Reporter cell lines to study the estrogenic effects of xenoestrogens

The Science of the Total Environment 233 Ž1999. 47]56

Reporter cell lines to study the estrogenic effects of xenoestrogens Patrick Balaguer a,1, Fabienne Franc¸ois b, Franck Comunale a , Helene ´ ` Fenet c , Anne-Marie Boussioux a , Michel Pons a , Jean-Claude Nicolas a , Claude Casellas b,U b

a INSERM 439, 70 rue de Na¨ acelles, 34090 Montpellier, France Departement Sciences de l’En ¨ ironnement et Sante´ Publique. UMR ] CNRS 5556-Faculte´ de Pharmacie A¨ enue Charles ´ Flahaut. 34060 Montpellier, France c Laboratoire de chimie analytique,Faculte ´ de Pharmacie A¨ enue Charles Flahaut. 34060 Montpellier, France

Abstract In order to characterize the estrogenic activity of chemicals, we established complementary in vitro recombinant receptor]reporter gene assays in stably transfected MCF-7 and HeLa cells. MCF-7 cells which express the endogenous estrogen receptor a ŽERa . were stably transfected with only an estrogen-regulated luciferase gene. These cells enable the detection of compounds which bind to ERa or interfere with the induction of ERa mediated gene expression. Furthermore, HeLa cells, which do not express endogenous ERs, were transfected with an ERa or an ERb construct together with an estrogen-regulated luciferase gene, or a chimeric GAL4-ERa receptor and the corresponding luciferase reporter gene. Finally, we tested these four cellular models as tools to check the estrogenic activities of several potential xenoestrogens and to detect estrogenic activity in wastewater sewage treatment effluents. In all of the models, nonylphenol mixture ŽNPm., 4 n-nonylphenol Ž4 nNP., 2,49-DDE, 4,49-DDE and wastewater sewage treatment effluent were active, while PCB mixture ŽAroclor 1254., PCB 77, atrazine and lindane Ž g hexachlorocyclohexane . were inactive. Dioxin partially activates the estrogen receptor in MCF-7 cells while in HeLa-derived cell lines, it decreased the estrogenic-induced expression of luciferase. Q 1999 Elsevier Science B.V. All rights reserved. Keywords: Xenoestrogens; Reporter cell lines; ERa; ERb; Wastewater effluent

U

Corresponding author. Tel.: q33-4-67-54-80-64; fax: q33-4-67-04-59-33. E-mail address: [email protected] ŽC. Casellas. 1 Information on cell lines can be requested from P. Balaguer INSERM 439, 70 rue de Navacelles, 34090 Montpellier, France. Tel.: 33-467043703; fax: 33-467043715. 0048-9697r99r$ - see front matter Q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 8 - 9 6 9 7 Ž 9 9 . 0 0 1 7 8 - 3

48

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

1. Introduction Environmental estrogens can be defined as any compounds that can bind the estrogen receptors ŽERa or b . and elicit or modulate an ER-mediated response. In many cases, these chemicals share no structural similarities with traditional ER ligands, but have been classified as environmental estrogens on the basis of their ability to induce physiological effects that resemble estrogenic responses ŽGillesby and Zacharewski, 1998.. Sources of suspected environmental estrogens include endogenous plant compounds and several xenobiotics ŽBalaguer et al., 1996; Connor et al., 1997; Moore et al., 1997; Le Bail et al., 1998., such as pesticides, estrogenic drugs, degradation products of alkylphenolethoxylate surfactants, and some precursors used in the manufacture of plastics. Several epidemiological studies have reported an association between exposure or body burden of suspected environmental estrogens and the incidence of some hormone-dependant diseases. For example, exposure to environmental estrogens has been associated with an increase in incidence of testicular and breast cancer, disorders of the male reproductive tract, and compromised reproductive capacity. In contrast, Adlercreutz Ž1995. noted that phytoestrogens may not contribute to the adverse effect seen in humans or wildlife but could in fact have a role in protection against cancer; nevertheless there is no evidence that man-made xenoestrogens have similar effects. Several in vivo and in vitro assays have been developed to assess the estrogenic-like activity of individual compounds or natural mixtures Žreviewed by Zacharewski, 1997.. Most of these assays use a variety of endpoints, including an increase in target organ weight, increased protein expression or cell proliferation. They often are time consuming and expensive, far from the criteria required for large-scale screenings and environmental monitoring. Moreover, these assays may not demonstrate unequivocally that the observed responses are receptor-mediated, and a substance may be misclassified as an environmental estrogen on the basis of studies using these endpoints. The development of in vitro assays such as recombinant reporter induction and re-

