the biological role of estrogen receptors in cancer · physiology of the reproductive process...

Critical Reviews in Oncology/Hematology 50 (2004) 3–22

The biological role of estrogen receptors� and� in cancer

Sandra Timm Pearce, V. Craig Jordan∗

Robert H. Lurie Comprehensive Cancer Center, The Feinberg School of Medicine, Olson Pavilion, Room 8258,Northwestern University, 303 E. Chicago Avenue, Chicago, IL 60611, USA

Accepted 19 September 2003

Contents

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2. Isoforms, domains, ligand binding characteristics and expression of ER� and ER� . . . . . . . . . . . . . . . . . . . 4

3. Transcriptional activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63.1. Coregulators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63.2. Ligands: E2, antiestrogens, phytoestrogens and subtype-specific ligands. . . . . . . . . . . . . . . . . . . . . . . . 73.3. Insight into the molecular basis for ER agonism and antagonism from crystal structures. . . . . . . . . . 73.4. ER phosphorylation and ligand-independent transcriptional activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.5. Non-classical pathways. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

4. Normal physiological roles of ER� and ER� . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104.1. Tissue distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104.2. Knockout mouse studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4.2.1. Reproductive phenotypes in ER� knockout mice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114.2.2. Reproductive phenotypes in ER� knockout mice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124.2.3. Reproductive phenotypes in ER�/ER� knockout mice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124.2.4. Bone density and cardioprotection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.3. Summary of findings with ER knockouts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124.4. Consequences of an ER� mutation in a man. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5. The role of ER� and ER� in cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135.1. Breast cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5.1.1. ER� and breast cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135.1.2. Tamoxifen resistance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145.1.3. ER� and breast cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5.2. Prostate cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155.3. Colon cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155.4. Ovarian cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

6. Current status of the ER and future research directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Biographies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Abstract

The temporal and tissue-specific actions of estrogen are mediated by estrogen receptors� and�. The ERs are steroid hormone receptorsthat modulate the transcription of target genes when bound to ligand. The activity of these transcription factors is regulated by a variety offactors, including ligand binding, phosphorylation, coregulators, and the effector pathway (ERE, AP1, SP1). The end result of target gene

∗ Corresponding author. Tel.:+1-312-908-4148; fax:+1-312-908-1372.E-mail address:[email protected] (V.C. Jordan).

1040-8428/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.critrevonc.2003.09.003

4 S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 3–22

transcription is to modulate physiological processes, such as reproductive organ development and function, bone density, and unfortunatelycontribute to the growth and development of breast and endometrial cancer. The complex biological effects mediated by ER� and ER� involvecommunication between many proteins and signaling pathways. An ultimate goal of current research is to enhance the value of the separateestrogen receptors as targets for therapeutic intervention.© 2003 Elsevier Ireland Ltd. All rights reserved.

Keywords:Estrogen receptor; Cancer; Coregulator; Knockout mice

1. Introduction

Estradiol (E2) regulates the growth, differentiation, andphysiology of the reproductive process through the estrogenreceptor (ER). E2 also affects other tissues, such as bone,liver, brain and the cardiovascular system. Because of thefunctional diversity displayed by estrogens through the ER,much of the current interest in understanding the basis of ERactions at the molecular level is focused towards the goal oftherapeutic intervention[1,2].

One of the earliest studies reporting a relationship be-tween breast cancer and ovarian hormones described breasttumor regression after removal of the ovaries[3], the majorsite of estrogen production in premenopausal women. How-ever, only one in three women respond to oophorectomy[4].The explanation for these observations became clear whenthe ER was discovered[5]. In the late 1960s and early 1970s,the ER was initially used as a predictor of breast cancerresponse to endocrine ablation. Tumors that were ER richwere more likely to respond to endocrine therapy than if thetumor was ER poor[6,7]. In the mid 1970s, before adjuvanttherapy became the standard of care, the ER was viewedas a prognostic indicator after surgery, with ER-positive pa-tients responding better than ER negative patients[8]. Fromthe 1970s to the present day, the ER has evolved to bethe most effective target for breast cancer therapy. Interac-tions between E2 and the ER can be blocked using a varietyof agents. Selective estrogen receptor modulators (SERMs)such as tamoxifen and raloxifene, are competitive inhibitorsof E2 at the ER and display agonist or antagonist behaviordepending on the tissue[9]. Pure antiestrogens, exemplifiedby fulvestrant (ICI 182,780), only produce antagonist effectsand are proving to be useful in treating advanced breast can-cer [10,11]. Aromatase inhibitors, such as anastrozole, thatblock the conversion of androstenedione or testosterone toestrone and estradiol, respectively, are a particularly inter-esting new approach to breast cancer treatment as the com-pounds appear to increase efficacy and reduce side effectscompared with tamoxifen[12–15]. The optimal combina-tions and sequential orders of treatment continue to be in-vestigated in clinical trials.

Although the primary focus of research for the first 30years (1960–1990) has been on the role of steroid receptorsin reproductive functions and breast cancer, there is reason tobelieve that there are opportunities to design new moleculestargeted to novel sites dominated by one ER or the other.This is especially true since the publication of the Women’s

Health Initiative did not demonstrate an overall health benefitfor women taking hormone replacement therapy (HRT)[16].Positive aspects of HRT include a decrease in the rate ofbone density loss, a decrease in total and LDL cholesterol,and a protective effect against colon cancer. However, therisk of breast cancer is increased in HRT users[17]. Thechallenge now is to dissect the individual roles of ER� andER� as transcription factors that participate in normal andabberant physiological processes. Clearly, the goal will be amenu of multifunctional medicines that can be used singlyor in combination to treat and prevent a range of diseasesassociated with menopause or reproductive function.

2. Isoforms, domains, ligand binding characteristicsand expression of ER� and ER�

The therapeutic targets estrogen receptors� (ER�) and�(ER�) are members of the nuclear receptor superfamily oftranscription factors. Other members of this family includethyroid receptor, Vitamin D receptor, retinoic acid receptor,and other steroid receptors such as the glucocorticoid recep-tor, androgen receptor, progesterone receptor and mineralo-corticoid receptor.

ER� was the first estrogen receptor cloned and it wasisolated from MCF-7 human breast cancer cells in the late1980s[18–20]. In accordance with its role as a transcriptionfactor, this 66 kDa ER� localizes primarily to the nucleus.A 46 kDa isoform (hER�46) that lacks the first 173 aminoacids of the 66 kDa form of ER� has also been preliminarilycharacterized[21]. In addition, several ER� splicing variantshave been described[22,23], but whether they are expressedas proteins that have a biological function remains unknown.

Ten years later, ER� was cloned from rat prostate us-ing degenerate PCR primers[24]. Mouse[25] and human[26–28] forms of ER� have also been cloned. The humanER� gene is located on chromosome 6 and the ER� gene ison chromosome 14, demonstrating that they are in fact en-coded by separate genes and are distinct[27]. A variety ofER� mRNA isoforms have been described in humans, pri-mates, rats and mice[29], but the 530 amino acid form ofER� [28] is considered to be the wild type, full length hu-man ER�. Because the functional significance and expres-sion of the various ER� isoforms are unclear, the reader isreferred to recent reviews[29,30].

Human ER� and ER� (long form) share common struc-tural domains, which are designated AF (Fig. 1). The A/B

S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 3–22 5

Fig. 1. Structure and homology between human ER� and the long formof ER�. The domains A-F are depicted, as well as the percent iden-tity between the individual domains at the amino acid level. Activationfunctions 1 (AF1) and 2 (AF2) are also indicated. Adapted from[28].

domain contains activation function 1 (AF1), a constitutiveactivation function contributing to the transcriptional activ-ity of the ER. This domain is one of the least conserveddomains between ER� and ER�, exhibiting only a 30%identity. Based on functional studies, ER� has been shownto lack AF1 activity[31]. The DNA binding domain, or Cdomain, is the most highly conserved region between ER�and ER�, with 96% identity. This allows both receptorsto bind to similar target sites. The D domain, or hinge re-gion, is not well conserved (30%) between the receptorsand it contains the nuclear localization signal. Finally, theE/F region encompasses the ligand binding domain (LBD),a coregulator binding surface, the dimerization domain, a

Fig. 2. Distribution of ER� and ER� in the human body. Adapted from[34].

Table 1The relative binding affinity of various ligands for ER� and ER�

Compound Relative binding affinitya

ER� ER�

E2 100 1004-OHT 178 3391C1164,384 85 166DES 468 295Genistein 5 36

a The relative binding affinity was calculated as a ratio of concen-trations of E2 or competitor required to reduce the specific radioligandbinding by 50%. Adapted from[32].

second nuclear localization signal, and activation function 2(AF2). In contrast to AF1, AF2 is a ligand-dependent acti-vation function. The E/F domains of ER� and ER� exhibita sequence identity of 53%.

Despite the 53% sequence identity between ER� andER� in the ligand binding domain, the two receptors ex-hibit subtle differences in ligand binding specificity. AKDof 0.6 nM for ER� and 0.24 nM for ER� for the ligand16�-iodo-E2 was determined using saturation binding as-says, which is similar to the range of E2 binding to ERs(0.1–1 nM) in various systems[32]. Additional studies us-ing E2 showed that theKD for ER� was 0.05 nM and forER� was 0.07 nM[33]. Despite the slight differences in the


affinity of ER� and ER� for E2, both receptors are consid-ered to have a similar affinity for E2. However, differenceswere observed for other ligands, such as antiestrogens andphytoestrogens.

A wide variety of structurally distinct compoundsbind to the ER with differing affinities. Certain ligands act asER agonists, and these include the natural ligand E2, as wellas the synthetic estrogen diethylstilbestrol (DES). Certainphytoestrogens, which are environmental compounds pro-duced by plants, can also be estrogenic. Genistein, presentin soya beans and soy products, is a widely utilized phytoe-strogen. Other compounds, such as ICI 182,780, are receptorantagonists. A final group of mixed agonists and antago-nists are comprised of the SERMs and examples includetamoxifen and raloxifene. The relative binding affinity(Table 1) of ER� for various ligands are: DES(468) >

4-OHT(178) > E2(100) > ICI 164, 384(85) > genistein(5).In contrast, ER� displayed the following relative bindingaffinities: 4-OHT(339) > DES(295) > ICI 164, 384(166) >

E2(100) > genistein(36) [32]. These differencescould result in functional consequences for a receptorsubtype.

ER ligands interact with ER subtypes in various partsof the human body (Fig. 2). The abundance and distribu-tion of the receptors will, in part, determine whether a lig-and will have a particular effect. Using RT-PCR, Northernblot analysis, immunohistochemistry and in situ hybridiza-tion techniques, ER� and ER� are known to be localized inthe breast, brain, cardiovascular system, urogenital tract andbone[27,32,34,35]. ER� is the main ER subtype in the liver,whereas ER� is the main ER in the colon. ER� and ER� mayalso localize to distinct cellular subtypes within each tissue.For example, within the ovary, ER� is largely present in thethecal and interstitial cells, whereas ER� is predominantlyin the granulosa cells[36,37]. In the prostate, ER� local-izes to the epithelium, whereas ER� localizes to the stroma[38].

3. Transcriptional activity

The transcriptional activity of the ER is mediated by AF1and AF2 (Fig. 1) [39–42] and these regions were largelydelineated using mutational studies. The activity of AF1and AF2 differs depending on the cellular environment andpromoter context[43]. In some cells, either AF1 or AF2 isdominant, and in others, both activation functions synergize[44]. In addition, AF1 and AF2 are differentially regulatedby ligand. E2 is an agonist regardless of whether AF1 orAF2 is dominant. The pure antiestrogen ICI 164,384 blocksboth AF1 and AF2, affects dimerization[45] and targets theER for degradation[46]. Tamoxifen acts by blocking AF2activity so it is an antagonist in cells where AF2 is dominantand a partial agonist where AF1 is dominant. ER� activityis mediated by AF2, since ER� does not contain AF1. Cellstherefore have the ability to distinguish between different

ER–ligand complexes and translate this information into thecoordinated regulation of gene transcription.

3.1. Coregulators

The initiation of transcription is complex and requiresthe interaction of many proteins at a target gene promoter.Transcriptional activation by the ER requires the recruit-ment of transcriptional regulators, such as general transcrip-tion factors, coactivators, corepressors, cointegrators, his-tone acetyltransferases, and histone deacetylases. (reviewedin [47–49]). All of these regulators interact to affect tran-scription and the accessibility of target gene promoters.

Coactivators are proteins that enhance transcription. Thecontact between coactivators and the ER is made through theLXXLL motif present in the coactivator[50], although thesite on the ER required for this interaction varies. Coactiva-tors include steroid receptor coactivator 1 (SRC-1), SRC-2and SRC-3, which are members of the p160 family. p300and CREB-binding protein (CBP) are cointegrators, in thatthey do not themselves bind DNA, but are recruited to pro-moters by other transcription factors, such as SRC-1. Core-pressors decrease transcription and include nuclear receptorcorepressor (NCoR) and silencing mediator for retinoid andthyroid hormone receptor (SMRT).

