The E-SCREEN Assay as a Tool to IdentifyEstrogens: An Update on EstrogenicEnvironmental PollutantsAna M. Soto,' Carlos Sonnenschein,1 Kerrie L. Chung,' Mariana F.Fernandez,1 Nicolas Olea,2 and Fatima Olea Serrano31Tufts University School of Medicine, Department of Anatomy and Cellular Biology, Boston, Massachusetts;2School of Medicine, University of Granada, Granada, Spain; 3School of Pharmacy, University of Granada,Granada, Spain

Estrogens are defined by their ability to induce the proliferation of cells of the female genital tract. The wide chemical diversity of estrogeniccompounds precludes an accurate prediction of estrogenic activity on the basis of chemical structure. Rodent bioassays are not suited for thelarge-scale screening of chemicals before their release into the environment because of their cost, complexity, and ethical concerns. The E-SCREENassay was developed to assess the estrogenicity of environmental chemicals using the proliferative effect of estrogens on their target cells as anend point. This quantitative assay compares the cell number achieved by similar inocula of MCF-7 cells in the absence of estrogens (negativecontrol) and in the presence of 1 7,B-estradiol (positive control) and a range of concentrations of chemicals suspected to be estrogenic. Among thecompounds tested, several "new" estrogens were found; alkylphenols, phthalates, some PCB congeners and hydroxylated PCBs, and theinsecticides dieldrin, endosulfan, and toxaphene were estrogenic by the E-SCREEN assay. In addition, these compounds competed with estradiolfor binding to the estrogen receptor and increased the levels of progesterone receptor and pS2 in MCF-7 cells, as expected from estrogen mimics.Recombinant human growth factors (bFGF, EGF, IGF-1) and insulin did not increase cell yields. The aims of the work summarized in this paper werea) to validate the E-SCREEN assay; b) to screen a variety of chemicals present in the environment to identify those that may be causing reproductiveeffects in wildlife and humans; c) to assess whether environmental estrogens may act cumulatively; and finally d) to discuss the reliability of thisand other assays to screen chemicals for their estrogenicity before they are released into the environment. - Environ Health Perspect 1 03(Suppl7):113-122 (1995)

Key words: xenobiotics, cell proliferation, endosulfan, phthalates, antioxidants, alkylphenols

IntroductionFor the last 40 years, substantial evidencehas surfaced on the hormonelike effects ofmany xenobiotics in fish, wildlife, andhumans (1). The endocrine and reproduc-tive effects of xenobiotics are believed to bedue to their a) mimicking effects ofendogenous hormones such as estrogensand androgens; b) antagonizing the effectsof normal, endogenous hormones; c) alter-ing the pattern of synthesis and metabolism

of natural hormones; and d) modifyinghormone receptor levels.

Among environmental chemicals foundto cause reproductive impairment inwildlife and humans there are estrogenmimics (xenoestrogens). Natural estrogenspromote cell proliferation and hypertrophyof female secondary sex organs and inducethe synthesis of cell type-specific proteins(2). Xenobiotics of widely diverse chemicalstructure have estrogenic properties (3,4).

This paper was presented at the Symposium on Estrogens in the Environment, III: Global Health Implicationsheld 9-11 January 1994 in Washington, DC. Manuscript received: March 15, 1995; manuscript accepted:April 4, 1995.

This work was partially supported by grants from the W. Alton Jones Foundation, EPA-CR 820301, NIH-CA-13410, and NSF-DCB-9105594. The assistance of The Center for Reproductive Research at Tufts University(P30 HD 28897) is gratefully acknowledged.

Address correspondence to Dr. Ana M. Soto, Tufts University School of Medicine, Department ofAnatomy and Cellular Biology, 136 Harrison Avenue, Boston, MA 02111. Telephone: (617) 636-6954. Fax:(617) 636-6536. E-mail: ASOTO@OPAL.TUFTS.EDU

Abbreviations used: PCB, polychlorinated biphenyls; bFGF, basic fibroblast growth factor; EGF, epidermalgrowth factor; IGF-1, insulinlike growth factor 1; DME, Dulbecco's modification of Eagle's medium; FBS, fetalbovine serum; E2, 17f-estradiol; DDT, dichlorodiphenyltrichloroethane; DDD, dichlorodiphenyidichloroethane;DDE, dichlorodiphenyidichloroethylene; NIEHS, National Institute of Environmental Health Sciences; DES,diethylstilbestrol; DMSO, dimethyl sulfoxide; CD, charcoaldextran; CDHuS, CD human serum in phenol red-free DME; td, doubling time; PE, proliferative effect; RPP, relative proliferative potency; RPE, relative prolifera-tive effect; PBS, phosphate-buffered saline; BHT, butylated hydroxytoluene; BHA, t-butylhydroxyanisole; PR,progesterone receptor; IC50, concentration that inhibits 50%; APEs, alkylphenol polyethoxylates; DCB, dichloro-biphenyl; RBA, relative binding affinity; TCB, tri- or tetrachlorobiphenyl.

This diversity makes it difficult to predictthe estrogenicity of xenobiotics solely onstructural bases. To overcome this short-coming, their identification as estrogens hasrelied on rodent bioassays. These assaysmeasure either vaginal cornification or theincrease in uterine wet weight; however, thelatter is not a specific estrogen response (5).To obviate problems inherent to animaltesting, quantitative bioassays using cells inculture have been developed. For example,the induction of prolactin in primary sheeppituitary cell culture has been proposed as ameasure of estrogen action (6); in thismodel, estrogens induce protein synthesisbut are ineffective at inducing cell prolifera-tion. The limitations of this assay are thatsome estrogen-inducible genes could also beinduced by nonestrogenic substances. Forexample, prolactin synthesis may beinduced by EGF, thyrotropin releasing fac-tor, and phorbol esters (7). Another estro-gen-inducible marker, ovalbumin synthesis,is stimulated by other steroids such as prog-esterone and glucocorticoids (8). Also,induction of reporter genes under controlby estrogen-responsive elements has been

Environmental Health Perspectives 113

SOTO ETAL.

proposed to assess estrogenicity; however,elevated basal expression in the absence ofestrogen often occurs, and this may raiseconcern about the reliability of these assays.Therefore, the proliferative effect of naturalestrogens on the female genital tractremains the hallmark of estrogen action.

Hertz argued convincingly that thisproliferative property should be adopted asthe one method to determine whether ornot a chemical is an estrogen (2). Thisrequires measuring the increase of mitoticactivity in tissues of the female genital tractafter estrogen administration. However,this method is not suitable for large-scalescreening of suspected chemicals and anequally reliable, easy, and rapid-to-performmethod would be preferable. The novelE-SCREEN assay fulfills these require-ments (9). This assay measures estrogen-induced increase of the number of humanbreast MCF-7 cells and is recognized asbiologically equivalent to the increase ofmitotic activity in the rodent endometrium(9,10). The objectives of this study were tovalidate the E-SCREEN assay and to testthe estrogenicity of chemicals released intothe environment in large volumes.

Materials and MethodsCell Line and Cell Culture ConditionsHuman breast cancer estrogen-sensitiveMCF-7 cells were obtained from theMichigan Cancer Foundation (Detroit,MI) (11). For routine maintenance, cellswere grown in Dulbecco's modification ofEagle's medium (DME) (GIBCo, GrandIsland, NY) supplemented with 5% fetalbovine serum (FBS) (Hyclone, Logan, UT)at 37°C in an atmosphere of 5% C02/95%air under saturating humidity.