ceptor-binding provide simple endpoints, which together demonstrate that the observed effect is receptor-mediated. Therefore, as a first large screening-tool, these assays will allow to identify and assess the activity of substances that are suspected of adversely affecting human health and wildlife ŽDesbrow et al., 1998.. The objective of this study was to evaluate the use of reporter cell lines with environmental samples. Pure chemicals were selected for their known or suspected estrogenic activity. They were chosen for their occurrence or persistence in the aquatic environment due to urban Že.g. surfactants. or agriculture Že.g. pesticides, DDE. activities. Sewage treatment work effluents were also studied.

2. Materials and methods Materials for cell culture were obtained from Life Technologies ŽCergy Pontoise, France.. Luciferin was synthesized by G. Auzou according to Bowie Ž1978.. Estradiol ŽE2. was purchased from Sigma Chemical Co ŽSt Louis, MO, USA.. 4-Hydroxytamoxifen ŽOH, Tam. and ICI 164.384 was obtained from Zeneca ŽMacclesfield, UK .. TTNPB was a gift from Hoffman-La-Roche ŽBasel, Switzerland. while dexamethasome ŽDEX. was obtained from Hoechst Marion Roussel ŽRomainville, France.. These effectors were dissolved in ethanol at 10y3 M. Nonylphenol mixture ŽNpm, ring and chain isomers, Sigma-Aldrich, purity 90%., 4 n-nonylphenol Ž4 nNP., 2,49-DDE, 4,49-DDE, ŽSigmaAldrich, purity ) 99%. were diluted to 10y2 M in methanol. A PCB mixture ŽAroclor 1254., 3,39 4,49-tetrachlorobiphenyl ŽPCB 77., atrazine, lindane Ž g hexachlorocyclohexane . and 2,3,7,8-tetrachlorodibenzo-p-dioxin ŽCIL Cluzeau, purity ) 98%. were prepared in dimethyl sulfoxide ŽDMSO. at 10y2 M. Effluent and influent were collected from a sewage treatment works. It receives domestic wastewater Ž3 = 10 5 equivalent inhabitant . and the treatment process used is aerated activated sludge medium loaded.

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

After sterile filtration on 0.2-mm membranes, they were stored at Žq48C.. All the equipment was rinsed with methanol prior to use to remove any organic compounds. ERE-bGlob-Luc-SVNeo was obtained by inserting in the Apa I site of ERE-bGlob-Luc ŽMetzger et al., 1995. the Apa I fragment containing the neomycin gene of pMAMneoLUC ŽCLONTECH, Palo Alto, CA, USA.. RARE-tkLuc-SVNeo plasmid was obtained by exchanging the fragment ERE-bGlob ŽSal I-Bgl II. from ERE-bGlob-Luc by a fragment RARE-tk ŽSal I-Bgl II. providing from RARE-tk-CAT ŽMader et al., 1993. and by adding in the ApaI site the fragment containing the neomycin gene. pSG5ERa puro was obtained by exchanging the fragment ŽNdeI-BamHI. from pSG5puro vector with the NdeI]BamHI fragment containing the sequence of ERa from pSG5-ERa ŽGreen et al., 1988.. RARE-tk-CAT, ERE-bGlob-Luc, 17m5bGlob-Luc, pAG-60, pSG5puro, pSG5-ERa and pSG5-GAL4-ERa were gifts from Hinrich Gronemeyer and Pierre Chambon. pSG5ERbpuro was a gift from Michelle Resche-Rigon. Standard protocols of recombinant DNA technology were used for plasmid preparations. MCF7 and HeLa-derived cell lines were cultured in DMEM with phenol red supplemented with 5% of fetal calf serum ŽFCS. for the strain culture. For experiments, in order to decrease their estradiol content and their background signal, cells were pretreated for 3 days with steroidfree calf serum. This was achieved by seeding cells in 24-well plates Žapprox. 5 = 10 4 HeLa or 2 = 10 5 MCF7-derived cells per well. in the presence of DMEM whithout phenol red, supplemented by 3% of dextran-coated charcoal-treated fetal calf serum ŽDCC-FCS.. Cells were then incubated for 16 h with the test compounds. Compound solvents never exceed 0.1% of the culture medium. For wastewater samples 1 vol. of water Ž250 ml. was mixed with 1 vol. of medium twofold concentrated, prior to adding to the wells which already contained 1 vol. of medium; thus a wastewater sample was threefold diluted in the culture medium Ž1:3.. At the end of incubation, cells were treated as described below for luciferase activity determination and protein assay.