Local chromatin structure is remodeled to allow for genetranscription[47–49]. Chromatin remodeling factors includeATP-dependent nucleosome remodeling complexes and pro-teins that contain acetyltransferase activity. Histone acety-lation correlates with transcription, whereas deacetylationcorrelates with gene repression. p300/CBP-associated fac-tor (PCAF), p300/CBP, SRC-1 and SRC-3 contain intrinsicacetyltransferase activity. In contrast, corepressors do notcontain histone deacetylase (HDAC) activity, but they re-cruit other proteins that have HDAC activity.

Although the majority of coregulators interact with mul-tiple members of the nuclear receptor family, there are ex-amples of coregulators that interact exclusively with the ER.In a screen to identify proteins that interact with the E/F do-main of the dominant negative L540Q mutant ER, the proteinrepressor of estrogen receptor activity (REA) was isolated[51]. REA is a selective ER� and ER� coregulator that en-hances the inhibitory effectiveness of the L540Q ER and ofantiestrogens. In contrast to other corepressors that interactwith the unliganded receptor, REA preferentially interactswith the liganded ER. In addition, REA binds to the L540Qmutant ER and antiestrogen-liganded ER more than the wildtype ER. REA also competes with the coactivator SRC-1for modulation of ER transcriptional activity. These studiessuggest that when cellular REA levels are high, antiestrogenaction will be enhanced[52].

Coactivators can also interact preferentially with a partic-ular activation function region. For example, p68 RNA he-licase is a coactivator specific for the ER� AF1 region[53].p68 binds CBP, so p68 may serve as a bridge to associateAF1 with AF2 coactivators. p68 enhanced the transcriptional


activity of the 4-OHT–ER� complex and the phosphoryla-tion of S118 of ER� was required for the ability of p68 toenhance transcription.

In addition to interacting with both ER� and ER� or aparticular activation function, coactivators can interact se-lectively with ER� or ER�. For example, SRC-3 enhancesER� and progesterone receptor (PR) stimulated transcrip-tion, but has no effect on ER�-mediated transcription[54].

Therefore, coregulators provide an additional layer ofspecificity and regulation to the transcriptional activity ofthe ER. In addition to being a general ER coregulator, thiscould be accomplished by targeting ER� or ER� specifi-cally, or interacting with AF1 or AF2. The discovery of newcoregulators will continue to define the selective role of ei-ther ER� or ER�.

3.2. Ligands: E2, antiestrogens, phytoestrogens andsubtype-specific ligands

Initial transcriptional activation studies using E2 were per-formed using ER� cloned from rat prostate. These studiesshowed that an ERE-luciferase reporter was activated in thepresence of E2 in CHO cells transfected with ER� [24].Studies in HepG2 and Hela cells showed that both AF1 andAF2 contribute to the activity of ER�, but the individualcontributions depend on the cell context[31]. In contrast,AF2 mediates the transcriptional activity of ER� to E2, sinceAF1 is inactive and inhibits the transcriptional activity me-diated by AF2. In the presence of sub-saturating amounts ofE2, ER� activity was actually suppressed by ER�, suggest-ing that the relative amounts of ER� and ER� can affect E2activity [31].

The agonist activity of the antiestrogen tamoxifen hasbeen shown to be dependent on cell type, promoter context,and ER subtype[55]. The agonist activity of tamoxifen ap-pears to be mediated through ER�, because no agonism isobserved with ER� [31,55]. In fact, the addition of ER� toER� and tamoxifen inhibited the agonist activity of tamox-ifen [31]. The lack of tamoxifen agonism at ER� is likelyto result from differences in the A/B region, and AF1 inparticular, between ER� and ER�.

Replacing the A/B domain of ER� with the A/B domainof ER� (ER�/�) increased the transcriptional activity of thechimera in response to E2 compared to that observed withER� alone[56]. In addition, the ER�/� chimera displayeda transcriptional response to tamoxifen, which was not ob-served with ER�. These studies further suggested the impor-tance of ER� AF1 in mediating the agonist effect of tamox-ifen. Therefore, differences in the A/B region between ER�and ER� are responsible for the cell, promoter and ligandspecificity displayed by the estrogen receptors.

Phytoestrogens are plant-derived compounds that are con-sumed in the diet. They contain inherent estrogenic activityor are converted to estrogenic compounds by bacteria in thegut. The first class of phytoestrogens is the isoflavonoids,which are present in soybean products, some fruits and veg-

etables, and red clover. Genistein, dadzein and glycitein arethe main dietary-derived isoflavones[57]. Many of thesecompounds have a greater affinity for one ER subtype. Forexample, the isoflavone genistein has a 20-fold greater bind-ing affinity for ER� compared to ER�, but it activates tran-scription through ER� and ER� [33]. The second class ofphytoestrogens is the lignans, such as enterodiol and en-terolactone, which are present in whole grain cereals, seeds,berries and nuts[58].

Interest in phytoestrogens and their potential role in theprevention of breast, prostate and colon cancer is a result of ahigher incidence of these cancers in the western world com-pared to Asian populations[59]. The Asian diet is largelyvegetarian or semivegetarian, which has a higher proportionof phytoestrogens compared to western diets that includemore animal protein and fat. In general, phytoestrogens mayprovide some protection against breast, prostate and coloncancer (reviewed in[57,58]), but more human studies areneeded before a definitive conclusion can be drawn.

All of the ER ligands described to date can bind to bothER� and ER�, but with differing affinities. The develop-ment of agonists and antagonists that are selective for eitherER� or ER� would be useful tools to analyze the individ-ual role of ER� or ER�. For example, theR,R-enantiomerof 5,11-cis-diethyl-5,6,11,12-tetrahydrochrysene-2,8-diol(THC) acts as an ER� agonist and an ER� antagonistand has a sixfold greater affinity for ER� over ER� [60].Another example is methyl-piperidino-pyrazole (MPP),a basic side chain pyrazole that has a 200-fold bindingselectivity for ER� over ER� and is an ER� selective an-tagonist[61]. The structure function relationships of ER�-and ER�-specific ligands are reviewed in detail elsewhere[1,2].

3.3. Insight into the molecular basis for ER agonism andantagonism from crystal structures

The crystal structures of ligands complexed with the ERhave provided invaluable insight into the structural basis ofreceptor agonism and antagonism (Fig. 3). The first stud-ies of the human ER� LBD complexed with E2 and ralox-ifene indicated that although E2 and raloxifene bind at thesame site within ER�, structural differences resulted. Helix12 (Fig. 3A, green) is positioned over the ligand bindingpocket in the ER�−E2 complex[62,63], thereby generat-ing a functional AF2 that is able to interact with coactiva-tors. In contrast, because the side chain of raloxifene is toolong to fit within the binding pocket, it displaces helix 12(Fig. 3C, compare arrows in 3A and 3C)[62,63]. This pre-vents the formation of a competent AF2 region. Subsequentstudies directly analyzed the interaction of DES (Fig. 3B),4-hydroxytamoxifen (4-OHT, the active metabolite of ta-moxifen [64]) and the coactivator GRIP1 with human ER�[65]. Like raloxifene, the bulky side chain of 4-OHT dis-places helix 12 (Fig. 3D, arrow) and places it in a positionwhere it mimics bound coactivator.


Fig. 3. Modulation of ER� and ER� by ligand and helix 12 positioning. The crystal structure of human ER� in complex with E2 [63], DES[65], raloxifene[63] and 4-OHT[65] have been solved. The crystal structure of human ER� has been studied in complex with genistein[63,66] and rat ER� has beenstudied in complex with ICI 164,384[67] and raloxifene[66]. The individual helices are numbered and the ligand is depicted as a space filling model. Foreach ligand, the structure on the left represents a frontal view and the structure on the right represents a 90o rotation to the left. The chemical structureof the ligands is also shown. Reprinted from[63,65,67]with permission from Elsevier and from[66] with permission from Oxford University Press.

The first structural description of ER� was in complexwith raloxifene or genistein[66]. The structure of the ER�LBD was similar to the previously reported ER� LBD struc-ture. In addition, the human ER�-genistein (Fig. 3E) and therat ER�–raloxifene (Fig. 3G) complexes are quite similar.As in the ER�-4-OHT structure, the side chain of raloxifenedisplaces helix 12 and prevents it from sealing the ligandin the ligand binding pocket (Fig. 3G, arrow). In contrast,genistein binds to ER� in a manner similar to E2. The maindifference is that instead of helix 12 assuming the typicalagonist position that E2 induces, helix 12 is in a more antag-onist position (Fig. 3E, arrow). Genistein is a partial agonist

at ER� and coactivators must displace helix 12 into a moreagonist conformation before activating transcription.

The only crystal structure of the pure antiestrogen ICI164,384 with an estrogen receptor is with the rat ER� LBD(Fig. 3F) [67]. The side chain of ICI 164,384 is longer thanthat of raloxifene. For the side chain of ICI 164,384 toachieve the same position that the side chain of raloxifeneoccupies, the steroidal core is flipped 180o around its longest(hydroxy-to-hydroxyl) axis. The position of the side chainexposes a large hydrophobic patch on the surface becausethe side chain of ICI 164,384 binds along the coactivatorrecruitment site. This abolishes the interaction of helix 12


with the LBD such that it is disordered. Therefore, helix 12cannot move into position to form a coactivator recruitmentsite. This overall conformation could favor the recruitmentof corepressors and resemble misfolded or denatured pro-teins, thereby targeting the ER for degradation.

In summary, when an agonist binds to the ER, a conforma-tional change occurs that forms the AF2 coactivator bindingsite. However, the large side chain of antagonists such as ta-moxifen, raloxifene, or ICI 164,384 protrudes from the bind-ing cavity and displaces helix 12, thereby disrupting AF2.The molecular events resulting from ligand binding subse-quently translate into agonism or antagonism at the ER (forreview, see[63]). However, it is important to note that onlythe ER LBD has been used to generate these crystal struc-tures, so the potential structural importance and interactionof the remaining part of the ER with the LBD has yet to bedetermined. In addition, a second binding site has been iden-tified in the LBD of ER� and ER� [68]. Although multipletetrahydrochrysene derivatives were docked into this secondsite, the classical steroid binding site was still the preferredsite. Future studies could determine whether these multiplebinding sites have consequences for the activity of the ER.

3.4. ER phosphorylation and ligand-independenttranscriptional activity

In contrast to the ligand-mediated transcriptional activityor genomic effects described above, the activity of the ERcan also involve ligand-independent or non-genomic effects.This is based on reports showing that many effects inducedby E2 occur within a short time frame of seconds to min-utes, which is faster than transcriptional events. These rapideffects may be mediated in part by plasma membrane asso-ciated forms of the ER[69–71]. Interestingly, ER� and ER�can both localize to the membrane[72].

The ligand-independent activity of the ER is a result ofphosphorylation of the ER (Fig. 4) and creates cross-talkbetween the ER and other signaling pathways[73,74]. Phos-phorylation of the ER is largely on serine residues in the

Fig. 4. Kinases and ligands induce ER phosphorylation at specific sites. Phosphorylation of the ER occurs within AF1 at S104/S106, S118 and S167 bya variety of ligands and kinases as indicated. S236, in the DNA binding domain, is phosphorylated by PKA and Y537 is phosphorylated by src familykinases.

AF1 region of ER�. The first reports showed that E2 inducedphosphorylation only on serine residues in MCF-7 cells[75].In addition, using a transient transfection system in COS-1cells, the ER was shown to be phosphorylated only on ser-ine residues by E2, as well as other ligands such as 4-OHTand ICI 164,384[76]. More specifically, E2, 4-OHT and ICI164,384 induced ER phosphorylation at S118[77]. Furtherevidence indicated that S118 was a major site of phospho-rylation by E2 and phorbol ester (TPA) in COS-1 cells[78]and that S118 is required for epidermal growth factor (EGF)activation of the ER via MAP kinase[79].

However, controversy exits as to which kinase phospho-rylates S118 and the possibility exists that multiple kinasescarry out this phosphorylation. S118 is a target of MAP ki-nase in vitro and in response to EGF or insulin-like growthfactor (IGF) treatment in vivo[80]. Other reports have shownthat MAP kinase does not phosphorylate S118 in MCF-7cells in response to E2, but EGF and PMA treatment resultin phosphorylation of S118 via MAP kinase[81]. Furtherinformation suggests that MAP kinase phosphorylates S118independent of ligand, whereas Cdk7 phosphorylates S118in response to E2 in COS-1 cells[82].

Serine residues in ER� other than S118 are also phos-phorylated. In response to E2, S167 is the major phosphory-lation site in recombinant ER expressed in Sf9 insect cellsand MCF-7 cells and it is phosphorylated by casein ki-nase II[83]. Upon EGF or phorbol myristate acetate stim-ulation, S167 is a target of pp90rsk1, which is phosphory-lated by MAP kinase[84]. Phosphorylation of S167 hasalso been implicated in the phosphatidylinositol-3-OH ki-nase (PI(3)K)/Akt pathway. E2 stimulates ER� binding tothe p85� subunit of PI(3)K in endothelial cells, whereas ER�does not exhibit any interaction with PI(3)K[85]. However,in HEK293 and MCF-7 cells, ER� is constitutively associ-ated with p85� because this interaction is not affected byE2 [86]. The end result is that Akt phosphorylates S167 ofER� and the consensus site is not present in ER� [87].

