruptors. http://iccvam.niehs.nih.gov. Date accessed 01/11/08. Islinger, M., Pawlowski, S., Hollert, H., Volkl, A., Braunbeck, T., 1999. Measurement of vitellogenin mRNA expression in primary cultures of rainbow trout hepatocytes in a non-radioactive dot blot/RNAse protection-assay. Sci. Total Environ. 233, 109–122. Jobling, S., Sumpter, J.P., 1993. Detergent components in sewage effluent are weakly oestrogenic to fish: an in vitro study using rainbow trout (Oncorhynchus mykiss) hepatocytes. Aquat. Toxicol. 27, 361–372. Katsu, Y., Lange, A., Ichikawa, R., Urushitani, H., Paull, G.C., Cahill, L.L., Jobling, S., Tyler, C.R., Iguchi, T., 2007. Functional associations between two estrogen receptors, environmental estrogens and sexual disruption in the roach (Rutilus rutilus). Environ. Sci. Technol. 41, 3368–3374. Kim, B.H., Takemura, A., 2003. Culture conditions affect induction of vitellogenin synthesis by estradiol-17 beta in primary cultures of tilapia hepatocytes. Comp. Biochem. Physiol. B 135, 231–239. Klaunig, J.E., Ruch, R.J., Goldblatt, P.J., 1985. Trout hepatocyte culture—isolation and primary culture. In vitro Cell. Dev. Biol. 21, 221–228. Kwon, H.C., Hayashi, S., Mugiya, Y., 1993. Vitellogenin induction by estradiol-17-beta in primary hepatocyte culture in the rainbow trout, Oncorhynchus mykiss. Comp. Biochem. Physiol. B 104, 381–386. Latonnelle, K., Le Menn, F., Bennetau-Pelissero, C., 2000. In vitro estrogenic effects of phytoestrogens in rainbow trout and Siberian sturgeon. Ecotoxicology 9, 115–125. Mommsen, T.P., Walsh, P.J., 1988. Vitellogenesis and oocyte assembly. In: Hoar, W.S., Randall, D.J. (Eds.), Fish Physiology XIA. Academic Press, San Diego. Navas, J.M., Segner, H., 2006. Vitellogenin synthesis in primary cultures of fish liver cells as an endpoint for in vitro screening of the (anti) estrogenic activity of chemical substances. Aquat. Toxicol. 80, 1–22. Navas, J.M., Segner, H., 2000. Antiestrogenicity of beta-naphthoflavone and PAHs in cultured rainbow trout hepatocytes: evidence for a role of the arylhydrocarbon receptor. Aquat. Toxicol. 51, 79–92. Nilsen, B.M., Berg, K., Eidem, J.K., Kristiansen, S.I., Brion, F., Porcher, J.M., Goksoyr, A., 2004. Development of quantitative vitellogenin ELISAs for fish test species used in endocrine disruptor screening. Anal. Bioanal. Chem. 378, 621–633. Okoumassoun, L.E., Averill-Bates, D., Gagné, F., Marion, M., Denizeau, F., 2002. Assessing the estrogenic potential of organochlorine pesticides in primary cultures of male rainbow trout (Oncorhynchus mykiss) hepatocytes using vitellogenin as a biomarker. Toxicology 178, 193–207. Pakdel, F., Delauney, F., Ducouret, B., Flouriot, G., Kern, L., Lazennec, G., Le Drean, Y., Petit, F., Salbert, G., Saligaut, D., Tujague, M., Valotaire, Y., 1997. Regulation of gene expression and biological activity of rainbow trout estrogen receptor. Fish Physiol. Biochem. 17, 123–133. Pelissero, C., Flouriot, G., Foucher, J.L., Bennetau, B., Dunogues, J., Legac, F., Sumpter, J.P., 1993. Vitellogenin synthesis in cultured hepatocytes—an in vitro test for the estrogenic potency of chemicals. J. Steroid Biochem. Mol. Biol. 44, 263–272. Pedersen, R.T., Hill, E.N., 2000. Biotransformation of the xenoestrogen 4-tertoctylphenol in hepatocytes of rainbow trout (Oncorhynchus mykiss). Xenobiotica 30, 867–879. Purdom, C.E., Hardiman, P.A., Bye, V.J., Eno, N.C., Tyler, C.R., Sumpter, J.P., 1994. Estrogenic effects of effluents from sewage treatment works. Chem. Ecol. 8, 275–285. Rankouhi, T.R., van Holsteijn, I., Letcher, R., Giesy, J.P., van den Berg, M., 2002. Effects of primary exposure to environmental and natural estrogens on vitellogenin production in carp (Cyprinus carpio) hepatocytes. Toxicol. Sci. 67, 75–80. Rankouhi, T.R., Sanderson, J.T., van Holsteijn, I., van Leeuwen, C., Vethaak, A.D., van den Berg, M., 2004. Effects of natural and synthetic estrogens and various environmental contaminants on vitellogenesis in fish primary hepatocytes: comparison of bream (Abramis brama) and carp (Cyprinus carpio). Toxicol. Sci. 81, 90–102.