49

The stably transfected cell lines were obtained as already described ŽGagne et al., 1994; Joyeux et al., 1997.. Briefly, to obtain MELN and HELN cells, we transfected MCF7, and HeLa cells respectively with the ERE-bGlob-Luc-SVNeo plasmid. HELN cells with ERa and ERb were obtained by a second transfection of the corresponding pSG5puro plasmids ŽpSG5ERa puro or pSG5ERbpuro, respectively.. HGELN cells were prepared by cotransfecting HeLa cells with pAG60, the plasmid responsible for conferring resistance to neomycin Žor G418., p17m5-bGlob-Luc and the chimeric receptor expression plasmid GAL4-ER ŽChen et al., 1995.. HeLa cells, stably transfected with the glucocorticoid responsive reporter gene MMTV-Luc-SVNeo ŽHMLN. or with the retinoid responsive gene RARE-tk-LucSVNeo ŽHRLN., were used in this study as negative controls. Selection by geneticin and puromycin was made at 0.5 mgrml and 0.5 mgrml, respectively. Luminescent and inducible clones were identified using photon-counting cameras ŽArgus 50 from Hamamatsu or Night Owl from Berthold. and the most responsive clones were isolated. Cell extract preparation was done essentially as recommended by Promega Corporation. In brief, cells were washed twice with 1 ml of PBS, lysed with 0.4 ml of lysis buffer Ž25 mM Tris phosphate, pH 7.8, 2 mM EDTA, 10% glycerol, 1% triton X-100.. After shaking plates for 10 min, lysed cells were collected and briefly centrifuged. Luciferase activity was measured on 100 ml supernatant aliquots, during 10 s after injection of 100 ml of luciferase detection solution w20 mM Tricine, pH 7.8, 1.07 mM ŽMgCO3 . 4Mg ŽOH. 2 5 H 2 O, 2.67 mM MgSO4 , 0.2 mM EDTA, ATP 0.53 mM, Co-enzyme A Ž0.27 mM., Luciferin Ž0.48 mM.x using a LKB-Wallac luminometer. The results were expressed as arbitrary units per 100 mg of protein as assayed by Lowry’s method ŽLowry et al., 1951..

3. Results In order to characterize the Žanti.estrogenic activity of chemicals, we established complemen-

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

50

tary in vitro recombinant receptor]reporter gene assays in stably transfected MCF7 and HeLa cells. MELN cells were derived from MCF7 cells. In these cells, that express endogenous ERa , we only stably transfected a luciferase gene driven by an estrogen responsive element ŽERE. in front of the bGlobin promoter ŽERE-bGlob-Luc.. MELN cells enable the detection of compounds which bind ERa or interfere with the induction of ERa-mediated gene expression. Other compounds which bind to other receptors ŽDemirpence et al., 1994; Kharat and Saatcioglu, 1996. could also act on the ER pathway by DNA binding or transcriptional interference or by modification of the expression of ERa in MCF7 cells. Since a new estrogen receptor ERb has been recently discovered ŽKuiper et al., 1997., we established two other reporter cell lines, with either an ERa or ERb specificity. In order to obtain comparable cell lines, the estrogen-responsive reporter gene was firstly stably transfected into Hela cells that do not express ERs. In these cells ŽHELN cell line., estrogens do not induce any luciferase gene expression Žresults not shown.. In a second step, these HELN cells were transfected with an ERa or ERb construct to obtain the HELNŽ a . or HELNŽb . cell lines, respectively. Since in these cells the exogenous ER genes are under control of a SV40 promoter which cannot a priori be regulated, Žanti.estrogenic activity of xenoestrogens could not be due to a modulated