The phosphorylation of S104 and S106 is mediated by thecyclin A-CDK2 complex in U-2 OS human osteosarcoma


cells [88]. S236 in the DNA binding domain, is phosphory-lated by protein kinase A and this phosphorylation regulatesdimerization[89].

Tyrosine phosphorylation has been detected at Y537 inMCF-7 and Sf9 cells[90] and p60c-src and p56lck mediatethis phosphorylation event. Y537 is not phosphorylated byE2 treatment, indicating that it is a basal phosphorylationsite.

Although it is clear that many studies have focused on therole of phosphorylation of ER�, few have studied ER�. Therole of phosphorylation in the interaction between ER� andcoactivators indicated that the interaction between mouseER� and SRC-1 increased in the presence of E2 [25]. Phos-phorylation of the unliganded receptor by MAP kinase atS106 and S124 (S118 in human) in AF1 enhanced the re-cruitment of SRC-1 and therefore the activity of ER� [91].The role of phosphorylation in the activity of ER� remainsto be elucidated.

Activation of ER� via phosphorylation at multiple sites(S104, S106, S118, S167, S236, Y537) by multiple kinasesis important because of the interaction between growth fac-tor signaling and the ER. Increased growth factor signalingmay account for the loss of E2 dependence, thereby produc-ing antiestrogen resistant tumors. In addition, an associationhas been observed between elevated MAP kinase phospho-rylation/activity and a poor response to endocrine therapy inbreast cancer patients[92]. Although the precise relationshipbetween ER phosphorylation and clinical outcome remainsto be elucidated, the ER phosphorylation state has the po-tential to be a predictive biomarker and intervention target.

3.5. Non-classical pathways

The classical scheme of ER action involves ligand bind-ing to the ER, dissociation of heat shock proteins from theER and receptor dimerization. The ER dimer then interactswith coregulatory proteins, binds to DNA sequences termedestrogen response elements (ERE) that are located in theregulatory regions of responsive target genes, and transcrip-tion is activated.

However, the ER can also mediate transcription via teth-ered interactions through protein–protein interactions at AP1(reviewed in[93]) and Sp1 sites. In Hela cells transfectedwith ER� and an AP1 reporter, E2, DES, raloxifene, tamox-ifen and ICI 164,384 stimulated reporter activity to varyingdegrees[94]. The amount of stimulation varied dependingon the cell type[95]. In contrast, ER� activated the AP1reporter in the presence of raloxifene, tamoxifen and ICI164,384, but not with E2 and DES. ER� and ER� thereforerespond differently to estrogens and antiestrogens at AP1sites. The regions of the ER that are required for the stimu-lation of AP1-mediated transcription vary depending on thecell type and ligand[95–97].

The ER also activates transcription of target genes throughER–Sp1 protein interactions at GC-rich promoter elements.E2-responsive genes that activate transcription through non-

consensus ERE half sites and GC-rich motifs include c-myc,creatine kinase B, cathepsin D, heat shock protein 27 andtransforming growth factor� [98]. ER� and ER� havebeen shown to bind to the C-terminal domain of the Sp1protein [99]. Transient transfections of MCF-7, Hela andMDA-MB-231 cells with a Sp1 reporter and ER� or ER�showed varying patterns of activation by estrogen and antie-strogens. In MCF-7 and MDA-MB-231 cells, E2, 4-OHTand ICI 182,780 activated Sp1 through ER�. In contrast, nochanges were observed in Hela cells with any ligand. 4-OHTactivated Sp1 through ER� in MCF-7 cells, but no changeswere observed by any ligand in MDA-MB-231 cells[99].All of the ligands decreased Sp1 reporter activity in Helacells in the presence of ER�. Additional results[99] sug-gested that the relative amounts of ER� and ER� present ina cell influence Sp1 activity. Amino acids 79–117 in AF1are important for the interaction of ER� with Sp1, and thisregion could mediate an association with other proteins thatare important for the ER�/Sp1 mechanism.

In summary, cell-specific regulation occurs as a result ofmultiple factors, including coregulator expression and re-cruitment, the ratio of ER� to ER�, the nature of the lig-and, ER phosphorylation and the pathway activated (ERE,Sp1 or AP1). The activity of ER� and ER� is also compli-cated by the fact that they can form functional homo andheterodimers[28,31,100,101], which may activate differenttarget genes. The inherent structural differences in the A/Bdomain between ER� and ER�, where ER� lacks a func-tional AF1 domain, may result in a large effect on the ac-tivation profiles of target genes, especially when coregula-tors that interact preferentially with the AF1 domain areconsidered.

4. Normal physiological roles of ER� and ER�

4.1. Tissue distribution

Analysis of the tissue distribution of ER� and ER� pro-vides insight into the potential for targeting specific tissues.The relative distribution of ER� and ER� mRNA was ini-tially determined in rat tissues using RT-PCR[32]. ER�mRNA was highly expressed in epididymis, testis, pituitarygland, ovary, uterus, kidney and adrenal. Moderate amountswere also present in the prostate gland, bladder, liver, thy-mus and heart. Highest amounts of ER� mRNA were de-tected in the prostate gland and ovary. In the rat ovary, ER�is the predominant ER in the granulosa cells, whereas ER�is largely present in the thecal and interstitial cells[36,37].Uterus, bladder, lung and testis showed intermediate levelsof ER�, whereas low but detectable levels of ER� were ob-served in epididymis, the pituitary gland, thymus, variousbrain sections and spinal cord. Target tissues with higher ex-pression levels of ERs are predicted to be more affected byER ligands. The expression in rat is similar to that observedin humans (seeSection 2) ([35] and references therein).


4.2. Knockout mouse studies

Transgenic mice are valuable experimental models to elu-cidate gene functions. Knockout mice, where genes of in-terest have been deleted, provide a basic insight into thenormal functions of genes during development and at ma-turity. The classical scheme of reproductive organ develop-ment is that female reproductive tract development is thedefault pathway and that estrogens are not required for theinitial differentiation and development of the female repro-ductive system. In contrast, testosterone is essential for theproper development of male structures. The potential role ofestrogen in male development has remained unclear. Knock-out mouse models have been utilized to analyze the role ofER� and ER� in the general development and physiologyof the mouse (Table 2). These models include ER� knock-out mice (�ERKO), ER� knockout mice (�ERKO) andboth ER� and ER� knockout mice (��ERKO) (reviewedin [102–105]). It is interesting to note that a loss of eitherof ER� and/or ER� is not lethal, and the mice survive toadulthood.

4.2.1. Reproductive phenotypes in ERα knockout miceER� knockout mice show no abnormal external pheno-

types. However, the most striking phenotypes occur in thetissues that predominantly express ER�, such as the uterusand mammary gland, but defects are also observed in theovary and in sexual behavior. One role of the ovary is to actas an endocrine organ by providing sex steroids to the fe-male and is essential for proper reproductive functions. Theovaries in�ERKO mice contained cystic and hemorrhagicfollicles that contain no corpora lutea and few granulosa

Table 2Phenotypes of ER knockout mice

�ERKO no ER�, ER� dominant �ERKO no ER�, ER� dominant ��ERKO no ER� or ER�

Female Male Female Male Female Male

Infertile Reduced fertility Reduced fertility Fertile Infertile InfertileReproductive

tractHypoplasticuterus; cystic andhemorrhagicfollicles; nocorpora lutea; fewgranulosa cells inovary

Low sperm count;low spermmotility; lowtestis weight

Normal uterus;many early atreticfollicles; Fewercorpora lutea inovary

Epithelialhyperplasia incollecting duct ofprostate andbladder wall

Hypoplasticuterus; developstructures similarto maleseminiferoustubules (sexreversal) in ovary

Low spermnumber; lowsperm motility

Mammaryglands

Develop to anewbornstructure; nopubertaldevelopment

N/A Normal N/A N/A

Sexualbehavior

No lordosis; noreceptivity

Normal mounts;reducedintro-missions;rarely ejaculates

Normal Normal No mounts,intro-missions orejaculation

Bone Decreaseddensity, diameterand length

Decreaseddensity, diameterand length

Normal Normal

N/A: not applicable.

cells[106]. Adult females have hypoplastic uteri and showedno responses to estrogen, such as increases in uterine wetweight, hyperemia, or the alteration of vaginal epithelial cellmorphology[106]. Females exhibit little sexual behavior inthat they display no lordosis posture or receptivity, indicat-ing a lack of estrogen responsiveness in the central nervoussystem[106]. As a result of these phenotypes, female micelacking ER� are infertile.

The mammary glands of the adult�ERKO look essen-tially like those of a female mouse before puberty, indicatingthat rudimentary mammary glands can develop independentof estrogen and ER�, but a fully differentiated gland re-quires ER� [107]. �ERKO mice were further used to studythe potential protective effect of dietary genistein on mam-mary tumor development. The rationale for this study is thesuggestion that genistein could be protective against breasttumors. Because genistein has a greater binding affinity forER� when compared to ER�, genistein is predicted to actthrough an ER� pathway.�ERKO mice or mice containingwild type ER� were fed genistein and treated with dimethyl-benz[a]anthracene. Tumors formed in the mice with wildtype ER�, whereas no tumors were present in the�ERKOmice[108]. This indicated that genistein did not provide pro-tection against tumor formation in the wild type mice andthat it could actually result in tumor formation. In situationswhere ER� is dominant, such as in the�ERKO mice, ER�(genistein) is protective against breast tumors, but when ER�is present, such as in the wild type mice, ER� (genistein)is no longer protective. These data, combined with the highexpression levels of ER� in the mammary gland, point to-ward ER� being the main mediator of estrogen action in themammary gland.


Male�ERKO mice show decreased sperm motility[109],a low sperm count, and low testis weight, resulting in re-duced male fertility[106]. In terms of sexual behavior, male�ERKO mice show normal levels of mounts, but also re-duced intromissions, and rare ejaculations[110], which alsocontribute to the reduction in fertility.

4.2.2. Reproductive phenotypes in ERβ knockout miceER� knockout mice have also provided additional infor-

mation as to the function of ERs in the mouse. The mostobvious phenotype occurs in the ovary, which is a tissue thatcontains the greatest expression of ER�. Ovaries in femaleslacking ER� have more early atretic follicles and fewer cor-pora lutea when compared to wild type females[111]. Fur-ther experiments indicate a partial arrest of follicular devel-opment and a decrease in the frequency of follicular mat-uration [111]. These female mice show normal sexual be-havior [112] and normal mammary gland structure. In ad-dition, ER� knockout females have fewer litters and fewerpups per litter. Therefore, ER� knockout mice show reducedfertility.

In contrast to ER� knockout females, ER� knockoutmales are fertile and show normal sexual behavior[112].However, older males show epithelial hyperplasia in thecollecting duct of the prostate and the bladder wall[111].�ERKO mice were developed by another group that exhib-ited many of the same phenotypes, except that no prostateor bladder hyperplasia was observed[113]. The reasonfor these differences is unknown.�ERKO mice also dis-play abnormalities in the brain, such as regional neuronalhypocellularity, which progressed to a degeneration ofneuronal cell bodies with age[114].

4.2.3. Reproductive phenotypes in ERα/ERβ knockout miceTo create a complete picture of ER function in male and

female development, mice with both ER� and ER� knock-outs (��ERKO) were developed. As with the ER� and ER�single knockouts,��ERKO mice survive to adulthood andexhibit no abnormal external phenotypes[115].

Young ��ERKO females show proper differentiationof the uterus, vagina and cervix. However, once the micereach 2.5–7 months of age, uterine hypoplasia is ob-served. Because this phenomenon is observed in both�ERKO and ��ERKO mice, it is a hallmark of ER�loss in the uterus. The most interesting phenotype oc-curs in the adult��ERKO ovary, where structures rem-iniscent of male seminiferous tubules of the testis wereobserved[115]. These structures were not present in theprepubertal ovaries. This is the first example of sex rever-sal in an adult mouse gonad because the female ovariancells are able to re-differentiate to a male Sertoli cellphenotype.

��ERKO males are infertile and show an 80% reductionin epididymal sperm number and a 5% decrease in spermmotility [115]. They also show no components of sexualbehaviors[116].

4.2.4. Bone density and cardioprotectionAn established association exists between the declining

levels of estrogen at menopause and the development ofosteoporosis, implicating estrogen in the maintenance ofbone mass.�ERKO females and males showed decreasesin femoral length and diameter as well as density[105].�ERKO mice had normal bone length and density[102],emphasizing the important role for ER� in bone.

Cardiovascular disease in women increases aftermenopause. As a result of the suggestion that estrogen mightbe cardioprotective, there has been an interest in studyingthe role of the ERs in the cardiovascular system. Indeed, thepublication of the negative results of the Women’s HealthInitiative [16] has increased interest in this aspect of femalephysiology to discover mechanisms of physiological bene-fit that can be separated from disadvantageous side effectssuch as fatalities from cardiovascular complications.