Steroids, Xenobiotics, andGrowth Factors Tested17,-Estradiol (E2) was obtained fromCalbiochem (Richmond, CA). Othersteroids were purchased from Steraloids(Keene, NH). R26008 (allenolic acid) wassupplied by Roussel-UCLAF, Romainville,France. Toxaphene (technical grade) andendosulfan (technical grade) were obtainedfrom Chem Services (West Chester, PA).Endosulfan a and P isomers, o,p'-DDT,p,p'-DDT, p,p'-DDD, p,p'-DDE, PCBcongeners, methoxychlor, dieldrin, phtha-late esters, and antioxidants were fromUltra Scientific (North Kingstown, RI).Hydroxylated biphenyls were a gift from J.A. McLachlan (National Institute ofEnvironmental Health Sciences, Research

Triangle Park, NC [NIEHS]). DES metab-olites were a gift of K.S. Korach (NIEHS).Estradiol was stored as a 1-mM stock solu-tion in ethanol at -20°C. Pesticides weredissolved in ethanol to a final concentra-tion of 10 mM, except endosulfan mixedisomers, dieldrin, and toxaphene, whichwere dissolved in dimethyl sulfoxide(DMSO); they were all diluted to desiredconcentrations in phenol red-free DMEimmediately before using. The final solventconcentration in culture medium did notexceed 0.1%; this concentration did notaffect cell yields. Human recombinantEGF, basic FGF, and IGF-1 were pur-chased from Collaborative Research(Lexington, MA).

Plasma-derived and Blood-derivedHuman SerumPlasma-derived human serum was preparedfrom outdated plasma supplied by the NewEngland Medical Center Blood Bank,(Boston, MA). Calcium chloride was addedto a final concentration of 30 mM to facili-tate clot formation. Blood-derived serumwas obtained using blood from healthyadult volunteers; blood was allowed to clotin glass centrifuge tubes for 2 to 4 hr toobtain serum. Plasma- and blood-derivedserum were clarified by centrifugation(2000xg for 10 min), heat-inactivated(56°C for 30 min), centrifuged, charcoal-dextran stripped, and stored in glass tubesat -20°C until use.

Removal ofSex Steroids byCharcoal-Dextran Treatment ofSerumCharcoal (Norit A, acid washed; SigmaChemical Co, St. Louis, MO) was washedtwice with cold sterile water immediatelybefore using. A 5% charcoal-0.5% dextranT70 (Pharmacia-LKB, Uppsala, Sweden)suspension was prepared. Charcoal-dex-tran (CD) suspension aliquots of a volumesimilar to the serum aliquots to beprocessed were centrifuged at 2500 rpm for10 min. Supernatants were aspirated andserum aliquots were mixed with the char-coal pellets. This charcoal-serum mixturewas maintained in suspension by rolling at4 cycles/min at 370C for 1 hr. This suspen-sion was centrifuged at 2000xg for 20 min.The supernatant was then filtered througha 0.45-pm Nalgene filter. Over 99% ofserum sex steroids were removed by thistreatment when determined by removal of3H-E2 (12); E2 levels after CD treatmentwere less than 0.01 pg/ml when measuredby radioimmunoassay. CD sera were storedat -20°C until needed. Samples kept for 1

year in the freezer maintained theirinhibitory properties on the proliferationof human estrogen-sensitive breast tumorMCF-7 cells; plasma- and blood-derivedsera were equally effective.

The E-SCREEN TestThe E-SCREEN assay was developed basedon the following premises: a) a humanserum-borne molecule specifically inhibitsthe proliferation of human estrogen-sensi-tive cells (12-16); and b) estrogens inducecell proliferation by canceling thisinhibitory effect (12,13,16). Nonestrogenicsteroids and growth factors did not abolishthe proliferative inhibition by mammalianserum (12,13).

Cloned MCF-7 cells were trypsinizedand plated into 12-well plates (Costar,Cambridge, MA) at initial concentrationsof 20,000 cells per well (9,10). Cells wereallowed to attach for 24 hr; then, the seedingmedium (5% FBS in DME) was removedand replaced by the experimental medium[5% CD human serum supplemented tophenol red-free DME (CDHuS)]. A rangeof concentrations of the test compounds wasadded to this medium. The bioassay wasterminated on day 6 (late exponentialphase) by removing the media from thewells, adding a cell lysing solution (10%ethylhexadecyl-dimethylammonium bro-mide [Eastman Kodak Co., Rochester,NY] in 0.5% Triton X-100, 2 mM MgCl2,15 mM NaCl, 5 mM phosphate buffer,pH 7.4) and counting the nuclei in aCoulter Counter Apparatus, Model ZM(Coulter Electronics, Hialeah, FL).

The best estimate of the proliferativebehavior of a cell population is td or dou-bling time. td is the time interval in whichan exponentially growing culture doublesits cell number. Determining td requiresmeasuring cell yields at several time inter-vals during the exponential proliferationphase. A less cumbersome alternative tomeasuring proliferation rates is comparingthe cell yield achieved by similar cell inocu-la harvested simultaneously during the lateexponential phase of proliferation. Theproliferative effect (PE) is measured as theratio between the highest cell yieldobtained with the test chemical and withthe hormone-free control. Under theseexperimental conditions, cell yield repre-sents a reliable estimate of the relative pro-liferation rate achieved by similar inoculaexposed to different proliferation regula-tors. In our experimental design, MCF-7cell yields were measured 6 days after to;however, significant differences between

Environmental Health Perspectives114

THE E-SCREEN TESTAND NOVEL XENOESTROGENS

control and estrogen-treated cultures areapparent after 4 days (16).

The estrogenic activity of xenobioticswas assessed a) by determining their relativeproliferative potency (RPP), which mea-sures the ratio between the minimal con-centration of estradiol needed for maximalcell yield and the minimal dose of the testcompound needed to achieve a similareffect; and b) by measuring their relativeproliferative effect (RPE), which is 100times the ratio between the highest cellyield obtained with the chemical and withE2. RPE is calculated as 100 x (PE-1) of thetest compound/(PE-1) of E2. Thus, theRPE indicates whether the compoundbeing tested induces a proliferative responsequantitatively similar to the one obtainedwith E2, that is, a full agonist (RPE = 100),or a proliferative yield significantly lowerthan the one obtained with E2, that is, apartial agonist (9). Figure 1 displays aschematic representation of these concepts.For screening purposes, the range of xenobi-otic concentrations was from 1 nM to 10pM, and for E2 from 0.1 pM to 1 nM, mea-sured at intervals of one order of magnitude.

Progesterone Receptor Assay andEstrogen Receptor ProcessingMCF-7 cells were seeded in 25-cm2 flasksin 5% FBS-supplemented DME. Twenty-four hours later, the medium was changedto 5% CDHuS, and the chemicals to betested were added. Control flasks weretreated with vehicle. After 72 hr of expo-sure to the test xenoestrogens, medium wasaspirated, the cell layer was rinsed withPBS, and the cells were frozen in liquid

600 -

= 400-E

c-a 200-

RPP

I0.1100

37.5

-{RPE

0 0.001 0.1 10 1000Concentration

Figure 1. Schematic representation of the dose-responsecurve to E2 (-*-), a full agonist (-O-), and a par-tial agonist (-A-). The horizontal bars indicate thatRPP is a comparison between effective concentrations ofthe agonist and E2. The vertical bars at the right of thegraph box illustrate that RPE compares the ability of E2and of agonists to increase the cell yield over the valuesobtained in untreated controls. The RPE of the full agonistin this figure is 100, that is, lOOx(500/100)-1(500/100)-l. The RPE of the partial agonist is 37.5,that is, 100 x(250/1 00)- 1/(500/100)-1.

N2. To extract receptor molecules, cellswere incubated with 1 ml of extractionbuffer (0.5 M KCI, 10 mM potassiumphosphate, 1.5 mM EDTA, and 1 mMmonothioglycerol, pH 7.4) at 40C for30 min (17). After centrifugation to pelletthe cell debris, receptor levels were mea-sured in 100--pl extract aliquots by enzymeimmunoassay using the Abbott estrogenand progesterone receptors kits (AbbottDiagnostics, Chicago, IL) according to themanufacturer's instructions.

pS2 AssayCulture media were harvested after 72 hrof exposure to the test chemicals and cen-trifuged to eliminate floating and detachedcells; samples were kept frozen at -80°Cuntil the immunoradiometric assay wasperformed following the manufacturer'sprotocol (ELSA-PS2, CIS Bio Inter-national, Gif-sur-Yvette, France).