expression of estrogen receptors. Another cell line ŽHEGLN. expressing a chimeric estrogen-receptor was also established. The receptor protein contains the DNA binding domain of the yeast transactivator GAL4 fused to the ligand-binding domain ŽEF regions. of ERa. The HEGLN cell line was obtained by transfecting both the GAL4ERa expression plasmid and a luciferase gene driven by a pentamer of the Gal4 recognition sequence Ž17m. in front of the bGlobin promoter Ž17Mx5-bGlob-Luc.. In this cell line, the luciferase expression is restricted to test compounds which directly interact with the ligand binding site. Thus, test compounds acting through the N-terminal domain of ERa cannot trigger the luciferase expression. Interference effect mediated by ligand-activated receptors, such as competition for the estrogen-responsive element ŽERE., cannot occur in these cells ŽDemirpence et al., 1994.. However, cross-talk involving coactivators interfering with the ER signaling pathway, through interactions with the ligand binding site, could be identified by the HGELN cell line. Fig. 1 summarizes the effects of 17b-estradiol and several xenoestrogens on ER-mediated luciferase induction in the MELN cell line. The EC 50 value for the 17b-estradiol-induced response was approximately 5 pM. After a 3 day culture in DMEM DCC-FCS, MELN cells were not absolutely free of intracellular estradiol since antiestrogens were able to decrease the ETOH

Table 1 Effect of the NP mixture and dioxin on the four estrogenic responsive cell lines a Sample

Control OHT 10y7 M ICI 10y7 M Npm 10y7 M Dioxin 10y7 M ICI 10y7 M, dioxin 10y7 M E2 10y1 1 M E2 10y1 1 M, ICI 10y7 M E2 10y1 1 M, Npm 10y7 M E2 10y1 1 M, dioxin 10y7 M a

% Luciferase activity Ž100% as 10y8 M E2. MELN

HELNa

10.5" 1 4.3" 0.5 2.5" 0.5 75.5" 2.5 28.8" 1.5 2.3" 0.4 92.5" 3 2.8" 0.5 93.4" 1 117.5" 1

20 " 2 24.5" 3 9.1" 1 65.5" 3.5 11.5" 2.3 9.3" 1.2 89.7 " 1 9.3" 1.5 91.6" 1 71.5" 1

Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M.

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

51

Fig. 1. Induction of luciferase activity by xenoestrogens in the MELN cell line. Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M Ž9500 arbitrary units r0.1 mg of protein..

Fig. 2. Induction of luciferase activity by xenoestrogens in the HELNa cell line. Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M Ž3100 arbitrary unitsr0.1 mg of protein..

52

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

value ŽTable 1.. This is probably due to residual estrogens in the serum and the potency of MELN to answer to low estradiol levels. All of the xenoestrogens elicited a dose-dependent induction. The ranking of the estrogenic potency is E2 4 nonylphenol mixture ) mixture ) 2,49-DDE) 4 nnonylphenol ) 4,49-DDE) dioxin. However, maximum induction was only achieved by the nonylphenol mixture at 3 = 10y7 M. 4 n-Nonylphenol, 2,49-DDE, 4,49-DDE and dioxin, at their highest concentration, only partially activated the reporter gene. However, nonylphenol mixture, 4 n-nonylphenol, 2,49-DDE and 4,49-DDE could be considered as full agonists since they were not able to decrease activation by estradiol ŽTable 1 and data not shown.. Surprisingly dioxin increased both basal and estradiol-induced luciferase expression ŽFig. 1 and Table 1.. Estrogenic activity of these compounds was completely inhibited with the pure antiestrogen ICI 164,384 at 10y7 M ŽTable 1 and results not shown.. Lindane, atrazine, PCB77 and Aroclor 1254 were also tested for their estrogenic activities and had

no significant effect in this cellular model Žresults not shown.. Fig. 2 summarizes the effects of 17b-estradiol and xenoestrogens on luciferase induction in the HeLa-derived cell line, HELNŽ a ., expressing ERa. Results obtained in HELNŽ a .cells were similar to that previouly described in MELN cells, particularly the estradiol sensitivity was approximately 5 pM. Only dioxin showing inhibitory effects on both basal and estradiol-induced luciferase expression ŽFig. 2 and Table 1., presented a different behavior. Such an antiestrogenic effect of dioxin was already observed and attributed to direct transcriptional interference between ARh and ER ŽKharat and Saatcioglu, 1996.. Fig. 3 summarizes the effects of 17b-estradiol and xenoestrogens on ER-mediated luciferase induction in the HELN b cell line. The EC 50 value Žapprox. 10 pM. for the 17b-estradiol induced response was slightly higher than that obtained in MELN and HELNŽ a . cells. Nonylphenol mixture, 2,49-DDE, 4 n-nonylphenol and 4,49-DDE