Models of carotid injury have been developed in knock-out mice to study the individual contribution of ER� andER� in cardioprotection. In wild type and�ERKO mice,E2 treatment inhibited the increase in carotid medial ves-sel wall area and vascular smooth muscle cell proliferationnormally observed in injured carotid arteries[117]. BecauseE2 inhibits markers of vascular injury in the�ERKO mice,ER� is clearly not required for this process. Similar resultswere observed in the�ERKO mice[118], suggesting thatER� and ER� are individually redundant in mediating thevasoprotective effects of E2. In contrast, E2 did not protectagainst the increase in carotid medial wall area after injuryin ��ERKO mice, but did inhibit vascular smooth musclecell proliferation[119]. Unfortunately, the specific batch of��ERKO mice exhibited uterine weight increases in thepresence of E2, suggesting residual ER� activity. In fact, ithas been suggested that the original�ERKO mice[106] ex-press a smaller ER transcript[120] that may have some func-tional activity[119,121]. As a result, these data[119] shouldbe interpreted with caution. Nevertheless, overall studies inER� or ER� knockout mice suggest that either ER� or ER�can mediate the E2-induced protection of vascular injury.

It is important to consider whether E2 also has a role inmaintaining the vasculature by mediating increases in ni-tric oxide synthase, which ultimately results in vasodila-tion. Wild type male mice have increased basal levels of en-dothelial nitric oxide in the aorta when compared with male�ERKO mice[122]. �ERKO mice also display increases inL-type Ca2+ channels in the heart, which could lead to ab-normalities in cardiac excitability[123]. These studies sug-gest that ER� is involved in cardiac modulation through theregulation of nitric oxide synthesis as well as cardiac Ca2+channel expression.

4.3. Summary of findings with ER knockouts

Overall, these transgenic mice studies provide insight intothe role of estrogen receptors in the development and func-tion of reproductive structures. In the case of the male, ERs


are not necessary for the development of reproductive struc-tures. However, ER� is required for male fertility, becauseER� loss results in a low sperm count and a decrease insperm motility. As would be expected, ER loss has moreof an impact on the female. Surprisingly, early differentia-tion of the reproductive tract occurs in the absence of ER�and ER�. Later in life, both ERs are necessary for properovarian function, whereas ER� is critical for uterine physi-ology. ER�, but not ER�, is required for the proper develop-ment of the mammary gland as well as the maintenance ofbone. In addition, ER� and ER� may play redundant rolesin E2-mediated cardioprotection.

4.4. Consequences of an ERα mutation in a man

Although only one example of an ER germ-line defectin humans has been described thus far[124], comparisonscan be made to the phenotypes observed in knockout mice.A man was shown to have a cytosine to thymine substi-tution at codon 157 in ER�, resulting in a premature stopcodon. The translated protein would therefore be truncatedand lack the DNA and hormone binding regions. The pa-tient presented with osteoporosis, unfused epiphyses and el-evated serum estrogen, among other phenotypes. These re-sults provide evidence for the critical role of estrogen inbone development and mineralization as well as epiphysealmaturation in males. Cardiovascular function was also im-paired, in that signs of early atherosclerosis were detected[125] and flow-induced dilation of the brachial artery wasimpaired[126].

5. The role of ER� and ER� in cancer

The analysis of knockout mice has provided a frameworkin which to study the potential functions of ER� and ER�in human target tissues. Phenotypes of�ERKO mice havepointed toward the importance of ER� in the uterus andmammary gland of females. In addition,�ERKO mice havesuggested an important function for ER� in the ovary infemales and in the prostate gland in males. The laboratorystudies in mice naturally advance the study of the complexrole of the individual ERs in human cancer.

5.1. Breast cancer

5.1.1. ERα and breast cancerThe ER is an important target to develop drugs for the

treatment and prevention of breast cancer[127]. The inter-action of estrogen with the ER can result in increased prolif-eration of target cells so the rationale for endocrine therapyis to block the interaction of estrogen with the ER. This goalcan be accomplished by blocking the production of estro-gen by ovariectomy, or inhibiting the conversion of steroidalprecursors to estrogen using aromatase inhibitors. The ERcan also be targeted directly using SERMs such as tamox-

Table 3ER and PR status and response to endocrine therapy

Status Response

Number (%)

ER+ 303/571 53ER+PR− 80/248 32ER+PR+ 223/323 69ER− 32/239 13PR+ 233/354 66ER−PR- 22/208 11ER−PR+ 10/31 32PR− 102/456 22

Adapted from[128].

ifen and raloxifene as competitive inhibitors of estrogen ac-tion, or by the removal and degradation of the ER by pureantiestrogens such as ICI 182,780 (fulvestrant).

Endocrine manipulations are among the least toxic andmost effective therapies for the treatment of hormone re-sponsive breast cancers. In the clinic, factors such as ER, PRand nodal status have historically predicted response to en-docrine therapy (Table 3). Patients that are ER+ show a 53%objective response rate to endocrine therapy, and this canbe divided into 69% for ER+PR+ and 32% for ER+PR−[128]. As expected, 13% of ER− patients respond to therapy,with 32% of the ER−PR+ and 11% of the ER−PR− groupexhibiting a response. 66% of patients with PR+ tumorswill respond, versus 22% with PR− tumors. The measure-ment of receptor status has changed from the ligand bindingassay to immunohistochemical methods. Nodal positivity isanother predictor of response to endocrine therapy. Tumorsthat contain positive nodes correlate with a lower diseasefree survival (DFS)[129]. If a tumor population is nodepositive or node negative, the presence of the ER correlateswith better DFS. Current treatment strategies have shownthat 5 years of adjuvant tamoxifen treatment is beneficial inpre- and postmenopausal women with ER-positive tumors[130]. In addition, tamoxifen can be used for the preventionof breast cancer[131].

SERMs act as estrogens in select target tissues but act asantiestrogens in other target tissues[132]. The ideal SERMwould be an estrogen agonist in bone, liver, the cardiovas-cular system and brain and an estrogen antagonist in thebreast and uterus. The activity of tamoxifen is dependent oncirculating levels of E2, which are high in premenopausalwomen and low in postmenopausal women. Tamoxifen is anantiestrogen in the breast, and decreases low density lipopro-tein cholesterol levels in postmenopausal women. Tamoxifentreatment decreases bone density in premenopausal womenbut increases bone density in postmenopausal women[133].Estrogenic activity of tamoxifen is observed in the uterus,which results in an increased incidence of endometrial can-cer in postmenopausal women[131]. The molecular basisfor the tissue specificity of tamoxifen is not well under-stood, but a possible explanation involves the interactionof the ER–tamoxifen complex with coregulators. In certain


breast cancer cell lines where tamoxifen is an antagonist, ta-moxifen recruits corepressors; however, in endometrial cellswhere tamoxifen is an agonist, tamoxifen recruits coactiva-tors[134]. SRC-1 is necessary for the agonist activity of ta-moxifen in endometrial cells[134]. These findings suggestthat differences in coregulator recruitment to a promoter candetermine the functionality of the ER in different tissues.

Changes in coregulator levels affect target gene expres-sion and may change the gene activation profile to one sup-porting a proliferative phenotype. This molecular mosaic ap-pears to correlate with clinical outcomes. Studies analyzingSRC-1 and tamoxifen showed that high SRC-1 levels maycorrelate with a favorable response to tamoxifen treatmentin women with recurrent breast cancer[135]. Gene amplifi-cation and overexpression of “amplified in breast cancer-1”(AIB1/SRC-3) has been documented in breast and ovariancancer cell lines and breast cancer biopsies[136,137]. Pa-tients with ER-positive breast tumors expressing high levelsof SRC-3 and HER2 display a poor outcome with tamox-ifen therapy[138]. In contrast, high SRC-3 levels are as-sociated with better prognosis in patients not receiving ad-juvant tamoxifen. The authors’ hypothesis is that increasedHER2 signaling results in activation of SRC-3 and the ERvia phosphorylation, which results in tamoxifen resistance.Clearly, the strategic goal of establishing the importance ofboth ER� and ER� in the outcomes of endocrine therapyshould be addressed and resolved using the tissue resourcesof the ATAC trial of tamoxifen versus anastrozole[12].

5.1.2. Tamoxifen resistanceDespite over 30 years of clinical experience with tamox-

ifen, for most patients, tumors that initially regress with ta-moxifen will eventually recur and require alternate treat-ment. The mechanisms of cellular resistance to tamoxifenare under investigation (reviewed in[139–142]).

Most resistant tumors continue to retain a functional ER[143], so loss of the ER is not sufficient to explain resis-tance. Because a functional ER is present, the cells havechanged how the ER–tamoxifen complex is perceived andhow it signals, or there is altered expression of genes tocounteract tamoxifen signaling. During the development ofresistance, tamoxifen could become an estrogen agonist byinducing E2-specific genes[144,145]. Tamoxifen resistancecould also be explained by ER mutations, coregulator ex-pression and recruitment, or interactions with other signal-ing pathways.

Additionally, a number of nonspecific mechanisms maycontribute to the tamoxifen resistant phenotype. The mech-anisms may include events that limit the intracellular avail-ability of tamoxifen, such as binding to other proteins, parti-tioning into lipophilic membrane domains, altered transportinto or out of the cell, the development of oxidative stressor the conversion of tamoxifen to other metabolites. Othermechanisms include overexpression of growth factors, in-creased angiogenesis or heterogeneity in the tumor cell pop-ulation (reviewed in[146]).

Mutations in ER� have been demonstrated in breast tu-mors, but there is no reason to believe they play a major rolein tamoxifen resistance at this point. Nevertheless, if thesemutations develop during the course of tamoxifen therapy,the tumor cells could begin to respond differently to tamox-ifen. The K303R mutation results in increased sensitivity toestrogen[147]; however, tamoxifen is still effective in block-ing estrogen action. In addition, the Y537N mutation resultsin constitutive activity of the ER that is unaffected by E2,tamoxifen and ICI 164,384[148]. The D351Y mutation wasdiscovered in tamoxifen stimulated MCF-7-derived breasttumor models[149]. Despite the fact that the mutation en-hances the estrogen-like action of SERMs[150,151], thereis no general enhancement of mutations in tamoxifen resis-tant disease[152–156]. Knowledge of the D351Y mutationin ER� and the close association of D351 with the antie-strogenic side chain of SERMs have provided an importantinsight into the molecular modulation of the SERM–ER�complex[12,157]. The balance between ER� and ER� mayalso play a role. ER� mRNA was shown to be upregulatedin tamoxifen resistant human tumors and cell lines, suggest-ing that ER� is a poor prognostic factor for the developmentof drug resistance[158].

Since the target of SERM action is the ER, it is clear thatincreases or decreases in coregulatory molecules will mod-ulate the SERM–ER complex to be estrogenic or antiestro-genic, respectively. This principle is illustrated by the find-ing of elevated levels of the coactivator SRC-3 in tumorsthat fail on tamoxifen, but only in the presence of cell sur-face signaling[138]. It is known that repression of gene ac-tivation by tamoxifen is an active process, where tamoxifenrecruits corepressors[159]. If corepressor levels are low ina particular tumor, tamoxifen may be unable to recruit a suf-ficient amount of corepressors to silence gene transcription,thereby contributing to drug resistance.

Alterations in signal transduction pathways are anothermechanism of tamoxifen resistance. Examples include re-sistance to the growth inhibitory effects of TGF-�1 [153],enhanced AP1 signaling[160–163] and upregulation ofAkt/PI3K [86,87], HER2[164–166], IGF-1 receptor[167],and active MAP kinase[168]. Activation of these pathwayscan lead to ligand-independent activation of the ER viaphosphorylation.

Overall, there are numerous potential mechanisms thatmay contribute singly or in combination to the developmentof drug resistance. Despite the possibility of drug resistance,there are potential treatments after the development of ta-moxifen resistance. These include the use of aromatase in-hibitors to block the production of estrogen[12] as well aspure antiestrogens to degrade the ER[11].

5.1.3. ERβ and breast cancerThe role of ER� in breast cancer growth and development

is not as clear as the role of ER� (reviewed in[169,170]).ER� might have a modulating effect in breast cancer be-cause it is expressed in normal and malignant breast tissue,


binds 17�-estradiol and ER� can heterodimerize with ER�[28,31,100,101]. One problem is that most of the ER� analy-ses have been at the RNA level using PCR based techniques.This is because the analysis of ER� protein has been diffi-cult with currently available antibodies as they continue toyield inconsistent results. Further, large studies comparingthe distribution of ER� and ER� using multiple antibodiesin multiple tissues are required to build on the preliminaryfindings of Taylor and Al-Azzawi[35]. The fact that thereare many different antibodies of doubtful value could leadto inconsistencies in the literature. For example, one studyindicates that ER� is a good prognostic indicator for breastcancer. Expression of ER� was associated with better sur-vival in patients receiving adjuvant tamoxifen[171]. Anotherstudy shows that ER� is associated with negative axillarynode status and low grade tumors[172]. In addition, ER�cases had a better disease free survival rate[173] and levelsof ER� were decreased in proliferative preinvasive tumors[174]. These studies suggest a protective role for ER� inbreast cancer. In contrast, evidence also suggests that ER�is a poor prognostic indicator. Tumors that expressed bothER� and ER� were node positive and of a higher grade[175]. ER� mRNA levels were also elevated in tumors thatdisplayed tamoxifen resistance[158].