Determination of RelativeBinding AffinitiesMCF-7 cells were grown in 150-cm2 flasksin 5% FBS; they were harvested during thelate exponential phase after 24 hr of expo-sure to 5% CDHuS. Cells were rinsed withPBS, and a suspension of 20 x 106 cells/mlof buffer (500 mM KCl, 1.5 mM EDTA,10 mM Tris-HCl, pH 7.4, at 40C) wassonified at 40C (5-sec pulses with 30-secintervals). The cell homogenate was cen-trifuged at 100,OOOxg for 40 min, andsupernatant aliquots were incubated with 2nM 3H-E2 alone and in combination withunlabeled competitors at concentrationsranging from 1 pM to 1 pM E2 or 1 nM to1 mM xenoestrogens for 16 hr at 40C. Thereaction mixture contained 15% DMSOto solubilize hydrophobic xenoestrogens.This treatment did not alter the shape ofthe competition curve for E2 and nonyl-phenol, the only two compounds fromwhich a competition curve could beobtained in the absence of DMSO.Separation of bound and free hormone wasdone by CD adsorption (18).

Statistical AaysisResults were expressed as the mean ± SE.Proliferation yield experiments conductedin duplicate wells were repeated at least aminimum of 5 times. Mean cell numbersfrom each experiment were normalized tothe steroid-free control (100%) to correctfor differences in the initial plating den-sity. Differences between the diversesteroid treatment groups were assessed byanalysis of variance and the a posteriori

Shaffe's test (19). A p value of < 0.05 wasregarded as significant.

ResultsProliferative Effect ofCompoundsKnown To Be Estrogenicin Animal Models

E2 induced maximal cell yields at 10 to100 pM using the E-SCREEN assay.Twenty-two compounds reported to haveestrogenic activity were also tested. TheirRPP is listed in Table 1. Their relativepotency measured by the E-SCREEN assaycorrelated with their relative bindingaffinity to the estrogen receptor and withtheir biological effect in uterotropic assays.Exceptions to these correlations have beenreported in the literature; they reflect ratesof clearance and metabolization of estro-gens (20). The E-SCREEN assay mimicsexposure to a constant level of hormone,much like that achieved in animals byusing estrogen-filled silastic implants.Estriol behaved as a full agonist in the E-SCREEN assay as it did when adminis-tered to animals in multiple doses (21).Similarly, the proliferative potency of DESmetabolites measured by the E-SCREENassay paralleled that in the uterotropicassay; however, pseudo-DES and indane-strol had poor uterotropic activity but werefull agonists when assayed by theE-SCREEN test (RPP = 10). The lowered

Table 1. Proliferative effect of diverse estrogens mea-sured by the E-SCREEN assay.

Compound RPP, %

17j-Estradiol 10017a-Estradiol 10Estrone 1Estriol 10Moxestrol 100016-Hydroxyestrone 0.1Diethylstilbestrol (DES) 1000cis-Tamoxifen 0.001Metabolite E (from tamoxifen) 0.001R26008 0.111 3-Chloromethylestradiol 1000Indanestrol 1 0Indenestrol-A 100Indenestrol-B 1 0Pseudo-DES 0.1Pseudo-DES-e 10Pseudo-DES-z 10Zearalenol 1Zearalenone 1d-Equilenin 1Coumestrol 0.001Ethynyl-estradiol 100

Relative proliferative potency (RPP) is the ratiobetween 170-estradiol and xenoestrogen dosesneeded to produce maximal cell yields x 100.

Volume 103, Supplement 7, October 1995 1 15

SOTO ETAL.

estrogenic potency of these two compoundswas attributed to slow processing or clear-ance of the ligand-bound receptor (20).This notion is at variance with the data dis-cussed below regarding alkylphenols,which are active both in uterotropic and E-SCREEN assays, while displaying dimin-ished processing of estrogen receptors.

Nonestrogenic compounds (natural andsynthetic progestagens, glucocorticoids, andpesticide derivatives such as mirex, chlor-dane-a isomer, chlordane, and heptachlor)did not affect the proliferation of MCF-7cells. No false positives were observed.Moreover, insulin, transferrin, and EGFdid not reverse the inhibitory effect ofserum (12,13,16). This conclusion was val-idated further by using human recombinantbFGF, EGF, and IGF-1 (Table 2). Theseresults strengthen the reliability andspecificity of this "in culture" assay.

Identification ofNew Xenoestrogensamong Antioxidants and PlasticizersAn estrogenic contaminant was isolatedfrom modified polystyrene centrifuge tubes(Corning Glass Co., Corning, NY, Cat.No. 25310-15). After purification by flashchromatography and reverse-phase HPLC,the estrogenic compound was identified bygas chromatography-mass spectrometry asa p-nonylphenol isomer (10). This nonyl-phenol was a full estrogen for MCF-7 cells(RPE=100; RPP=0.0003 %, Table 3).Nonylphenol also increased the mitoticindex of the endometrial epithelium inadult ovariectomized rats. As expectedfrom a genuine estrogen, it also inducedprogesterone receptor in MCF-7 cells.p-Nonylphenol is 10 to 50 times morepotent an estrogen than kepone ando,p'-DDT, and it mimics both the prolifer-ative and inductive properties of naturalestrogens. Alkylphenols with at least athree-carbon alkyl chain were also found tobe estrogenic; p-octylphenol was the morepotent one (RPP = 0.03). Other phenolicantioxidants were tested; among them

Table 2. Effect of recombinant human growth factorson the proliferation of MCF-7 cells grown in mediumcontaining 5% charcoal-dextran stripped human serum(CDHuS).

Compound Concentration Cell number

Control, 5% CDHuS - 38,140 ± 5,9005% CDHuS + estradiol 10 pM 214,120 ±3,4005% CDHuS + bFGF 10 ng/ml 27,310 ±2855% CDHuS + EGF 10 ng/ml 29,820 ±2,8445% CDHuS + IGF-1 10 ng/ml 49,140 ±1,519Cells were harvested after 6 days of exposure to theexperimental media.

polyalkylated, hindered phenols such asbutylated hydroxytoluene (BHT) andIrganox 1640 were not estrogenic, whereast-butylhydroxyanisole (BHA) was estro-genic. Among phthalate esters used as plas-ticizers, those derived from alkylalcoholssuch as dibutylphthalate and diamylphtha-late were not estrogenic, whereas butylben-zylphthalate was estrogenic.

Identification ofEstrogenicPCB CongenersAroclor 1221 (9), 18 PCB congeners, and10 hydroxylated PCBs were tested (Table4). Five PCB congeners were estrogenic inthe E-SCREEN assay indicating that theyare estrogenic per se or they have under-gone hydroxylation by MCF-7 cells; theirRPP was 0.0001. None of the estrogenicPCBs was coplanar. Among the hydroxy-lated PCBs assayed, the most potent were2', 5'-dichloro-4-hydroxybiphenyl,(RPP=0.0 1-0.001), 2',4',6'-trichloro-4-hydroxybiphenyl (RPP=0.01), and2',3',4',5'-tetrachloro-4-hydroxybiphenyl(RPP=0.001). 2',5'-dichloro-3-hydroxy-biphenyl and 2',3',4',5'-tetrachloro-3-hydroxybiphenyl were also estrogenic,albeit 10-fold less potent than their 4-hydroxy isomers (Figure 2). More con-geners should be assayed to derivemeaningful structure-activity relationships

among the 209 PCB congeners and theirmetabolites; the E-SCREEN assay shouldfacilitate this undertaking.

Identification ofEstrgenic PesticidesTable 5 lists a number of pesticides andindustrial chemicals assayed by the E-SCREEN test. None of the herbicides andfungicides tested were estrogenic. Theestrogenicity of DDT, its metabolites,and chlordecone was confirmed by theE-SCREEN test. Among DDT isomers,o,p'-DDT was slightly more potent thanp,p'-DDT, showing significant activity at1 pM, albeit lower than at 10 pM; p,p'-DDT showed estrogenic activity only at10 pM. Methoxychlor was expected to beinactive in the E-SCREEN assay because itrequires metabolic activation, probably inthe liver. However, methoxychlor (98%pure, U.S. EPA standard) and p,p'-meth-oxychlor (99.6% pure) induced the prolif-eration of MCF-7 cells. From these datawe infer that MCF-7 cells have the enzy-matic complement necessary to activateproestrogens. Comparable results werereported by White et al. (22) regardingalkylphenol diethoxylates. A heptachlorderivative, 1-hydroxychlordene, showedsubmaximal estrogenic activity (RPE= 40)at 10 pM. Technical grade endosulfan anda and ,B endosulfan isomers were estrogenic

Table 3. Estrogenic effect of industrial chemicals measured by the E-SCREEN assay.