Fig. 3. Induction of luciferase activity by xenoestrogens in the HELNb cell line. Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M Ž2600 arbitrary unitr0.1 mg of protein..

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

53

Fig. 4. Induction of luciferase activity by xenoestrogens in the HGELN cell line. Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M Ž1100 arbitrary unitsr0.1 mg of protein..

also elicited a dose-dependent induction but in this model, 4 n-nonylphenol transactivate better than 2,49-DDE. Dioxin has an inhibitory effect while lindane, atrazine, PCB77 and Aroclor 1254 have also been tested and were not found to have a significant effect Žresults not shown.. Fig. 4 summarizes the effects of 17b-estradiol and xenoestrogens on ER-mediated luciferase in-

duction in the HGELN cell line. The EC 50 values for the 17b-estradiol-induced response was 40 pM for the HGELN cell line. The nonylphenol mixture, 2,49-DDE, 4 n-nonylphenol and 4,49-DDE also elicited a dose-dependent induction. In order to check the estrogenic specificity of these test-compounds, we took advantage of possessing the two HeLa-derived cell lines HMLN

Table 2 Absence of effect of the NP mixture and dioxin on the glucocorticoid and retinoid responsive cell lines a Sample

% Luciferase activity HMLN

Control Npm 10y7 M Dioxin 10y7 M DEX 10y7 M DEX 10y7 M, Npm 10y7 M DEX 10y7 M, dioxin 10y7 M a

2.6" 1.1 2.4" 2.5 2.3" 1.5 100 " 3 98.5" 1.2 97.8" 1.5

HRLN Control Npm 10y7 M Dioxin 10y7 M TTNPB 10y7 M TTNPB 10y7 M, Npm 10y7 M TTNPB 10y7 M, dioxin 10y7 M

15 " 2 14.7" 2 15.2" 2 100 " 3 98.5" 1.2 97.8" 1.5

Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of 10y7 M DEX for HMLN and 10y7 M TTNPB for HRLN cell lines.

54

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

and HRLN cells, to assess their potential glucocorticoid or retinoid activities. As expected for xenoestrogen compounds, no stimulating nor inhibiting effects on luciferase induction in these glucocorticoid or retinoid responsive cell lines were observed ŽTable 2 and data not shown.. Table 3 shows the ability of these assays for testing estrogenic activity in environmental samples. Sewage treatment works influent was tested using the four estrogenic cell lines. Water samples showed potent estrogenic activity but did not increase luciferase expression in the glucocorticoid and retinoid responsive cell lines. The antiestrogen ICI completely blocked this activity. Compared to the estradiol dose-response curves described in Figs. 1]4 and taking into account the water dilution, these crude results suggested that the estrogenic activity were equivalent to 60]100 pM estradiol in the influent and was reduced by a factor of 4]10 in the effluent. These results mainly show that cell lines gave similar results.

4. Discussion In vitro recombinant receptor]reporter gene assays provide methods that are sensitive, selective, well adapted to high throughput screening assays and environmental biomonitoring. The stable transfectants used in this study exhibited good responsiveness following treatment to 17bestradiol Ždetection limit 1]10 pM.. The most sensitive were MELN and HELNŽ a . cell lines with an EC 50 of approximately 5 pM. The HELNŽb . cells are slightly less sensitive ŽEC 50 10