Overall, the majority of studies suggest that the presenceof ER� is a good prognostic marker for breast cancer. How-ever, the relative amounts of ER� and ER� must be con-sidered. As normal breast tissue becomes tumorigenic, theamount of ER� increases whereas the amount of ER� de-creases[176]. The majority of ER present in breast tumorsis ER� so the biological relevance of ER� in breast cancerremains a topic of debate. Nevertheless, large studies thatcan correlate tumor characteristics with precise determina-tions of ER� mRNA and protein are needed to identify spe-cific situations where ER� may be a critical player in eithercarcinogenesis or disease progression.

5.2. Prostate cancer

Several lines of evidence suggest that ER� could beinvolved in prostate cancer. Most importantly, ER� isexpressed at high levels in the prostate[32]. Within theprostate, ER� localizes primarily in the epithelium, whereasER� is in the stroma[38]. ER� has also been detected innormal and malignant prostate tissue. Because the prostategland expresses a large amount of ER�, it may be moresusceptible to the effects of environmental estrogens[177].Secondly, one report analyzing ER� knockout mice showedthat these mice displayed prostatic hyperplasia[111],whereas no evidence of hyperplasia was observed in the�ERKO mice. This suggests that ER� may protect againstabnormal growth in the prostate. Although the role of ER�in the prostate remains unclear, most findings support anantiproliferative role. If this scenario is proven correct,ER� selective ligands could potentially be used for prostatecancer therapy.

5.3. Colon cancer

The distribution of ER� and ER� has been evaluated incolon cancer cell lines as well as human colon cancer sam-ples. ER� is present in the human colon cancer cell linesHCT116, HCT8, DLD-1, LoVo, HT29, Colo320, SW480and Colo205 but ER� is not present[178–180]. Studies onhuman samples showed that ER� protein was expressed atextremely low levels in normal and malignant colon tissuecompared to ER� levels[181]. Although no differences wereobserved in ER� mRNA levels using RT-PCR between nor-mal and malignant colon tissue, a loss of ER� protein wasobserved during malignant transformation[181]. In addi-tion, the localization of ER� in normal colon was nuclear,whereas a cytoplasmic localization was also observed in col-orectal carcinoma tissue[182]. Therefore, multiple studiessupport the idea that ER� is the primary ER expressed in thecolon, and that the loss and change in localization of ER�is associated with the progression to cancer.

The recent results of the Women’s Health Initiative (WHI)trial have provided an interesting insight on the role of es-trogen in colon carcinogenesis[16]. The primary goal ofthe WHI was to evaluate the use of hormone replacementtherapy (HRT, estrogen plus medroxyprogesterone acetate)in postmenopausal women aged 50–79 for the preventionof cardiovascular disease. Other outcomes were also docu-mented. The HRT group had a 26% increase in breast can-cer rates, which was anticipated[17,183]. Colorectal can-cer rates were reduced by 37%, which was also suggestedin previous studies[184]. The decreased incidence of coloncancer could be mediated by ER�. In women, ER� mRNAlevels have been shown to be decreased in colon tumors com-pared to normal tissue, whereas ER� levels did not changeand are much lower than ER� levels [185]. This evidencesuggests that the activation of ER� in the colon by HRT(estrogen) provides protection against colon cancer. This issimilar to the situation observed in prostate cancer, whereER� is thought to play a protective role.

5.4. Ovarian cancer

Although approximately two-thirds of ovarian cancersare ER positive, responses to endocrine therapy are modest[186]. However, contraceptives that combine estrogen andprogestins decrease the risk of ovarian cancer, such that 5years of contraceptive use confers a 50% risk reduction thatpersists for at least 10 years after the cessation of use[187].

ER� is the predominant ER in the ovary, where it is foundin the granulosa cells, whereas ER� localizes to the thecaland interstitial cells. Knockout mouse studies have shownthat �ERKO mouse ovaries contained cystic and hemor-rhagic follicles that contained no corpora lutea and few gran-ulosa cells[106]. Ovaries in�ERKO mice have more earlyatretic follicles, fewer corpora lutea, a partial arrest of fol-licular development and a decrease in the frequency of fol-licular maturation[111].


Studies of normal and malignant human ovaries haveyielded conflicting results. One study showed an increase inER� mRNA relative to ER� in ovarian cancer compared tonormal ovary[188]. Another study showed varying amountsof ERs in normal ovary, lower levels of ER� in ovarianepithelial primary tumors, and only ER� in metastatic tu-mors[189]. ER� levels were lower than ER� levels in ovar-ian cancer compared to normal tissue[190]. In contrast, adecrease in ER� mRNA was observed relative to ER� inhuman ovarian surface epithelial cells[191]. The majorityof studies therefore support a scenario in which ER� be-comes the dominant ER in ovarian cancer. This implies amechanism that results in ER� overexpression or a selec-tive growth advantage for ER�-positive cells. Further stud-ies are needed to fully determine the contributions of ER�and ER� to ovarian cancer.

Overall, ER� appears to play a protective role againstthe development of breast, prostate and colon cancer. Avariety of mechanisms supporting an antiproliferative rolefor ER� have been proposed to explain these findings. InER�+/ER�− colon cancer cell lines, estrogen has no effecton cell growth, but genistein slightly inhibited cell growth[179]. In addition, transfections of the cDNA for ER� inhib-ited the growth, invasion and motility of the MDA-MB-231breast cancer cell line[192]. These data support a scenario inwhich activation of ER�-mediated pathways is able to sup-press cell growth. Another possibility is that the presence ofER� could simply antagonize the growth stimulatory effectsmediated by ER�. This is suggested by a study in whichER� inhibited the agonist activity of ER�–tamoxifen com-plex [31]. Further studies are needed to dissect the preciseinteractions between ER� and ER� in cell growth control.

6. Current status of the ER and future researchdirections

It is clear that ER� and ER� are extremely importantcomponents of a complex signal transduction pathway thatspecifically regulates the growth and development of targettissues and tumors. At the molecular level, ERs act as tran-scription factors to target a variety of genes using the clas-sical ERE pathway or tethering mechanisms utilizing AP1or SP1. Usually, transcriptional activity is in response to en-dogenous ligands such as steroidal estrogens or other ligandssuch as antiestrogens or phytoestrogens. The transcriptionalactivation of the ER results in the activation of target genesthat are involved in normal physiological processes such asthe maintenance of bone density, proper reproductive organdevelopment, fertility and behavior. Abberant roles for ER�have been demonstrated in breast cancer and ER� could beinvolved in prostate cancer and colon cancer.

Knowledge of the role of estrogen in physiology andpathology has resulted in the development of effective andrelatively safe drugs that target endocrine-related breastcancer, postmenopausal osteoporosis and resulted in re-

cent advances toward the prevention of breast cancer[1,2].The developing molecular knowledge of estrogen actioncan be further exploited to design better drugs to targetER� and ER� selectively. By way of a recent example,2,3-bis(4-hydroxyphenyl) propionitrile (DPN), an ER� se-lective ligand, has a 70-fold higher relative binding affinityand a 170-fold higher potency for ER� over ER� in tran-scription assays[193]. Propyl pyrazole triol (PPT) is anER� selective ligand, with a 400-fold affinity for ER�over ER� [194]. PPT treated rats exhibited uterine weightgain, increased complement C3 mRNA in the uterus andincreased progesterone receptor mRNA in the brain. There-fore, ER� and ER� selective ligands could be utilized todelineate the specific physiological processes mediated byER� and ER�. This knowledge can be further exploited todevelop ER subtype-specific therapeutic drugs.

The application of subtype-specific drugs for therapeu-tic use requires consideration of ER subtype expressionin target tissues. For example, the cellular environment ofbreast cancer is comprised of high levels of ER� and ex-tremely low levels of ER�. Because estrogen is a growthstimulus through ER�, the use of aromatase inhibitors orER�-specific antagonists to block this interaction may beoptimal. Since evidence[171–174] indicates that ER� isprotective in the breast, the combined use of an ER� ag-onist could provide further benefit. A different scenariooccurs in the prostate gland and colon, where ER� is thedominant ER and little or no ER� is present. ER� is viewedas protective in these tissues, so the use of an ER� agonistmay provide the most benefit. However, the situation be-comes more complicated in tissues that contain high levelsof both ER� and ER�, because the interaction that will oc-cur by activating or blocking both ERs separately must bedeciphered.

The issue of receptor interaction is extremely difficult toaddress in tissues throughout the body. Nevertheless, one is-sue that has important physiological implications is whethereven a small amount of ER� can cause a significant responsein the presence of a large amount of ER�. The question tobe answered is how much of an ER is required to elicit a re-sponse. In other words, if a large amount of ER� is present,can the activation of even a small amount of ER� causea significant response through alternate signal transductionpathways? If future strategies to design ER subtype-specificdrugs are pursued, then the physical chemistry of the agentswill be extremely important. It is extremely difficult to con-trol the pharmacokinetics and pharmacodynamics of orallyactive medicines, especially when the compliance of patientsis extremely variable. Treatment strategies involve blockingthe activation of a signal transduction pathway, but in the fu-ture, a specific ER agonist may be used that inadvertently isdetrimental to some unknown target. Thus, the developmentof a selective ligand with a high affinity for one ER sub-type may inadvertently interact with the other ER subtypein some patients if dosing is too high or drug interactionsresult in inappropriate accumulation.


Although the complexities of ER� and ER� function intarget tissues remain to be fully characterized, the knowledgegained to date has provided a solid foundation for furtherprogress. Future discoveries could include the developmentof the perfect SERM. Additionally, a combination of ERsubtype-specific ligands may be able to control cancer inspecific target sites or improve current treatment strategiesfor a variety of debilitating diseases linked to estrogen with-drawal following menopause. Most importantly, the eluci-dation of the complex signal transduction pathways mightopen the door to novel therapeutic strategies not previouslyconsidered appropriate. Clearly, the close cooperation oflaboratory science with clinical outcomes has enhanced ourknowledge of the disease process in breast and endometrialcancer with established agents for treatment. The promisefor the future is to target the ERs to prevent colon, prostateand ovarian cancer. Indeed, this concept would not haveappeared to be reasonable a decade ago and has resultedfrom advances in molecular biology in the laboratory.

References

[1] Jordan VC. Antiestrogens and selective estrogen receptor modula-tors as multifunctional medicines. 1. Receptor interactions. J MedChem 2003;46:883–908.

[2] Jordan VC. Antiestrogens and selective estrogen receptor modu-lators as multifunctional medicines. 2. Clinical considerations andnew agents. J Med Chem 2003;46:1081–111.

[3] Beatson G. On the treatment of inoperable cases of the carcinomaof the mamma: suggestions for a new method of treatment, withillustrative cases. The Lancet 1896;2:104–7.

[4] Boyd S. On oophorectomy in cancer of the breast. Br Med J1900;2:1161–7.

[5] Jensen EV, Jacobson HI. Basic guides to the mechanism of estrogenaction. Recent Prog Hormone Res 1962;18:387–414.

[6] Jensen EV, Block GE, Smith S, Kyser K, DeSombre ER. Estrogenreceptors and breast cancer response to adrenalectomy. Natl CancerInst Monogr 1971;34:55–70.

[7] McGuire WL, Carbone PP, Vollmer EP. In: McGuire, WL, Carbone,PP, Sears, ME, Escher, GC, editors. Estrogen receptors in humanbreast cancer. New York: Raven Press; 1975. p. 1–7.

[8] Knight WA, Livingston RB, Gregory EJ, McGuire WL. Estrogenreceptor as an independent prognostic factor for early recurrencein breast cancer. Cancer Res 1977;37:4669–71.

[9] Jordan VC, Gapstur S, Morrow M. Selective estrogen receptormodulation and reduction in risk of breast cancer, osteoporosis, andcoronary heart disease. J Natl Cancer Inst 2001;93:1449–57.

[10] Osborne CK, Pippen J, Jones SE, et al. Double-blind randomizedtrial comparing the efficacy and tolerability of fulvestrant versusanastrozole in postmenopausal women with advanced breast cancerprogressing on prior endocrine therapy: results of a North Americantrial. J Clin Oncol 2002;20:3386–95.

[11] Howell A, Robertson JF, Quaresma Albano J, et al. Fulvestrant, for-merly ICI 182,780, is as effective as anastrozole in postmenopausalwomen with advanced breast cancer progressing after prior en-docrine treatment. J Clin Oncol 2002;20:3396–403.

[12] Anastrozole alone or in combination with tamoxifen versus tamox-ifen alone for adjuvant treatment of postmenopausal women withearly breast cancer: first results of the ATAC randomised trial,Lancet 2002;359:2131–9.

[13] Bonneterre J, Thurlimann B, Robertson JF, et al. Anastrozole ver-sus tamoxifen as first-line therapy for advanced breast cancer in668 postmenopausal women: results of the Tamoxifen or ArimidexRandomized Group Efficacy and Tolerability study. J Clin Oncol2000;18:3748–57.

[14] Nabholtz JM, Buzdar A, Pollak M, et al. Anastrozole is superior totamoxifen as first-line therapy for advanced breast cancer in post-menopausal women: results of a North American multicenter ran-domized trial. Arimidex Study Group. J Clin Oncol 2000;18:3758–67.