Compound Concentrationa RPE, %C RPP, %d

Estradiol 30 pM 100 100Phenol 10 PMb 04-Ethylphenol 10 pMb 54-Propylphenol 10 pM 174-sec-Butylphenol 10 pM 76 0.00034-tert-Butylphenol 10 pM 71 0.00034-tert-Pentylphenol 10 pM 105 0.00034-Isopentylphenol 10 pM 93 0.00034-Butoxyphenol 10 pMb 04-Hexyloxyphenol 10 pMb 04-Hydroxybiphenyl 10 pM 87 0.00034,4'-Dihydroxybiphenyl 10 pM 84 0.00031-Naphthol 10 pMb 02-Naphthol 10 pMb 05,6,7,8-Tetrahydronaphthol-2 10 pMb 06-Bromonaphthol-2 10 PM 385-Octylphenol 100 nM 100 0.034-Nonylphenol 1 pM 100 0.003Nonylphenol, technical grade 10 pM 102 0.0003t-Butylhydroxyanisole 50 pM 30 0.00006Benzylbutylphthalate 10 pM 90 0.0003

'The lowest concentration needed for maximal cell yield. bIndicates the highest concentration tested in culture.cThe relative proliferative effect is calculated as 100x(PE-1) of the test compound/(PE-1) of E2; a value of 100indicates that the compound tested is a full agonist, a value of 0 indicates that the compound lacks estrogenicityat the doses tested, and intermediate values suggest that the xenobiotic is a partial agonist. dRelative prolifera-tive potency is the ratio between E2 and xenobiotic doses needed to produce maximal cell yields x 100. All com-pounds designated as full or partial agonists increased cell yields significantly over the hormoneless control(p< 0.05).

Environmental Health Perspectives116

THE E-SCREEN TESTAND NOVEL XENOESTROGENS

Table 4. Estrogenic effect of polychlorinated biphenyl (PCB) congeners and hydroxylated PCBs measured by theE-SCREEN assay.

Compound Name Concentrationa RPE, %C RPP, %d1 2-Monochlorobiphenyl lo pMb 4.42 4-Monochlorobiphenyl 10 PMb 2.1 -

3 2,5-Dichlorobiphenyl 10 pMb 3.7 -

4 2,6-Dichlorobiphenyl 10 pMb 3.4 -

5 3,5-Dichlorobiphenyl 10 pMb 2.7 -

6 2,3,4-Trichlorobiphenyl 10 pM 77.0 0.00017 2,3',5-Trichlorobiphenyl 10 pMb 2.2 -

8 2,3,6-Trichlorobiphenyl 10 PMb 5.8 -

9 2,2',4,5-Tetrachlorobiphenyl 10 pM 61.6 0.000110 2,3,4,4'-Tetrachlorobiphenyl 10 pMb 4.7 -

1 1 2,3,4,5-Tetrachlorobiphenyl 10 pM 39.2 0.000112 2,3,5,6-Tetrachlorobiphenyl 10 pMb 3.1 -

13 2,4,4',6-Tetrachlorobiphenyl 10 PM 75.7 0.000114 2,3,3',4,5-Pentachlorobiphenyl 10 pMb 1.0 -

15 2,3,4,5,6-Pentachlorobiphenyl 10 pMb 4.416 2,2',3,3',5,5'-Hexachlorobiphenyl 10 pMb 1.0 -

17 2,2',3,3',6,6'-Hexachlorobiphenyl 10 PM 61.6 0.000118 Decachlorobiphenyl 10 pMb 3.5 -

19 2',5'-Dichloro-2-hydroxybiphenyl 10 pM 13.0 0.000120 2',5'-Dichloro-3-hydroxybiphenyl 10 PM 69.9 0.000121 2',5'-Dichloro-4-hydroxybiphenyl 1 pMe 71.2 0.00122 3,5-Dichloro-2-hydroxybiphenyl 10 pMb 5.4 -

23 3,5-Dichloro-4-hydroxybiphenyl 10 pMb 1.5 -

24 2,2', 5-Trichloro-4-hydroxybiphenyl 1 pM 37.8 0.00125 2',4',6'-Trichloro-4-hydroxybiphenyl 100 nM 99.8 0.0126 2',3',4',5'-Tetrachloro-3-hydroxybiphenyl 10 PM 35.3 0.000127 2',3',4',5'-Tetrachloro-4-hydroxybiphenyl 1 pM 92.0 0.00128 2',3',4',5,5'-Pentachloro-2-hydroxybiphenyl 10 pMb 4.8 -

aIndicates the lowest concentration needed for maximal cell yield. bIndicates the highest concentration tested inculture. CRelative proliferative effect is calculated as 100 x (PE-1) of the test compound/(PE-1) of E2; a value of 100indicates that the compound tested is a full agonist, a value of 0 indicates that the compound lacks estrogenicityat the doses tested, and intermediate values suggest that the xenobiotic is a partial agonist. dRelative prolifera-tive potency is the ratio between E2 and xenobiotic doses needed to produce maximal cell yields x 100. All com-pounds designated as full or partial agonists increased cell yields significantly over the hormoneless control(p<0.05). 'This compound exhibited submaximal proliferative activity at 100 nM (RPE:62%).

450-

E30001

HC A B C D E F

W NC M- 10plM 1ijaWIOnlJm 011p

Figure 2. Proliferative activity of HC, hormonelesscontrol; A, estradiol; B, 2',5',2-hydroxy-DCB; C, 2',5',3-hydroxy-DCB; D, 2',5',4-hydroxy-DCB; E, 2',3',4',5',3-hydroxy-TCB; and F, 2',3',4',5',4-hydroxy-TCB.Asterisks (*) indicate significant differences with hor-moneless control (p<0.05).

at concentrations of 10 to 25 pM. Itshould be noted that the RPE of all thesechemicals is lower than that of estradiol(endosulfanmixed isomers RPE 81 %,a isomer RPE = 77%, 3 isomer RPE 78%)(Table 6). Dieldrin and toxaphene were

found to be estrogenic at 10 pM. TheRPEs of these compounds were lower thanthose of endosulfan (dieldrin, RPE= 55%;toxaphene, RPE = 52%) (23). The fact thatthese compounds have toxic effects at con-centrations one order of magnitude higherthan those needed to evoke a proliferativeresponse precluded assessing whetherhigher concentrations would attain fullestrogenic activity in this bioassay.

Induction ofProgserone Receptorand pS2 by the Newly IdentifiedX oestognsPesticides found to be estrogenic by theE-SCREEN assay were also effective induc-ers of markers of estrogen action such asprogesterone receptors (PR) and pS2 inMCF-7 cells (Table 7). Dieldrin signifi-cantly increased pS2 levels; PR levelsincreased slightly at the maximal dosetested (10 IM). Dieldrin is toxic at concen-trations higher than 10 pM; this predudedtesting whether higher concentrationswould result in full induction of PR.

Nonylphenol (J0) and octylphenol (22)also induced PR.

Xenoestrogens and the Prssingof Estrogen ReceptorsEstradiol treatment of MCF-7 cells for aperiod of 72 hr decreased the level of estro-gen receptors by 50% (receptor process-ing); this effect is interpreted by some as animportant step on estrogen action (24).Others have reported that the proliferativeeffect of estradiol also occurs in the absenceof processing (25). Treatment with xeno-estrogens such as endosulfan, toxaphene,dieldrin (Table 8), and nonylphenol (notshown) did not decrease estrogen receptorlevels in MCF-7 cells; to the contrary, aslight increase of up to 57% was recorded.