pM., probably due to the lower affinity of E2 for ERb than for ERa ŽKuiper et al., 1998.. The HGELN cell line which express the chimeric GAL4-ER receptor was the less sensitive ŽEC 50 40 pm.. Significant induction in reporter gene activity was observed following treatment with nonylphenol mixture, 4 n-nonylphenol, 2,49-DDE, 4,49-DDE in the four models indicating that they were capable of binding to } and activating } the estrogen receptors. Soto et al. Ž1995. showed with the MCF-7 proliferation assay an estrogenic activity of 4-nonylphenol and the mixture of nonylphenol. Maximal induction of luciferase was only obtained with the mixture of nonylphenols indicating that other nonylphenol compounds were more active than 4 n-nonylphenol. This could be due to different possible branching forms, nonylphenol contains isomeric branching structures that have different activities ŽRoutledge and Sumpter, 1997.. Dioxin only induced luciferase expression in the MCF-7-derived cell line. Moreover, in the HeLa-derived cell lines, dioxin decreased the basal and the estradiol-induced luciferase activity. These results indicate that dioxin does not interact directly with ER suggesting that dioxin could bind the aryl-receptor which in turn interfere with the activated estrogen receptor positively in MCF7 cells and negatively in HeLa cells. The results derived from the wastewater samples clearly indicate that they contain estrogenic compounds. This was observed by many authors ŽSumpter and Jobling, 1995; Desbrow et al., 1998.. Future work will consist of identifing these com-

Table 3 Estrogenic activity of water samples in the four different cell lines a Sample

Control Influent Ž1r3. Influent Ž1r3., ICI 10y7 M Effluent Ž1r3.

% Luciferase activity MELN

HELNa

HELNb

HGELN

HMLN

HRLN

10.5" 1.1 92.5" 2.7 2.5" 0.5 35.5" 0.7

20.3" 1.2 89 " 1.6 9.1" 1.5 41.5" 0.9

23.5" 1 69.4" 2.2 20.5" 1.2 34.3" 3.2

12.5" 0.5 57.3" 1.3 7.6" 0.8 29.3" 13

2.6" 1.1 2.4" 0.6 2.1" 0.5 2.6" 0.8

15 " 1.2 14.7" 0.8 15.3" 0.6 13.5" 1.3

a Results are expressed as a percentage of luciferase activity measured per well Žmean " S.E.M. of triplicates .. The 100% value represents the value obtained in presence of E2 10y8 M for MELN, HELNa , HELNb and HGELN, 10y7 M DEX for HMLN and 10y7 M TTNPB for HRLN cell lines

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

pounds. We have to determine whether the estrogenic activity is due to xenoestrogens or natural estrogens. Effluents from sewage treatment works are an important potential source of estrogenic compounds contaminating the aquatic environment. Another important point will be to measure estrogenic activity in the river downstream of the treatment plant. This study demonstrates the utility of in vitro recombinant receptor]reporter gene bioassays for identifying chemicals and environmental samples with estrogenic activities. These assays with simple end-points based on receptor-mediated observed effect, could be proposed as a first step in an assessment process. They must be complemented by in vitro assays taking into account the complexity of routes of exposure for assessing adverse effects on organisms. Development of stably transfected cell lines eliminates the requirement for transient transfection, thus significantly improving the utility of the assay as a screen to identify estrogenic substances. As antiestrogenic effects could be induced by a variety of mechanisms, the development of different estrogenic responsive cell lines could show differences of activation with xenoestrogens.

Acknowledgements This work was supported by funds from the EC ŽBMH4]96]0181. and the Agence de l’eau Seine-Normandie. The authors are grateful to Dr H. Gronemeyer, Prof P. Chambon and Dr M. Resche-Rigon for plasmids and to Dr T. Zacharewski for helpful discussion. References Adlercreutz H. Phytoestrogens: Epidemiology and a possible role in cancer protection. Environ Health Perspect 1995;103ŽSuppl 7.:103]112. Balaguer P, Joyeux A, Denison M, Vincent R, Gillesby B, Zacharewski T. Assessing the estrogenic and dioxin like activities of chemicals and complex mixtures using in vitro recombinant receptorrreporter gene assays. Can J Physiol Pharmacol 1996;94:216]222. Bowie LJ. Methods in enzymology. Synthesis of firefly luciferin and structural analogs. New York LVII: Academic Press, 1978:15]28.