[15] Mouridsen H, Gershanovich M, Sun Y, et al. Superior efficacy ofletrozole versus tamoxifen as first-line therapy for postmenopausalwomen with advanced breast cancer: results of a phase III studyof the International Letrozole Breast Cancer Group. J Clin Oncol2001;19:2596–606.

[16] Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits ofestrogen plus progestin in healthy postmenopausal women: principalresults From the Women’s Health Initiative randomized controlledtrial. JAMA 2002;288:321–33.

[17] Collaborative Group on Hormonal Factors in Breast Cancer. Breastcancer and hormone replacement therapy: collaborative reanalysisof data from 51 epidemiological studies of 52,705 women withbreast cancer and 108,411 women without breast cancer. Lancet1997;350:1047–59.

[18] Walter P, Green S, Greene G, et al. Cloning of the human estrogenreceptor cDNA. Proc Natl Acad Sci USA 1985;82:7889–93.

[19] Greene GL, Gilna P, Waterfield M, Baker A, Hort Y, Shine J. Se-quence and expression of human estrogen receptor complementaryDNA. Science 1986;231:1150–4.

[20] Green S, Walter P, Kumar V, et al. Human oestrogen receptorcDNA: sequence. Nature 1986;320:134–9.

[21] Flouriot G, Brand H, Denger S. Identification of a new isoform ofthe human estrogen receptor-alpha (hER-alpha) that is encoded bydistinct transcripts and that is able to repress hER-alpha activationfunction 1. EMBO J 2000;19:4688–700.

[22] Murphy LC, Dotzlaw H, Leygue E, Douglas D, Coutts A, WatsonPH. Estrogen receptor variants and mutations. J Steroid BiochemMol Biol 1997;62:363–72.

[23] Poola I, Koduri S, Chatra S, Clarke R. Identification of twentyalternatively spliced estrogen receptor alpha mRNAs in breast cancercell lines and tumors using splice targeted primer approach. J SteroidBiochem Mol Biol 2000;72:249–58.

[24] Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA.Cloning of a novel receptor expressed in rat prostate and ovary.Proc Natl Acad Sci USA 1996;93:5925–30.

[25] Tremblay GB, Tremblay A, Copeland NG, et al. Cloning, chromo-somal localization, and functional analysis of the murine estrogenreceptor beta. Mol Endocrinol 1997;11:353–65.

[26] Mosselman S, Polman J, Dijkema R. ER beta: identification andcharacterization of a novel human estrogen receptor. FEBS Lett1996;392:49–53.

[27] Enmark E, Pelto-Huikko M, Grandien K, et al. Human estrogenreceptor beta-gene structure, chromosomal localization, and expres-sion pattern. J Clin Endocrinol Metab 1997;82:4258–65.

[28] Ogawa S, Inoue S, Watanabe T, et al. The complete primary struc-ture of human estrogen receptor beta (hER beta) and its heterodimer-ization with ER alpha in vivo and in vitro. Biochem Biophys ResCommun 1998;243:122–6.

[29] Lewandowski S, Kalita K, Kaczmarek L. Estrogen receptor beta.Potential functional significance of a variety of mRNA isoforms.FEBS Lett 2002;524:1–5.

[30] Nilsson S, Makela S, Treuter E, et al. Mechanisms of estrogenaction. Physiol Rev 2001;81:1535–65.

[31] Hall JM, McDonnell DP. The estrogen receptor beta-isoform (ER-beta) of the human estrogen receptor modulates ERalpha transcrip-tional activity and is a key regulator of the cellular response toestrogens and antiestrogens. Endocrinology 1999;140:5566–78.


[32] Kuiper GG, Carlsson B, Grandien K, et al. Comparison of the ligandbinding specificity and transcript tissue distribution of estrogenreceptors alpha and beta. Endocrinology 1997;138:863–70.

[33] Kuiper GG, Lemmen JG, Carlsson B, et al. Interaction of estro-genic chemicals and phytoestrogens with estrogen receptor beta.Endocrinology 1998;139:4252–63.

[34] Gustafsson JA. Estrogen receptor beta—a new dimension in estro-gen mechanism of action. J Endocrinol 1999;163:379–83.

[35] Taylor AH, Al-Azzawi F. Immunolocalisation of oestrogen receptorbeta in human tissues. J Mol Endocrinol 2000;24:145–55.

[36] Hiroi H, Inoue S, Watanabe T, et al. Differential immunolocalizationof estrogen receptor alpha and beta in rat ovary and uterus. J MolEndocrinol 1999;22:37–44.

[37] Sar M, Welsch F. Differential expression of estrogen receptor-betaand estrogen receptor-alpha in the rat ovary. Endocrinology1999;140:963–71.

[38] Weihua Z, Warner M, Gustafsson JA. Estrogen receptor beta in theprostate. Mol Cell Endocrinol 2002;193:1–5.

[39] Kumar V, Green S, Stack G, Berry M, Jin JR, Chambon P. Func-tional domains of the human estrogen receptor. Cell 1987;51:941–51.

[40] Tora L, White J, Brou C, et al. The human estrogen receptor hastwo independent nonacidic transcriptional activation functions. Cell1989;59:477–87.

[41] Berry M, Metzger D, Chambon P. Role of the two activat-ing domains of the oestrogen receptor in the cell-type andpromoter-context dependent agonistic activity of the anti- oestrogen4-hydroxytamoxifen. EMBO J 1990;9:2811–8.

[42] Danielian PS, White R, Lees JA, Parker MG. Identification of aconserved region required for hormone dependent transcriptionalactivation by steroid hormone receptors. EMBO J 1992;11:1025–33.

[43] Tzukerman MT, Esty A, Santiso-Mere D, et al. Human estrogenreceptor transactivational capacity is determined by both cellularand promoter context and mediated by two functionally distinctintramolecular regions. Mol Endocrinol 1994;8:21–30.

[44] Kraus WL, McInerney EM, Katzenellenbogen BS.Ligand-dependent, transcriptionally productive association of theamino- and carboxyl-terminal regions of a steroid hormone nuclearreceptor. Proc Natl Acad Sci USA 1995;92:12314–8.

[45] Fawell SE, White R, Hoare S, Sydenham M, Page M, Parker MG.Inhibition of estrogen receptor-DNA binding by the “pure” antie-strogen ICI 164,384 appears to be mediated by impaired receptordimerization. Proc Natl Acad Sci USA 1990;87:6883–7.

[46] Dauvois S, Danielian PS, White R, Parker MG. Antiestrogen ICI164,384 reduces cellular estrogen receptor content by increasing itsturnover. Proc Natl Acad Sci USA 1992;89:4037–41.

[47] McKenna NJ, Lanz RB, O’Malley BW. Nuclear receptor coregula-tors: cellular and molecular biology. Endocr Rev 1999;20:321–44.

[48] Klinge CM. Estrogen receptor interaction with co-activators andco-repressors. Steroids 2000;65:227–51.

[49] Tremblay GB, Giguere V. Coregulators of estrogen receptor action.Crit Rev Eukaryot Gene Exp 2002;12:1–22.

[50] Heery DM, Kalkhoven E, Hoare S, Parker MG. A signature motif intranscriptional co-activators mediates binding to nuclear receptors.Nature 1997;387:733–6.

[51] Montano MM, Ekena K, Delage-Mourroux R, Chang W, MartiniP, Katzenellenbogen BS. An estrogen receptor-selective coregulatorthat potentiates the effectiveness of antiestrogens and represses theactivity of estrogens. Proc Natl Acad Sci USA 1999;96:6947–52.

[52] Katzenellenbogen BS, Montano MM, Ediger TR, et al. Estrogenreceptors: selective ligands, partners, and distinctive pharmacology.Recent Prog Horm Res 2000;55:163–93.

[53] Endoh H, Maruyama K, Masuhiro Y, et al. Purification and identi-fication of p68 RNA helicase acting as a transcriptional coactivatorspecific for the activation function 1 of human estrogen receptoralpha. Mol Cell Biol 1999;19:5363–72.

[54] Suen CS, Berrodin TJ, Mastroeni R, Cheskis BJ, Lyttle CR, FrailDE. A transcriptional coactivator, steroid receptor coactivator-3,selectively augments steroid receptor transcriptional activity. J BiolChem 1998;273:27645–53.

[55] Watanabe T, Inoue S, Ogawa S, et al. Agonistic effect of tamoxifenis dependent on cell type, ERE-promoter context, and estrogenreceptor subtype: functional difference between estrogen receptorsalpha and beta. Biochem Biophys Res Commun 1997;236:140–5.

[56] McInerney EM, Weis KE, Sun J, Mosselman S, KatzenellenbogenBS. Transcription activation by the human estrogen receptor subtypebeta (ER beta) studied with ER beta and ER alpha receptor chimeras.Endocrinology 1998;139:4513–22.

[57] Adlercreutz H. Phyto-oestrogens and cancer. Lancet Oncol2002;3:364–73.

[58] Mishra SI, Dickerson V, Najm W. Phytoestrogens and breastcancer prevention: what is the evidence? Am J Obstet Gynecol2002;188:S66–70.

[59] Adlercreutz H, Mazur W. Phyto-oestrogens and Western diseases.Ann Med 1997;29:95–120.

[60] Shiau AK, Barstad D, Radek JT, et al. Structural characterization ofa subtype-selective ligand reveals a novel mode of estrogen receptorantagonism. Nat Struct Biol 2002;9:359–64.

[61] Sun J, Huang YR, Harrington WR, Sheng S, Katzenellenbogen JA,Katzenellenbogen BS. Antagonists selective for estrogen receptoralpha. Endocrinology 2002;143:941–7.

[62] Brzozowski AM, Pike AC, Dauter Z, et al. Molecular basisof agonism and antagonism in the oestrogen receptor. Nature1997;389:753–8.

[63] Pike AC, Brzozowski AM, Hubbard RE. A structural biologist’sview of the oestrogen receptor. J Steroid Biochem Mol Biol2000;74:261–8.

[64] Jordan VC, Collins MM, Rowsby L, Prestwich G. A monohydrox-ylated metabolite of tamoxifen with potent antioestrogenic activity.J Endocrinol 1977;75:305–16.

[65] Shiau AK, Barstad D, Loria PM, et al. The structural basis ofestrogen receptor/coactivator recognition and the antagonism of thisinteraction by tamoxifen. Cell 1998;95:927–37.

[66] Pike AC, Brzozowski AM, Hubbard RE, et al. Structure of theligand-binding domain of oestrogen receptor beta in the presence ofa partial agonist and a full antagonist. EMBO J 1999;18:4608–18.

[67] Pike AC, Brzozowski AM, Walton J, et al. Structural insightsinto the mode of action of a pure antiestrogen. Structure (Camb)2001;9:145–53.

[68] van Hoorn WP. Identification of a second binding site in the estrogenreceptor. J Med Chem 2002;45:584–9.

[69] Zhang Z, Maier B, Santen RJ, Song RX. Membrane association ofestrogen receptor alpha mediates estrogen effect on MAPK activa-tion. Biochem Biophys Res Commun 2002;294:926–33.

[70] Levin ER. Cellular functions of plasma membrane estrogen recep-tors. Steroids 2002;67:471–5.

[71] Santen RJ, Song RX, McPherson R, et al. The role ofmitogen-activated protein (MAP) kinase in breast cancer. J SteroidBiochem Mol Biol 2002;80:239–56.

[72] Razandi M, Pedram A, Greene GL, Levin ER. Cell membrane andnuclear estrogen receptors (ERs) originate from a single transcript:studies of ERalpha and ERbeta expressed in Chinese hamster ovarycells. Mol Endocrinol 1999;13:307–19.

[73] Lannigan DA. Estrogen receptor phosphorylation. Steroids2003;68:1–9.

[74] Kato S. Estrogen receptor-mediated cross-talk with growth factorsignaling pathways. Breast Cancer 2001;8:3–9.

[75] Denton RR, Koszewski NJ, Notides AC. Estrogen receptor phos-phorylation. Hormonal dependence and consequence on specificDNA binding. J Biol Chem 1992;267:7263–8.

[76] Le Goff P, Montano MM, Schodin DJ, Katzenellenbogen BS.Phosphorylation of the human estrogen receptor. Identification of


hormone-regulated sites and examination of their influence on tran-scriptional activity. J Biol Chem 1994;269:4458–66.

[77] Ali S, Metzger D, Bornert JM, Chambon P. Modulation of tran-scriptional activation by ligand-dependent phosphorylation of thehuman oestrogen receptor A/B region. EMBO J 1993;12:1153–60.

[78] Joel PB, Traish AM, Lannigan DA. Estradiol and phorbol estercause phosphorylation of serine 118 in the human estrogen receptor.Mol Endocrinol 1995;9:1041–52.

[79] Bunone G, Briand PA, Miksicek RJ, Picard D. Activation of theunliganded estrogen receptor by EGF involves the MAP kinasepathway and direct phosphorylation. EMBO J 1996;15:2174–83.

[80] Kato S, Endoh H, Masuhiro Y, et al. Activation of the estrogenreceptor through phosphorylation by mitogen-activated protein ki-nase. Science 1995;270:1491–4.