Competitive Binding to theEstrgn Receptoro,p'-DDT, endosulfan, toxaphene, andnonylphenol competed with 3H-E2 forbinding to the receptor; their relativebinding affinities correlated with their rel-ative proliferative potency. Table 9 com-pares the relative binding affinities of thesexenoestrogens, their IC50 values and theirrelative proliferative potency. The concen-tration of estradiol and xenoestrogens nec-essary to decrease 3H-E2 binding by 50%were about one order of magnitude higherthan the concentration needed to achievemaximal cell proliferation yields, PR andpS2 induction.

Cumulative Effect ofXnoestogensAnalysis of fat or serum from wildlife andhumans often reveals the simultaneouspresence of several xenoestrogens (1); thesefindings suggest that xenoestrogens may actcumulatively. By using the E-SCREENassay we verified this suggestion. Figure 3shows the additive effect of 10 xenoestro-gens, each administered at one tenth of theminimal effective dose that produces aproliferative effect.

DiscussionThe deleterious impact of xenoestrogens onthe reproductive success, development, andhealth of animals is well documented; therealization that humans are also exposedand at risk has become increasingly obvious(26). Data showing a lowering of spermquality and quantity, increased infertility,and spontaneous abortion rates in humanssuggest that environmental estrogens play arole in the toxicology of human reproduc-tion and development (27,28). An objec-tive causal relationship between detrimental

Volume 103, Supplement 7, October 1995 11 7

SOTO ETAL.

Table 5. Xenobiotics tested by the E-SCREEN assay.

Estrogenic xenobiotics

HerbicidesNone 2,4-D

AlachlorButylateDacthalHexazinonePropazineTrifluralin

Insecticidesp,p'-DDTo,p'-DDTo,p'-DDEp,p'-DDEDDTaDieldrinChlordecone (kepone)Endosulfanaa-EndosulfanP-EndosulfanMethoxychlorToxaphene

FungicidesNone

Industrial chemicals2,3,4-TCB2,2',4,5-TCB2,3,4,5-TCB2,4,4',6-TCB2,2',3,3',6,6'-HCB2-OH-2',5'-DCB3-OH-2',5'-DCB4-OH-2',5'-DCB4-OH-2,2',5-TCB4-OH-2',4',6'-TCB3-OH-2',3',4',5'-TCB4-OH-2',3',4',5'-TCB4-OH-alkyl-phenolsBisphenol-A4-OH-biphenylt-ButylhydroxyanisoleBenzylbutylphthalate

a Denotes a technical grade isomer mixture.

Nonestrogenic xenobiotics

2,4-DBAtrazineCyanazineDinosebMetolachlorPicloramSimazine

BendiocarbChlordaneDiazinonHeptachlorKelthaneMalathionMethopreneParathionRotenone

ChlorothalonilHexachlorobenzeneManebMetiram

Butylated hydroxytoluene2,5-DCB2,6-DCB3,5-DCB2,3',5-TCB2,3,4,4'-TCB2,3,4,5,6-PCB2-OH-2',3',4',5,5'-PCB2-OH-3,5-DCB4-OH-3,5-DCB1 ,2-DichloropropaneIrganox 16402,3,7,8-TCDDTetrachloroethylene

CarbofuranChlordimeformChlorpyrifosCarbarylLindaneMirexPyrethrum

ThiramZinebZiram

2-CB4-CB2,3,6-TCB2,3,5,6-TCB2,3,3'4,5'-PCB2,2',3,3',5,5'-HCB

DecaCBDiamyl phthalateDibutyl phthalateDimethyl isophthalateDimethyl terephthalateDinonyl phthalateOctachlorostyreneStyrene

Table 6. Estrogenic effect of pesticides according tothe E-SCREEN assay.

Compound Concentration RPE, % RPP, %

Estradiol 10 pM 100 100DDT* 10 pM 79.61 0.0001o,p'-DDT 10 pM 86.14 0.0001p,p'-DDT 10 pM 71.00 0.0001Dieldrin 10 pM 54.89 0.0001Endosulfan* 10 pM 81.25 0.0001,B-Endosulfan 10 pM 78.26 0.0001ax-Endosulfan 10 pM 77.17 0.00011-Hydroxychlordene 10 pM 40.00 0.0001Kepone 10 pM 84.00 0.0001Toxaphene 10 pM 51.90 0.0001Methoxychlor 10 pM 57.00 0.0001

Concentration describes the dose at which an estro-genic effect is detected; maximal cell yield is obtainedat concentrations between 10 and 100 pM estradiol.Most xenobiotics are active at 10 pM. The RPE mea-sures the ratio between the maximal cell yieldachieved by the xenobiotic and that of estradiol. TheRPP measures the ratio between the dose of estradioland that of xenobiotic needed to achieve a prolifera-tive effect. Asterisks (*) denote that the compoundreferred to was of technical grade.

Table 7. Effect of xenoestrogens on progesteronereceptor and pS2 levels.

Concen-Compound tration PRa pS2b

Control - 7.9 ± 5.4 53.9± 16.7Estradiol 1 nM 153.1 +49.5* 179.6±50.2Endosulfan 1 pM 71.0 ± 13.1 * 129.2 ± 6.2*Endosulfan 10 pM 136.0 ± 27.2* 172.1 ± 61.6*Toxaphene 1 pM 18.1 ±9.2 59.9±6.2Toxaphene 10 pM 128.7 ± 19.1* 108.0 ± 17.6*Dieldrin 1 pM 8.2 ± 1.7 210.3 ± 29.9*Dieldrin 10 pM 32.3 ± 11.7 232.7 ± 32.6*

aProgesterone receptor (PR) levels are expressed asfemtomoles per mg cellular protein. bpS2 levels in theculture medium are expressed as ng/106 cells. Eachresult represents the mean ± SD of three experiments.An asterisk (*) denotes a significant difference with thehormoneless control (p<0.05).

health effects and their presumed causationby xenoestrogens is tempered however bythe lack of appropriate technology toexplore this subject on a large scale. Thefirst obstacle encountered is that the estro-genicity of chemicals cannot be predictedsolely on structural bases; therefore, it isunknown how many xenoestrogens are pre-sent in the ecosystem. The E-SCREENbioassay is a reliable tool to rapidly assessestrogenicity on a large number of com-pounds; in this paper we describe how itsuse helped to identify estrogens amongenvironmental pollutants. The secondobstacle to overcome is how to identifycausal agents when signs of estrogen

exposure have been verified. The finding ofreproductive effects caused by xenobioticshas largely been accidental. For example,workers at a kepone-producing plant devel-oped azoospermia and impotence (29); thiswas the first observation of reproductivetoxicity by kepone. Because of the occupa-tional nature of this case, the culprit keponebecame readily apparent. In contrast,wildlife are exposed to a combination ofxenobiotics. It became clear in the GreatLakes studies that it was difficult to sort outwhich one of the xenobiotics played acausal role or whether the signs of intoxica-tion were due to cumulative interactionamong the chemicals present in affected

Table 8. Effect of xenoestrogens on estrogen receptorprocessing.

Compound Concentration Estrogen receptoraControl - 183.9 ± 59Estradiol 1 nM 92.3 ± 23.1Endosulfan 1 pM 205.1 ± 21.7cEndosulfan 10 pM 238.3 ± 17.6cToxaphene 1 pM 261.9±1521,cToxaphene 10 pM 288.7±31.7bcDieldrin 1 pM 236.3±21.2bcDieldrin 10 pM 252.2±17.1 ,c

aEstrogen receptor levels are expressed as femto-moles/mg cellular protein. Each result represents themean ±SD of at least three experiments. bDenotes astatistically significant difference with the hormone-less control. cDenotes a statistically significant differ-ence with the estradiol-treated group (p< 0.05).

Environmental Health Perspectives118

THE E-SCREEN TESTAND NOVEL XENOESTROGENS

Table 9. Relative binding affinities (RBAs) and IC50 values of xenoestrogens.