55

Chen J-Y, Penco S, Ostrowski J, Balaguer P, Pons M, Starret J, Reczek P, Chambon P, Gronemeyer H. RAR-specific agonistrantagonists which dissociate transactivation and AP1 transrepression inhibit anchorage-independent cell proliferation. EMBO J 1995;14:1187]1197. Connor K, Ramamoorthy K, Moore M, Mustain M, Chen I, Safe S, Zacharewski T, Gillesby B, Joyeux A, Balaguer P. Hydroxylated polychlorinated biphenyl ŽPCBs. as estrogens and antiestrogens: structure ]activity relatioships. Toxicol Appl Pharmacol 1997;145:111]123. Demirpence E, Balaguer P, Trousse F, Pons M, Gagne D. Antiestrogenic effects of all-trans retinoic acid and 1,25 dihydroxyvitamin D3 in breast cancer cells occur at the estrogen response element but through different molecular mechanisms. Cancer Res 1994;54:1458]1464. Desbrow C, Routledge EJ, Brighty GC, Sumpter JP, Waldock M. Identification of estrogenic chemicals in STW effluent. 1. Chemical fractionation and in vitro biological screening. Environ Sci Technol 1998;32:1549]2158. Gagne D, Balaguer P, Demirpence E, Chabret C, Trousse F, Nicolas J-C, Pons M. Stable transfectant cells for studying steroid receptor biological activity. J Biolum Chem 1994;9:201]209. Gillesby BE, Zacharewski T. Exoestrogens : mechanisms of actions and strategies for identification and assessment. Environ Toxicol Chem 1998;17:3]14. Green S, Issemann I, Sheer E. A versatile in vivo and in vitro eukaryotic expression vector for protein engineering. Nucl Acids Res 1988;16:369. Joyeux A, Balaguer P, Nicolas J-C. Engineered cell lines as a tool to monitor biological activity of endocrine hormones. Anal Biochem 1997;249:119]130. Kharat I, Saatcioglu F. Antiestrogenic effects of 2,3,7,8 tetrachlorodibenzo-p-dioxin are mediated by direct transcriptional interference with the liganded estrogen receptor. J Biol Chem 1996;271:10533]10537. Kharat I, Saatcioglu F. Antiestrogenic effects of 2,3,7,8 tetrachlorodibenzo-p-dioxin are mediated by direct transcriptional interference with the ligand estrogen receptor. J Biol Chem 1996;271:10533]10537. Kuiper G, Carlsson B, Grandien K, Enmark E, Haggblad J, Nilsson S, Gustafsson J-A. Comparison of the ligand binding specificity and transcript tissue description of estrogen receptors a and b. Endocrinology 1997;138:863]870. Kuiper G, Lemmen JG, Carlsson B, Corton JC, Safe SH, van der Saag PT, van der Burg B, Gustafsson J-A. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor b. Endocrinology 1998;13:4252]4263. Le Bail JC, Varnat F, Nicolas JC, Habrioux G. Estrogenic and antiproiferative activities on MCF7 human breast cancer cells by flavonoids. Cancer Lett 1998;130:209]216. Lowry OH, Rosenbrough NI, Farz AL, Randall RJ. Protein measurements with the folin phenol reagent. J Biol Chem 1951;193:265]275. Mader S, Leroy P, Chen J-Y, Chambon P. Multiple parameters control the selectivity of nuclear receptors for their response elements. J Biol Chem 1993;12:591]600.

56

P. Balaguer et al. r The Science of the Total En¨ ironment 233 (1999) 47]56

Metzger D, Berry M, Ali S, Chambon P. Effect of antagonists on DNA binding properties of the human estrogen receptor in vitro and in vivo. Mol Endo 1995;9:579]591. Moore M, Mustain M, Daniel K, Chen I, Safe S, Zacharewski T, Gillesby B, Joyeux A, Balaguer P. Antiestrogenic activity of hydroxylated polychlorinated biphenyl congeners identified in human serum. Toxicol Appl Pharmacol 1997;142: 160]168. Routledge EJ, Sumpter JP. Structural features of alkylphenolic chemicals associated with estrogenic activity. J Biol Chem 1997;272:3280]3288.

Soto A, Sonnenschein C, Chung KL, Fernandez MF, Olea N, Serrano FO. The E-screen assay as a tool to identify estrogens: an update on estrogenic environmental pollutants. Environ Health Perspect 1995;103ŽSuppl 7.:113]122. Sumpter JP, Jobling S. Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ Health Perspect 1995;103ŽSuppl 7.:173]178. Zacharewski T. In vitro bioassays for assessing estrogenic substances. Environ Sci Technol 1997;31:613]623.