[81] Joel PB, Traish AM, Lannigan DA. Estradiol-induced phosphoryla-tion of serine 118 in the estrogen receptor is independent of p42/p44mitogen-activated protein kinase. J Biol Chem 1998;273:13317–23.

[82] Chen D, Washbrook E, Sarwar N, et al. Phosphorylation of hu-man estrogen receptor alpha at serine 118 by two distinct signaltransduction pathways revealed by phosphorylation-specific antis-era. Oncogene 2002;21:4921–31.

[83] Arnold SF, Obourn JD, Jaffe H, Notides AC. Serine 167 is themajor estradiol-induced phosphorylation site on the human estrogenreceptor. Mol Endocrinol 1994;8:1208–14.

[84] Joel PB, Smith J, Sturgill TW, Fisher TL, Blenis J, Lannigan DA.pp90rsk1 regulates estrogen receptor-mediated transcription throughphosphorylation of Ser-167. Mol Cell Biol 1998;18:1978–84.

[85] Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW,Liao JK. Interaction of oestrogen receptor with the regulatory sub-unit of phosphatidylinositol-3-OH kinase. Nature 2000;407:538–41.

[86] Sun M, Paciga JE, Feldman RI, et al. Phosphatidylinositol-3-OHKinase (PI3K)/AKT2, activated in breast cancer, regulates and isinduced by estrogen receptor alpha (ERalpha) via interaction be-tween ERalpha and PI3K. Cancer Res 2001;61:5985–91.

[87] Campbell RA, Bhat-Nakshatri P, Patel NM, Constantinidou D, AliS, Nakshatri H. Phosphatidylinositol 3-kinase/AKT-mediated acti-vation of estrogen receptor alpha: a new model for anti-estrogenresistance. J Biol Chem 2001;276:9817–24.

[88] Rogatsky I, Trowbridge JM, Garabedian MJ. Potentiation of humanestrogen receptor alpha transcriptional activation through phospho-rylation of serines 104 and 106 by the cyclin A-CDK2 complex. JBiol Chem 1999;274:22296–302.

[89] Chen D, Pace PE, Coombes RC, Ali S. Phosphorylation of humanestrogen receptor alpha by protein kinase A regulates dimerization.Mol Cell Biol 1999;19:1002–15.

[90] Arnold SF, Obourn JD, Jaffe H, Notides AC. Phosphorylation ofthe human estrogen receptor on tyrosine 537 in vivo and by srcfamily tyrosine kinases in vitro. Mol Endocrinol 1995;9:24–33.

[91] Tremblay A, Tremblay GB, Labrie F, Giguere V.Ligand-independent recruitment of SRC-1 to estrogen receptorbeta through phosphorylation of activation function AF-1. MolCell 1999;3:513–9.

[92] Gee JM, Robertson JF, Ellis IO, Nicholson RI. Phosphorylation ofERK1/2 mitogen-activated protein kinase is associated with poorresponse to anti-hormonal therapy and decreased patient survivalin clinical breast cancer. Int J Cancer 2001;95:247–54.

[93] Kushner PJ, Agard DA, Greene GL, et al. Estrogen receptor path-ways to AP-1. J Steroid Biochem Mol Biol 2000;74:311–7.

[94] Paech K, Webb P, Kuiper GG, et al. Differential ligand activationof estrogen receptors ERalpha and ERbeta at AP1 sites. Science1997;277:1508–10.

[95] Webb P, Lopez GN, Uht RM, Kushner PJ. Tamoxifen activationof the estrogen receptor/AP-1 pathway: potential origin for thecell-specific estrogen-like effects of antiestrogens. Mol Endocrinol1995;9:443–56.

[96] Webb P, Nguyen P, Valentine C, et al. The estrogen receptor en-hances AP-1 activity by two distinct mechanisms with different

requirements for receptor transactivation functions. Mol Endocrinol1999;13:1672–85.

[97] Weatherman RV, Scanlan TS. Unique protein determinants of thesubtype-selective ligand responses of the estrogen receptors (ERal-pha and ERbeta ) at AP-1 sites. J Biol Chem 2001;276:3827–32.

[98] Safe S. Transcriptional activation of genes by 17 beta-estradiolthrough estrogen receptor-Sp1 interactions. Vitam Horm2001;62:231–52.

[99] Saville B, Wormke M, Wang F, et al. Ligand-, cell-, and estrogenreceptor subtype (alpha/beta)-dependent activation at GC-rich (Sp1)promoter elements. J Biol Chem 2000;275:5379–87.

[100] Cowley SM, Hoare S, Mosselman S, Parker MG. Estrogen re-ceptors alpha and beta form heterodimers on DNA. J Biol Chem1997;272:19858–62.

[101] Pettersson K, Grandien K, Kuiper GG, Gustafsson JA. Mouseestrogen receptor beta forms estrogen response element-bindingheterodimers with estrogen receptor alpha. Mol Endocrinol1997;11:1486–96.

[102] Curtis Hewitt S, Couse JF, Korach KS. Estrogen receptor transcrip-tion and transactivation: Estrogen receptor knockout mice: whattheir phenotypes reveal about mechanisms of estrogen action. BreastCancer Res 2000;2:345–52.

[103] Couse JF, Curtis Hewitt S, Korach KS. Receptor null mice revealcontrasting roles for estrogen receptor alpha and beta in reproductivetissues. J Steroid Biochem Mol Biol 2000;74:287–96.

[104] Curtis SH, Korach KS. Steroid receptor knockout models: pheno-types and responses illustrate interactions between receptor signal-ing pathways in vivo. Adv Pharmacol 2000;47:357–80.

[105] Couse JF, Korach KS. Estrogen receptor null mice: what have welearned and where will they lead us? Endocr Rev 1999;20:358–417.

[106] Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, SmithiesO. Alteration of reproductive function but not prenatal sexual devel-opment after insertional disruption of the mouse estrogen receptorgene. Proc Natl Acad Sci USA 1993;90:11162–6.

[107] Bocchinfuso WP, Korach KS. Mammary gland development andtumorigenesis in estrogen receptor knockout mice. J MammaryGland Biol Neoplasia 1997;2:323–34.

[108] Day JK, Besch-Williford C, McMann TR, Hufford MG, LubahnDB, MacDonald RS. Dietary genistein increased DMBA-inducedmammary adenocarcinoma in wild-type, but not ER alpha KO,mice. Nutr Cancer 2001;39:226–32.

[109] Eddy EM, Washburn TF, Bunch DO, et al. Targeted disruptionof the estrogen receptor gene in male mice causes alteration ofspermatogenesis and infertility. Endocrinology 1996;137:4796–805.

[110] Ogawa S, Lubahn DB, Korach KS, Pfaff DW. Behavioral effectsof estrogen receptor gene disruption in male mice. Proc Natl AcadSci USA 1997;94:1476–81.

[111] Krege JH, Hodgin JB, Couse JF, et al. Generation and reproductivephenotypes of mice lacking estrogen receptor beta. Proc Natl AcadSci USA 1998;95:15677–82.

[112] Ogawa S, Chan J, Chester AE, Gustafsson JA, Korach KS, PfaffDW. Survival of reproductive behaviors in estrogen receptor betagene-deficient (betaERKO) male and female mice. Proc Natl AcadSci USA 1999;96:12887–92.

[113] Dupont S, Krust A, Gansmuller A, Dierich A, Chambon P, MarkM. Effect of single and compound knockouts of estrogen receptorsalpha (ERalpha) and beta (ERbeta) on mouse reproductive pheno-types. Development 2000;127:4277–91.

[114] Wang L, Andersson S, Warner M, Gustafsson JA. Morphologicalabnormalities in the brains of estrogen receptor beta knockout mice.Proc Natl Acad Sci USA 2001;98:2792–6.

[115] Couse JF, Hewitt SC, Bunch DO, et al. Postnatal sex reversal of theovaries in mice lacking estrogen receptors alpha and beta. Science1999;286:2328–31.

[116] Ogawa S, Chester AE, Hewitt SC, et al. From the cover: abolition ofmale sexual behaviors in mice lacking estrogen receptors alpha and


beta (alpha beta ERKO). Proc Natl Acad Sci USA 2000;97:14737–41.

[117] Iafrati MD, Karas RH, Aronovitz M, et al. Estrogen inhibits thevascular injury response in estrogen receptor alpha-deficient mice.Nat Med 1997;3:545–8.

[118] Karas RH, Hodgin JB, Kwoun M, et al. Estrogen inhibits thevascular injury response in estrogen receptor beta-deficient femalemice. Proc Natl Acad Sci USA 1999;96:15133–6.

[119] Karas RH, Schulten H, Pare G, et al. Effects of estrogen on thevascular injury response in estrogen receptor alpha, beta (double)knockout mice. Circ Res 2001;89:534–9.

[120] Couse JF, Curtis SW, Washburn TF, et al. Analysis of transcriptionand estrogen insensitivity in the female mouse after targeted disrup-tion of the estrogen receptor gene. Mol Endocrinol 1995;9:1441–54.

[121] Mendelsohn ME. Protective effects of estrogen on the cardiovascularsystem. Am J Cardiol 2002;89:12E–7E, discussion 17E-18E.

[122] Rubanyi GM, Freay AD, Kauser K, et al. Vascular estrogen receptorsand endothelium-derived nitric oxide production in the mouse aorta.Gender difference and effect of estrogen receptor gene disruption.J Clin Invest 1997;99:2429–37.

[123] Johnson BD, Zheng W, Korach KS, Scheuer T, Catterall WA,Rubanyi GM. Increased expression of the cardiac L-type cal-cium channel in estrogen receptor-deficient mice. J Gen Physiol1997;110:135–40.

[124] Smith EP, Boyd J, Frank GR, et al. Estrogen resistance caused bya mutation in the estrogen-receptor gene in a man. N Engl J Med1994;331:1056–61.

[125] Sudhir K, Chou TM, Chatterjee K, et al. Premature coronary arterydisease associated with a disruptive mutation in the estrogen receptorgene in a man. Circulation 1997;96:3774–7.

[126] Sudhir K, Chou TM, Messina LM, et al. Endothelial dysfunction ina man with disruptive mutation in oestrogen-receptor gene. Lancet1997;349:1146–7.

[127] Jensen EV, Jordan VC. The estrogen receptor: a model for molecularmedicine. The Dorothy P. Landon AACR Prize for TranslationalResearch. Clin Cancer Res 2003, in press.

[128] Horwitz KB. The central role of progesterone receptors and proges-tational agents in the management and treatment of breast cancer.Semin Oncol 1988;15:14–9.

[129] Clark GM, McGuire WL. Steroid receptors and other prognosticfactors in primary breast cancer. Semin Oncol 1988;15:20–5.

[130] Early Breast Cancer Trialists’ Collaborative Group. Tamoxifen forearly breast cancer: an overview of the randomised trials. Lancet1998;351:1451–67.

[131] Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen forprevention of breast cancer: report of the National Surgical Ad-juvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst1998;90:1371–88.

[132] Jordan VC. Selective estrogen receptor modulation: a personal per-spective. Cancer Res 2001;61:5683–7.

[133] Powles TJ, Hickish T, Kanis JA, Tidy A, Ashley S. Effect of tamox-ifen on bone mineral density measured by dual-energy X-ray ab-sorptiometry in healthy premenopausal and postmenopausal women.J Clin Oncol 1996;14:78–84.

[134] Shang Y, Brown M. Molecular determinants for the tissue specificityof SERMs. Science 2002;295:2465–8.

[135] Berns EM, van Staveren IL, Klijn JG, Foekens JA. Predictive valueof SRC-1 for tamoxifen response of recurrent breast cancer. BreastCancer Res Treat 1998;48:87–92.

[136] Anzick SL, Kononen J, Walker RL, et al. AIB1, a steroid re-ceptor coactivator amplified in breast and ovarian cancer. Science1997;277:965–8.

[137] List HJ, Reiter R, Singh B, Wellstein A, Riegel AT. Expression ofthe nuclear coactivator AIB1 in normal and malignant breast tissue.Breast Cancer Res Treat 2001;68:21–8.

[138] Osborne CK, Bardou V, Hopp TA, et al. Role of the estro-gen receptor coactivator AIB1 (SRC-3) and HER-2/neu in tamox-

ifen resistance in breast cancer. J Natl Cancer Inst 2003;95:353–61.

[139] Clarke R, Skaar TC, Bouker KB, et al. Molecular and pharmaco-logical aspects of antiestrogen resistance. J Steroid Biochem MolBiol 2001;76:71–84.

[140] Clarke R, Leonessa F, Welch JN, Skaar TC. Cellular and molecularpharmacology of antiestrogen action and resistance. Pharmacol Rev2001;53:25–71.

[141] Ali S, Coombes RC. Endocrine-responsive breast cancer and strate-gies for combating resistance. Nat Rev Cancer 2002;2:101–15.

[142] Schiff R, Massarweh S, Shou J, Osborne CK. Breast cancer en-docrine resistance: how growth factor signaling and estrogen recep-tor coregulators modulate response. Clin Cancer Res 2003;9:447S–54S.