Compound RBA RPP IC50 Concentration*

Estradiol 100 100 1.5±0.4 nM 100 pM,B-Endosulfan 0.00024 0.0001 631.0±88 pM 10 pMToxaphene 0.00032 0.0001 470.0±38 pM 10 pMo,p'-DDT 0.00031 0.0001 485.0±42 pM 10 pMp-Nonylphenol 0.021 0.001 7.2±3 pM 1 pM

*Concentration needed for maximal proliferative effect. All xenoestrogen IC50 values were significantly differentfrom that of estradiol.

animals (1). Exposure to environmentalestrogens, singly or in combination, maybe easily assessed in male fish, reptiles, orbirds used as sentinels by measuring theirvitellogenin plasma levels. Instead, expo-sure of females to xenoestrogens is moredifficult to ascertain through a marker suchas vitellogenin because the serum levels ofthis protein are high in animals laying eggs(30,31). In addition, there is no compara-ble marker to ascertain exposure inmammals. Again, the E-SCREEN assayrepresents the best alternative to resolvethis second obstacle.

Animal Bioassays and the E-SCREENAssay: Differences and SimilaritiesDiverse animal models and assays havebeen used to measure estrogenicity. Allenand Doisy (32) and other pioneers ofestrogen research used mouse and ratactivity units to follow their estrogenpurification protocol; the end point oftheir assay was vaginal cornification.Dodds and Lawson(33) used both theAllen and Doisy assay (32) and the femi-nization of the feather pattern in brownleghorn capons (33). Others adopted theuterotropic assay using single or multipledoses of estrogens over 24- to 72-hr peri-ods in immature or ovariectomized miceand rats. This diversity of end points indi-cates that there is no universal "gold stan-dard" of estrogen action among animalbioassays. The E-SCREEN assay appears tobe the best candidate for establishing aquantitative standard of estrogenic activityat the target organ level. As shown inTables 1 through 6, no false positives ornegatives were observed among the estro-gens and nonestrogens tested.

No qualitative differences could befound when comparing animal assays andthe E-SCREEN assay; that is, the estro-genic properties of compounds character-ized using animal bioassays was alsoascertained using the E-SCREEN test.From a pharmacokinetic perspective, thelatter measures estrogenicity at the targetcell level under conditions where estrogen

levels are mostly constant, much like theones achieved when animals are treatedwith estrogen-filled silastic implants. Thisapproach is more relevant to chronic envi-ronmental exposure than that of measuringacute effects after a single dose. In bothtypes of assays, metabolism of the suspectedxenoestrogen into more or less active com-pounds is uncertain and should be definedindividually for each compound. For exam-ple, nonylphenol diethoxylate was estro-genic for MCF-7 cells; since it does notcompete for estradiol binding to the estro-gen receptor, it is likely that estrogenicactivity results from nonylphenol diethoxy-late metabolism to the free phenol (22).Methoxychlor was also believed to be inac-tive until metabolized to free phenols, pre-sumably in the liver; again, methoxychlortested positive when assayed by the E-SCREEN test. Therefore, even though theputative proestrogens tested so far wereestrogenic when assayed by the E-SCREEN test, an added step in the questfor identifying all xenoestrogens mayinclude their metabolic activation by livermicrosome extracts prior to their testing bythe E-SCREEN assay.

Regarding quantitative effects, whilekepone is 100,000 to 1,000,000 times lesspotent than estradiol according to theE-SCREEN assay, an increase of the ratuterine wet weight comparable to that ofestradiol occurred with a 1000- to 5000-foldhigher dose of chlordecone than that ofestradiol (3). This discrepancy may be dueto rapid metabolism of estradiol and persis-tence and bioaccumulation of chlordeconein animals. Differences between results inculture and in live animals reflect the dif-ferent parameters used as a measure ofestrogenicity. On one hand, the rodentassay measures the increase of uterine wetweight [water imbibition, hypertrophy(which is also produced by estrogen antag-onists), and hyperplasia] (5), while on theother hand, the human E-SCREEN bio-assay measures cell proliferation only. Thisis a necessary and sufficient parameter todefine estrogen action (2).

400-

e 300- *

100

A B C D E F G H I J K L M

Figure 3. Proliferative activity of A, hormoneless con-trol; B, 10 pM 1-endosulfan; C, lpM 3-endosulfan; D,1 pM a-endosulfan; E, 1 pM toxaphene; F, 1 pM dield-rin; G, 1 pM 2,3,4,5-tetrachlorobiphenyl; H, 1 pM p,p'-DDT; I, 1 pM 2,2',3,3',6,6'-hexachlorobiphenyl; J, 1 pMp,p'-DDD; K, 1 pM p,p'-DDE; L, 1 pM methoxychlor;and M, mixture of the 10 chemicals indicated as C to L,each at 1 pM. Asterisks (*) indicate significant differ-ences with hormoneless control (p< 0.05).

New Estrogns Identfied by theE-SCREENAssayNovel xenoestrogens were found among

antioxidants, plasticizers, polychlorinatedbiphenyl (PCB) congeners, and pesticides.

Alkylphenols are used as antioxidantsand in the synthesis of detergents[alkylphenol polyethoxylates, (APEs)].APEs are used as industrial detergents inthe textile and paper industries, in toi-letries, and as spermicides. Four hun-dred and fifty million pounds of APEswere sold in the United States in 1990.APEs are not estrogenic per se; however,they are degraded during sewage treat-

ment. The polyethoxylate chain isshortened, and free alkylphenols as wellas mono and diethoxylates are pro-

duced. The free phenols are estrogenic(9,10). Recently, White et al. (22) haveshown that the diethoxylates are alsoestrogenic. These APE degradationproducts have been detected in drinkingwater (34). Nonylphenol has beenreported to leach from PVC tubing formilk processing (35) and plastics usedin food packaging (36). APEs such as

those used as spermicides are degradedto free nonylphenol when administeredto rodents (37). The contribution ofAPEs and alkylphenols to the xenoestro-

gen burden of humans is unknown;however, it has been reported that thesechemicals are present in sewage outletsin concentrations sufficient to feminizesentinel fish (38). Alkylphenols accu-

mulate in river sediment and in the fatof exposed fish (39). Some phenolicantioxidants such as butylated hydroxy-toluene (BHT) and t-butylhydroxy-anisole (BHA) are used to prolong theshelf life of foodstuffs and to reduce

Volume 103, Supplement 7, October 1995 1 19

SOTO ETAL.

nutritional losses by retarding oxidation.Interesting observations pertaining tostructure-function relationships weremade: a) the alkyl chain must at leasthave 3 carbons; b) the p-isomers aremore potent estrogens than the m-iso-mers; c) polyalkylated, hindered phenolslike BHT and Irganox 1640 (Ciba-Geigy, Basel, Switzerland) are not estro-genic while being effective antioxidants;and d) fused rings like naphthols arenot estrogenic in spite of being an inte-gral part of the A and B ring of naturalsteroids. Instead, substituted naphtholssuch as 6-Br naphthol and allenolic acidare estrogenic; more studies are neededto assess whether these substitutednaphthols are active due to a bulk effect,electronegativity, or because flat mole-cules sucli as naphthols and coplanarPCBs are unable to bind tightly to theestrogen receptor.In addition to the estrogenic alkylphenol

antioxidants described above, we found thatBHA was estrogenic. BHA is a widely usedantioxidant; because it controls oxidation ofshort-chain fatty acids such as coconut oil(40). Maximal usage levels ofBHA permit-ted by U.S. Food and Drug Administration(U.S. FDA) varies according to the foodtype, from 50 ppm in dry breakfast cerealsto 1000 ppm in active yeast (41).

Plasticizers are used to decrease therigidity of certain polymers. For themost part, they are di- and triesters oforganic acids. Phthalate esters are widelyused plasticizers (42). These compoundsleach from plastics, and they have beenfound to be ubiquitously distributed inthe environment, including marineecosystems (43,44). Among phthalateesters butyl benzyl phthalate was estro-genic whereas those derived from alkylalcohols such as dibutyl phthalate anddiamyl phthalate were not.