203

Riley, L.G., Hirano, T., Grau, E.G., 2004. Estradiol-17 beta and dihydrotestosterone differentially regulate vitellogenin and insulin-like growth factor-I production in primary hepatocytes of the tilapia Oreochromis mossambicus. Comp. Biochem. Physiol. C 138, 177–186. Routledge, E.J., Sheahan, D., Desbrow, C., Brighty, G.C., Waldock, M., Sumpter, J.P., 1998. Identification of estrogenic chemicals in STW effluent. 2: In vivo responses in trout and roach. Environ. Sci. Technol. 32, 1559–1565. Schirmer, K., 2006. Proposal to improve vertebrate cell cultures to establish them as substitutes for the regulatory testing of chemicals and effluents using fish. Toxicology 224, 163–183. Schmid, T., Gonzalez-Valero, J., Rufli, H., Dietrich, D.R., 2002. Determination of vitellogenin kinetics in male fathead minnows (Pimephales promelas). Toxicol. Lett. 131, 65–74. Segner, H., 1998. Isolation and primary culture of teleost hepatocytes. Comp. Biochem. Physiol. A 120, 71–81. Segner, H., Blair, J.B., Wirtz, G., Miller, M.R., 1994. Cultured trout liver cells—utilization of substrates and response to hormones. In vitro Cell. Dev. Biol. 30A, 306–311. Segner, H., Bohm, R., Kloas, W., 1993. Binding and bioactivity of insulin in primary cultures of carp (Cyprinus carpio) hepatocytes. Fish Physiol. Biochem. 11, 411–420. Segner, H., Cravedi, J.P., 2001. Metabolic activity in primary cultures of fish hepatocytes. Altern. Lab. Anim. 29, 251–257. Silversand, C., Hyllner, S.J., Haux, C., 1993. Isolation, immunochemical detection, and observations of the instability of vitellogenin from 4 teleosts. J. Exp. Zool. 267, 587–597. Smeets, J.M.W., Rankouhi, T.R., Nichols, K.M., Komen, H., Kaminski, N.E., Giesy, J.P., van den Berg, M., 1999a. In vitro vitellogenin production by carp (Cyprinus carpio) hepatocytes as a screening method for determining (anti)estrogenic activity of xenobiotics. Toxicol. Appl. Pharmacol. 157, 68–76. Smeets, J.M.W., van Holsteijn, I., Giesy, J.P., Seinen, W., van den Berg, M., 1999b. Estrogenic potencies of several environmental pollutants, as determined by vitellogenin induction in a carp hepatocyte assay. Toxicol. Sci. 50, 206–213. Strategy for a Future Chemicals Policy, 2001. Commission of the European Communities. Brussels, 27.2.2001 COM(2001) 88 Final http://eurlex.europa.eu/ LexUriServ/site/encom/2001/com2001 0088en01.pdf. Date accessed 01/11/08. Thorpe, K.L., Cummings, R.I., Hutchinson, T.H., Scholze, M., Brighty, G., Sumpter, J.P., Tyler, C.R., 2003. Relative potencies and combination effects of steroidal estrogens in fish. Environ. Sci. Technol. 37, 1142–1149. Tollefsen, K.E., Mathisen, R., Steenersen, J., 2003. Induction of vitellogenin synthesis in an Atlantic salmon (Salmo salar) hepatocyte culture: a sensitive in vitro bioassay for the oestrogenic and anti-oestrogenic activity of chemicals. Biomarkers 8, 394–407. Tremblay, L., Van der Kraak, G., 1998. Use of a series of homologous in vitro and in vivo assays to evaluate the endocrine modulating actions of ␤-sitosterol in rainbow trout. Aquat. Toxicol. 43, 149–162. Tyler, C.R., Filby, A.L., Bickley, L.K., Cumming, R.I., Gibson, R., Labadie, P., Katsu, Y., Liney, K.E., Shears, J.A., Silva-Castro, V., Urushitani, H., Lange, A., Winter, M.J., Iguchi, T., Hill, E.M., 2009. Environmental health impacts of equine estrogens derived from hormone replacement therapy. Environ. Sci. Technol. 43, 3897–3904. Tyler, C.R., Jobling, S., Sumpter, J.P., 1998. Endocrine disruption in wildlife: a critical review of the evidence. Crit. Rev. Toxicol. 28, 319–361. Tyler, C.R., van Aerle, R., Hutchinson, T.H., Maddix, S., Trip, H., 1999. An in vivo testing system for endocrine disruptors in fish early life stages using induction of vitellogenin. Environ. Toxicol. Chem. 18, 337–347. Tyler, C.R., van der Eerden, B., Jobling, S., Panter, G., Sumpter, J.P., 1996. Measurement of vitellogenin, a biomarker for exposure to oestrogenic chemicals, in a wide variety of cyprinid fish. J. Comp. Physiol. [B] 166, 418–426. Vaillant, C., Leguellec, C., Pakdel, F., Valotaire, Y., 1988. Vitellogenin gene expression in primary culture of male rainbow trout hepatocytes. Gen. Comp. Endocrinol. 70, 284–290. Vos, J.G., Dybing, E., Greim, H.A., Ladefoged, O., Lambre, C., Tarazona, J.V., Brandt, I., Vethaak, A.D., 2000. Health effects of endocrine-disrupting chemicals on wildlife, with special reference to the European situation. Crit. Rev. Toxicol. 30, 71–133.