* Aroclor 1221 was not estrogenic by theE-SCREEN assay (data not shown).Many congeners are present in Aroclormixtures; therefore, it is likely that themaximal concentration used, 10 pM,resulted in levels of individual con-geners lower than those needed toinduce cell proliferation. PCB mixturessuch as Aroclor 1221 were reported tobe estrogenic using uterotropic assays asend points (increased uterine wetweight, increased uterine glycogen con-tent) (45). However, the magnitude ofthe uterine wet weight increase wasonly marginally significant, analyzed byinappropriate statistical tests and lower

than that achieved with E2 or DDT(45,46). Therefore, the estrogenicity ofPCB mixtures would be best ascer-tained by first determining which con-geners are estrogenic. Korach et al. (47)demonstrated the ability of certainhydroxy-PCBs to bind to estrogenreceptors and to produce an uterotropiceffect that correlated with their relativebinding affinity to the estrogen recep-tor. It is generally assumed that hydrox-ylated PCBs are estrogenic whilenonhydroxylated PCBs are not. In thispaper we show that 5 of the 18 con-geners studied were estrogenic in theE-SCREEN assay (Table 4). It isunknown whether they were estrogenicper se or they were hydroxylated byMCF-7 cells; hydroxylated PCBs aremore potent than their nonhydroxy-lated counterparts. The phenyl rings inthe active compounds were not copla-nar. Among the hydroxylated PCBs,those p-hydroxylated were more potentthan m-hydroxylated; o-hydroxylatedcompounds were even less active.Hindered phenols, such as in 4-OH-3,5,-DCB were less estrogenic thanunobstructed ones.The pesticides dieldrin and toxapheneare estrogenic. Their use has beenrestricted in the United States since1974 and 1982, respectively (23).These compounds are highly lipophilicand bioaccumulate in ecosystems; theyare still found in wildlife, coincidentallywith signs of reproductive impairment.Toxaphene is a main airborne pollutantin North America, and its residuesappeared in regions where it has neverbeen used, like the Arctic andScandinavia (48). It is present in Arcticand Baltic salmon muscle fat at concen-trations of 700 to 7000 ppb (49); thisconcentration is well within thoseproducing estrogenic effects in theE-SCREEN assay (10 pM = 4800 ppb).Endosulfan was introduced in 1954; it

is presently used for agricultural purposesin the United States and other countries(50). Proliferative, estrogenlike effects inMCF-7 cells were found at doses of 10 pM(4060 ppb). Endosulfan was shown to pro-duce testicular atrophy in male rats fed adiet containing 10 ppm (51,52); it alsolowered gonadotrophin and testosteroneplasma levels (53). These results are consis-tent with its estrogenicity revealed by theE-SCREEN test.

These newly identified estrogens notonly induced cell proliferation but pS2 and

PR as well; this confirms their estrogen-mimicking properties and the specificity ofthe E-SCREEN assay as a tool to identifyestrogens. These xenoestrogens competewith estradiol for binding to its receptor;their RBA's correlated well with theirpotency to induce cell proliferation, pS2,and PR. Recent data suggest that alkylphe-nols bind to the estrogen-binding domainof the estrogen receptor (22). Binding tothe receptor is a necessary but insufficientproperty to define estrogenicity. Tetra-hydronaphthol, which is not estrogenic, iseffective in preventing the forward bindingof estradiol to its receptor while alkylphe-nols, which are estrogenic, are effective indisplacing prebound estradiol from itsreceptor (54). These data were interpretedas indicative of interactions with more thanone site on the estradiol receptor.

Cumulative Effect ofXenoestrogensXenoestrogens may act cumulatively (Figure3) and with endogenous estrogens thus dis-rupting the endocrine system of exposedwildlife and humans. Hence, measuringthe total estrogenic burden due to envi-ronmental contaminants present in aplasma/tissue sample may be more mean-ingful than to assess exposure by measur-ing the levels of each of the knownxenoestrogens. Recently, an epidemiologi-cal study showed a positive correlationbetween breast cancer and serum levels ofDDE, a DDT metabolite (55-57), leavingopen the possibility that xenoestrogen expo-sure increases the incidence of breast can-cer. Since xenoestrogens are postulated tobe a risk factor for breast cancer, measuringa single xenoestrogen may not be a reliableindicator of exposure because different per-sons eating different diets may be exposedto different xenoestrogens. Therefore, mea-suring total xenoestrogen burden representsa more reliable approach to assess the linkbetween xenoestrogens and breast cancer.The E-SCREEN test may be used to thisend once a protocol is developed to separateenvironmental estrogens from endogenousones. In addition, in a preventive approach,the E-SCREEN test may be used to screenchemicals for their estrogenicity before theyare released into the environment.

REFERENCES

1. Colborn T, vom Saal FS, Soto AM.Developmental effects of endocrine-disrupting chemicals in wildlife andhumans. Environ Health Perspect101:378-384 (1993).

120 Environmental Health Perspectives

THE E-SCREEN TESTAND NOVEL XENOESTROGENS

2. Hertz R. The estrogen problem - retrospect and prospect. In:Estrogens in the Environment II-Influences on Development(McLachlan JA, ed). New York:Elsevier, 1985;1-1 1.

3. Hammond B, Katzenellenbogen BS, Kranthammer N,McConnell J. Estrogenic activity of the insecticide chlordecone(kepone) and interaction with uterine estrogen receptors. ProcNatl Acad Sci USA 76:6641-6645 (1979).

4. Meyers CY, Matthews WS, Ho LL, Kolb VM, Parady TE.Carboxylic acid formation from kepone. In: Catalysis inOrganic Synthesis (Smith GW, ed). New York:Academic Press,1977;213-255.

5. Clark JH, Watson C, Upchurch S, Padykula H, MarkaverichB, Hardin JW. Estrogen action in normal and abnormal cellgrowth. In: Estrogens in the Environment (McLachlan JA, ed).New York:Elsevier, 1980;53-67.

6. Lieberman ME, Maurer RA, Gorski J. Estrogen control of pro-lactin synthesis in vitro. Proc Natd Acad Sci USA 75:5946-5949(1978).

7. Ramsdell JS, Tashjian AH. Thyro-tropin releasing hormoneand epidermal growth factor stimulate prolactin synthesis by apathway(s) that differs from that used by phorbol esters: disso-ciation of actions by calcium dependency additivity.Endocrinology 117:2050-2060 (1985).

8. Palmiter RD. Quantitation of parameters that determine therate of ovalbumin synthesis. Cell 4:189-197 (1975).

9. Soto AM, Lin T-M, Justicia H, Silvia RM, Sonnenschein C.An "in cu!ture" bioassay to assess the estrogenicity of xenobi-otics. In: Chemically Induced Alterations in SexualDevelopment: The Wildlife/Human Connection (Colborn T,Clement C, eds). Princeton, NJ:Princeton ScientificPublishing, 1992;295-309.

10. Soto AM, Justicia H, Wray JW, Sonnenschein C. p-Nonyl-phenol: an estrogenic xenobiotic released from "modified"polystyrene. Environ Health Perspect 92:167-173 (1991).

11. Soule HD, Vasq uez J, Long A, Albert S, Brennan MJ. Ahuman cell line from a pleural effusion derived from a breastcarcinoma. J Natl Cancer Inst 51:1409-1413 (1973).

12. Soto AM, Sonnenschein C. The role of estrogens on the prolif-eration of human breast tumor cells (MCF-7). J SteroidBiochem 23:87-94 (1985).

13. Soto AM, Sonnenschein C. Mechanism of estrogen action oncellular proliferation: evidence for indirect and negative controlon cloned breast tumor cells. Biochem Biophys Res Commun122:1097-1103 (1984).

14. Lykkesfeldt AE, Briand P. Indirect mechanism of oestradiolstimulation of cell proliferation of human breast cancer celllines. BrJ Cancer 53:29-35 (1986).

15. Briand P. Lykkesfeldt AE. Long-term cultivation of a humanbreast cancer cell line, MCF7, in chemically defined medium.Effect of estradiol. Anticancer Res 6:85-90 (1986).

16. Soto AM, Silvia RM, Sonnenschein C. A plasma-borne specificinhibitor of the proliferation of human estrogen-sensitive breasttumor cells (estrocolyone-I). J Steroid Biochem Mol Biol43:703-712 (1992).

17. Maddedu L, Legros N, Devleeschouwer N, Bosman C, PiccartM, Leclercq G. Estrogen receptor status and estradiol sensitivityofMCF-7 cells in exponential growth phase. Eur J Cancer ClinOncol 24:385-390 (1988).

18. Sonnenschein C, Olea N, Pasanen ME, Soto AM. Negativecontrols of cell proliferation: human prostate cancer cel s andandrogens. Cancer Res 49:3474-3481 (1989).

19. Winer BJ. Design and analysis of single-factor experiments. In:Statistical Principles in Experimental Design. 2nd ed. NewYork:McGraw-Hill, 1962; 149-260.

20. Korach KS, Metzler M, McLachlan JA. Diethylstilbestrolmetabolites and analogs. J Biol Chem 254:8963-8968 (1979).

21. Martin L, Pollard JW, Fagg B. Oestriol, oestradiol-17,B and theproliferation and death of uterine cells. J Endocrinol 69:103-115 (1976).

22. White R, Jobling S, Hoare SA, Sumpter JP, Parker MG.Environmentally persistent alkylphenolic compounds.Endocrinology 135: 175-182 (1994).

23. Soto AM, Chung KL, Sonnenschein C. The pesticides endosul-.fan, toxaphene, and dieldrin have estrogenic effects on humanestrogen sensitive cells. Environ Health Perspect 102:380-383(1994).

24. Gyling M, Leclercq G. Estrogenic and antiestrogenic down-reg-ulation of estrogen receptor levels: evidence for two differentmechanisms. J Recept Res 10:217-234 (1990).

25. Reiner GCA, Katzenellenbogen BS. Characterization of estro-gen and progesterone receptors and the dissociated regulationof growth and progesterone receptor stimulation by estrogen inMDA-MB-134 human breast cancer cells. Cancer Res46:1124-1131 (1986).

26. Thomas KB, Colborn T. Organochlorine endocrine disprup-tors in human tissue. In: Chemically-induced Alterations inSexual Development: The Wildlife/Human Connection(Colborn T, Clement C, eds). Princeton,NJ:PrincetonScientific Publishing, 1992;365-394.

27. Sharpe RM, Skakkebaek NE. Are oestrogens involved in fallingsperm count and disorders of the male reproductive tract?Lancet 341:1392-1395 (1993).

28. Giwercman A, Carlsen E, Keiding N, Skakkebaek NE.Evidence for increasing incidence of abnormalities of thehuman testis: a review. Environ Health Perspect 101(Suppl2):65-71 (1993).

29. Guzelian PS. Comparative toxicology of chlorodecone (kepone)in humans and experimental animals. Annu Rev PharmocolToxicol 22:89-113 (1982).

30. Bergink EW, Wallace RA, VandeBerg JA, Bos ES, Ab G.Estrogen-induced synthesis of yolk proteins in roosters. AmZool 14:1177 (1974).

31. Redshaw MR, Follett BK. Physiology of egg yolk productionby the fowl: the measurement of circulating levels of vitel-logenin employing a specific radioimmunoassay. CompBiochem Physiol A 55:399-405 (1976).

32. Allen E, Doisy EA. An ovarian hormone: preliminary report onits localization, extraction and partial purification, and action intest animals. JAMA 81:819-821 (1923).

33. Dodds EC, Lawson W. Molecular structure in relation tooestrogenic activity. Compounds without a phenathrenenucleus. Proc Roy Soc Lond B Biol Sci 125:222-232 (1938).

34. Clark LB, Rosen RB, Hartman TG, Louis JB, Suffet IH,Lippincott RL, Rosen JD. Determination of alkylphenolethoxylates and their acetic acid derivatives in drinking waterby particle beam liquid chromatography/mass spectrometry. IntJ Environ Anal Chem 47:167-180 (1992).

35. Junk GA, Svec HJ, Vick RD, Avery MJ. Contamination ofwater by synthetic polymer tubes. Environ Sci Technol8:1100-1106 (1974).

36. Gilbert MA, Shepherd MK, Startin JR, Wallwork MA.Identification by gas chromatography-mass spectrometry ofvinylchloride oligomers and other low molecular weight com-ponents of poly(vinylchloride) resins for food package applica-tions. J Chromatogr 237:249-261 (1992).

37. Knaak JB, Elridge JM, Sullivan LJ. Excretion of certain poly-ethylene glycol ether adducts of nonylphenol by the rat.Toxicol Appl Pharmacol 9:331-340 (1966).

38. Purdom CE, Hardiman PA, Bye VJ, Eno NC, Tyler CR,Sumpter JP. Estrogenic effects from sewage treatment works.Chem Ecol 8:275-285 (1994).

39. Shiraishi H, Carter DS, Hites RA. Identification and determi-nation of tert-amylphenols in carp from the Trenton channel ofthe Detroit river, Michigan, U.S.A. Biomed Environ MassSpectrom 18:478-483 (1989).

40. Shahidi F, Wanasundara PK. Phenolic antioxidants. Crit RevFood Sci Nutr 32:67-103 (1992).

41. Singer PL. Occupational oligospermia. JAMA 140:1249(1949).

42. Tepper L. Phthalic acid esters - an overview. Environ HealthPerspect 3:179-182 (1973).

43. Giam CS, Neff GS, Atlas EL, Chan HS. Phthalate ester plasti-cizers: a new class of marine pollutant. Science 199:419-421(1978).

Volume 103, Supplement 7, October 1995 121

SOTO ETAL.

44. Randall R, Ozretich R, Boese B. Acute toxicity of butyl benzylphthalate to the saltwater fish English sole, Parophrys vetulus.Environ Sci Technol 17:670-672 (1983).

45. Ecobichon DJ, MacKenzie DO. The uterotropic activity ofcommercial and isomerically-pure chlorobiphenyls in the rat.Res Commun Chem Path Pharmacol 9:85-94 (1974).

46. Gellert RJ. Uterotrophic activity of polychlorinated biphenyls(PCB) and induction of precocious reproductive aging inneonatally treated female rats. Environ Res 16:123-130 (1978).

47. Korach KS, Sarver P, Chae K, McLachlan JA, McKinney JD.Estrogen receptor-binding activity of polychlorinated hydroxy-biphenyls: conformationally restricted structural probes. MolPharmacol 33:120-126 (1987).

48. ATSDR. Toxicological Profile for Toxaphene. Rpt No TP-90-26. Atlanta:Agency for Toxic Substances and Disease Registry,1990.

49. Paasivirta J, Rantio T. Chloroterpenes and other organochlo-rines in Baltic, Finnish and Arctic wildlife. Chemosphere22(1):47-55 (1991).

50. ATSDR. Toxicological Profile for Endosulfan. Rpt No TP-91/16. Atlanta:Agency for Toxic Substances and DiseaseRegistry, 1991.

51. NCI. Bioassay of endosulfan for possible carcinogenicity. TechRpt Ser NCI-CG-TR-62. National Cancer Institute 62:18-54(1978).

52. Gupta PK, Gupta RC. Pharmacology, toxicology and degrada-tion of endosulfan. A review. Toxicology 13:115-130 (1979).

53. Singh SK, Pandey RS. Effect of sub-chronic endosulfan expo-sures on plasma gonadotrophins, testosterone, testicular testos-terone and enzymes of androgen biosynthesis in rat. Indian JExp Biol 28:953-956 (1990).

54. Mueller GC, Kim U-H. Displacement of estradiol from estro-gen receptors by simple alkylphenols. Endocrinology102:1429-1435 (1978).

55. Wolff MS, Toniolo PG, Lee EW, Rivera M, Dubin N. Bloodlevels of organochlorine residues and risk of breast cancer. JNatl Cancer Inst 85:648-652 (1993).

56. Krieger N, Wolff M. Breast cancer and serum organochlorines:a prospective study among white, black, and Asian women. JNatl Cancer Inst 86:589-599 (1994).

57. Savitz D. Re: Breast cancer and serum organochlorines: aprospective study among white, black and Asian women. J NatdCancer Inst 86: 1255 (1994).

122 Environmental Health Perspectives