[143] Johnston SRD, Saccanti-Jotti G, Smith IE, Newby J, DowsettM. Change in estrogen receptor expression and function intamoxifen-resistant breast cancer. Endocr Relat Cancer 1995;2:105–10.

[144] Vignon F, Lippman ME, Nawata H, Derocq D, Rochefort H. Induc-tion of two estrogen-responsive proteins by antiestrogens in R27, atamoxifen-resistant clone of MCF7 cells. Cancer Res 1984;44:2084–8.

[145] Westley J, Steiner DF, Westley B. Effects of antiestrogens on theestrogen-regulated pS2 RNA and the 52- and 160-kilodalton proteinsin MCF7 cells and two tamoxifen-resistant sublines. J Biol Chem1984;259:6641–9.

[146] MacGregor JI, Jordan VC. Basic guide to the mechanisms of antie-strogen action. Pharmacol Rev 1998;50:151–96.

[147] Fuqua SA, Wiltschke C, Zhang QX, et al. A hypersensitive estrogenreceptor-alpha mutation in premalignant breast lesions. Cancer Res2000;60:4026–9.

[148] Zhang QX, Borg A, Wolf DM, Oesterreich S, Fuqua SA. Anestrogen receptor mutant with strong hormone-independent activityfrom a metastatic breast cancer. Cancer Res 1997;57:1244–9.

[149] Wolf DM, Jordan VC. The estrogen receptor from a tamoxifenstimulated MCF-7 tumor variant contains a point mutation in theligand binding domain. Breast Cancer Res Treat 1994;31:129–38.

[150] Levenson AS, Tonetti DA, Jordan VC. The oestrogen-like effectof 4-hydroxytamoxifen on induction of transforming growth factoralpha mRNA in MDA-MB-231 breast cancer cells stably expressingthe oestrogen receptor. Br J Cancer 1998;77:1812–9.

[151] Liu H, Lee ES, de los Reyes A, Zapf JW, Jordan VC. Silencing andreactivation of the selective estrogen receptor modulator–estrogenreceptor alpha complex. Cancer Res 2001;61:3632–9.

[152] Bilimoria MM, Assikis VJ, Muenzner HD, Wolf DM, Satyaswa-roop PG, Jordan VC. An analysis of tamoxifen-stimulated humancarcinomas for mutations in the AF-2 region of the estrogen recep-tor. J Steroid Biochem Mol Biol 1996;58:479–88.

[153] Herman ME, Katzenellenbogen BS. Response-specific antiestrogenresistance in a newly characterized MCF-7 human breast cancer cellline resulting from long-term exposure to trans-hydroxytamoxifen.J Steroid Biochem Mol Biol 1996;59:121–34.

[154] Mullick A, Chambon P. Characterization of the estrogen receptorin two antiestrogen-resistant cell lines, LY2 and T47D. Cancer Res1990;50:333–8.

[155] Karnik PS, Kulkarni S, Liu XP, Budd GT, Bukowski RM. Estrogenreceptor mutations in tamoxifen-resistant breast cancer. Cancer Res1994;54:349–53.

[156] Jiang SY, Wolf DM, Yingling JM, Chang C, Jordan VC. An estrogenreceptor positive MCF-7 clone that is resistant to antiestrogens andestradiol. Mol Cell Endocrinol 1992;90:77–86.

[157] Liu H, Park WC, Bentrem DJ, et al. Structure-function relationshipsof the raloxifene–estrogen receptor-alpha complex for regulatingtransforming growth factor-alpha expression in breast cancer cells.J Biol Chem 2002;277:9189–98.


[158] Speirs V, Malone C, Walton DS, Kerin MJ, Atkin SL. Increasedexpression of estrogen receptor beta mRNA in tamoxifen-resistantbreast cancer patients. Cancer Res 1999;59:5421–4.

[159] Yamamoto Y, Wada O, Suzawa M, et al. The tamoxifen-responsiveestrogen receptor alpha mutant D351Y shows reducedtamoxifen-dependent interaction with corepressor complexes. J BiolChem 2001;276:42684–91.

[160] Dumont JA, Bitonti AJ, Wallace CD, Baumann RJ, Cashman EA,Cross-Doersen DE. Progression of MCF-7 breast cancer cells toantiestrogen-resistant phenotype is accompanied by elevated levelsof AP-1 DNA-binding activity. Cell Growth Differ 1996;7:351–9.

[161] Johnston SR, Lu B, Scott GK, et al. Increased activator protein-1DNA binding and c-Jun NH2-terminal kinase activity in humanbreast tumors with acquired tamoxifen resistance. Clin Cancer Res1999;5:251–6.

[162] Smith LM, Wise SC, Hendricks DT, et al. cJun overexpressionin MCF-7 breast cancer cells produces a tumorigenic. Oncogene1999;18:6063–70.

[163] Schiff R, Reddy P, Ahotupa M, et al. Oxidative stress and AP-1activity in tamoxifen-resistant breast tumors in vivo. J Natl CancerInst 2000;92:1926–34.

[164] Chung YL, Sheu ML, Yang SC, Lin CH, Yen SH. Resistance totamoxifen-induced apoptosis is associated with direct interactionbetween Her2/neu and cell membrane estrogen receptor in breastcancer. Int J Cancer 2002;97:306–12.

[165] Kurokawa H, Lenferink AE, Simpson JF, et al. Inhibition ofHER2/neu (erbB-2) and mitogen-activated protein kinases enhancestamoxifen action against HER2-overexpressing, tamoxifen-resistantbreast cancer cells. Cancer Res 2000;60:5887–94.

[166] Pietras RJ, Arboleda J, Reese DM, et al. HER-2 tyro-sine kinase pathway targets estrogen receptor and promoteshormone-independent growth in human breast cancer cells. Onco-gene 1995;10:2435–46.

[167] Parisot JP, Hu XF, DeLuise M, Zalcberg JR. Altered expressionof the IGF-1 receptor in a tamoxifen-resistant human breast cancercell line. Br J Cancer 1999;79:693–700.

[168] Rabenoelina F, Semlali A, Duchesne MJ, Freiss G, Pons M, BadiaE. Effect of prolonged hydroxytamoxifen treatment of MCF-7 cellson mitogen activated kinase cascade. Int J Cancer 2002;98:698–706.

[169] Palmieri C, Cheng GJ, Saji S, et al. Estrogen receptor beta in breastcancer. Endocr Relat Cancer 2002;9:1–13.

[170] Gustafsson JA, Warner M. Estrogen receptor beta in the breast:role in estrogen responsiveness and development of breast cancer.J Steroid Biochem Mol Biol 2000;74:245–8.

[171] Mann S, Laucirica R, Carlson N, et al. Estrogen receptor betaexpression in invasive breast cancer. Hum Pathol 2001;32:113–8.

[172] Jarvinen TA, Pelto-Huikko M, Holli K, Isola J. Estrogen receptorbeta is coexpressed with ERalpha and PR and associated with nodalstatus, grade, and proliferation rate in breast cancer. Am J Pathol2000;156:29–35.

[173] Omoto Y, Inoue S, Ogawa S, et al. Clinical value of the wild-typeestrogen receptor beta expression in breast cancer. Cancer Lett2001;163:207–12.

[174] Roger P, Sahla ME, Makela S, Gustafsson JA, Baldet P, RochefortH. Decreased expression of estrogen receptor beta protein in prolif-erative preinvasive mammary tumors. Cancer Res 2001;61:2537–41.

[175] Speirs V, Parkes AT, Kerin MJ, et al. Coexpression of estrogenreceptor alpha and beta: poor prognostic factors in human breastcancer? Cancer Res 1999;59:525–8.

[176] Leygue E, Dotzlaw H, Watson PH, Murphy LC. Altered estrogenreceptor alpha and beta messenger RNA expression during humanbreast tumorigenesis. Cancer Res 1998;58:3197–201.

[177] Chang WY, Prins GS. Estrogen receptor-beta: implications for theprostate gland. Prostate 1999;40:115–24.

[178] Fiorelli G, Picariello L, Martineti V, Tonelli F, Brandi ML. Func-tional estrogen receptor beta in colon cancer cells. Biochem Bio-phys Res Commun 1999;261:521–7.

[179] Arai N, Strom A, Rafter JJ, Gustafsson JA. Estrogen receptorbeta mRNA in colon cancer cells: growth effects of estrogen andgenistein. Biochem Biophys Res Commun 2000;270:425–31.

[180] Qiu Y, Waters CE, Lewis AE, Langman MJ, Eggo MC.Oestrogen-induced apoptosis in colonocytes expressing oestrogenreceptor beta. J Endocrinol 2002;174:369–77.

[181] Foley EF, Jazaeri AA, Shupnik MA, Jazaeri O, Rice LW. Selectiveloss of estrogen receptor beta in malignant human colon. CancerRes 2000;60:245–8.

[182] Witte D, Chirala M, Younes A, Li Y, Younes M. Estrogen receptorbeta is expressed in human colorectal adenocarcinoma. Hum Pathol2001;32:940–4.

[183] Hulley S, Furberg C, Barrett-Connor E, et al. Noncardiovasculardisease outcomes during 6.8 years of hormone therapy: Heart andEstrogen/progestin Replacement Study follow-up (HERS II). Jama2002;288:58–66.

[184] Grodstein F, Newcomb PA, Stampfer MJ. Postmenopausal hormonetherapy and the risk of colorectal cancer: a review and meta-analysis.Am J Med 1999;106:574–82.

[185] Campbell-Thompson M, Lynch IJ, Bhardwaj B. Expression of es-trogen receptor (ER) subtypes and ERbeta isoforms in colon cancer.Cancer Res 2001;61:632–40.

[186] Makar AP. Hormone therapy in epithelial ovarian cancer. EndocrRelat Cancer 2000;7:85–93.

[187] Hankinson SE, Colditz GA, Hunter DJ, Spencer TL, Rosner B,Stampfer MJ. A quantitative assessment of oral contraceptive useand risk of ovarian cancer. Obstet Gynecol 1992;80:708–14.

[188] Pujol P, Rey JM, Nirde P, et al. Differential expression of estrogenreceptor-alpha and -beta messenger RNAs as a potential marker ofovarian carcinogenesis. Cancer Res 1998;58:5367–73.

[189] Rutherford T, Brown WD, Sapi E, Aschkenazi S, Munoz A, Mor G.Absence of estrogen receptor-beta expression in metastatic ovariancancer. Obstet Gynecol 2000;96:417–21.

[190] Sakaguchi H, Fujimoto J, Aoki I, Toyoki H, Khatun S, Tamaya T.Expression of oestrogen receptor alpha and beta in uterine endome-trial and ovarian cancers. Eur J Cancer 2002;38(Suppl 6):S74–5.

[191] Lau KM, Mok SC, Ho SM. Expression of human estrogenreceptor-alpha and -beta, progesterone receptor, and androgen re-ceptor mRNA in normal and malignant ovarian epithelial cells. ProcNatl Acad Sci USA 1999;96:5722–7.

[192] Lazennec G, Bresson D, Lucas A, Chauveau C, Vignon F. ERbeta inhibits proliferation and invasion of breast cancer cells. En-docrinology 2001;142:4120–30.

[193] Meyers MJ, Sun J, Carlson KE, Marriner GA, KatzenellenbogenBS, Katzenellenbogen JA. Estrogen receptor-beta potency-selectiveligands: structure–activity relationship studies of diarylpropioni-triles and their acetylene and polar analogues. J Med Chem2001;44:4230–51.

[194] Harris HA, Katzenellenbogen JA, Katzenellenbogen BS. Character-ization of the biological roles of the estrogen receptors, ERalphaand ERbeta, in estrogen target tissues in vivo through the use ofan ERalpha-selective ligand. Endocrinology 2002;143:4172–7.

Biographies

Sandra Timm Pearceis a postdoctoral fellow in the RobertH. Lurie Comprehensive Cancer Center of NorthwesternUniversity. After receiving a PhD in biology from the Uni-versity of Virginia in 2001, she joined the laboratory of Dr.V. Craig Jordan. Dr. Pearce is supported by the Programin Signal Transduction and Cancer (Grant T32-CA70085)and the Avon Foundation. Her research interests include the


molecular mechanisms underlying tamoxifen action at theestrogen receptor.

V. Craig Jordan, Diana, Princess of Wales Professor ofCancer Research, director of the Lynn Sage Breast Can-cer Research Program at the Robert H. Lurie Comprehen-sive Cancer Center of Northwestern University, serves as theprincipal investigator for a National Cancer Institute SpecialProgram of Research Excellence (SPORE) in breast can-cer (P50 CA89018-02). His research has been recognizedwith the receipt of the American Cancer Society’s Medal of

Honor, and the Dorothy P. Landon AACR Prize in Transla-tional Research with Elwood V. Jensen. Dr. Jordan was ap-pointed Officer of the Most Excellent Order of the BritishEmpire by Her Majesty the Queen for services to Interna-tional Breast Cancer Research in 2002. In 2003, Dr. Jor-dan received the Charles F. Kettering Prize and gold medalfrom the General Motors Cancer Research Foundation foradvances in breast cancer treatment with tamoxifen and thedevelopment of SERMs. Dr. Jordan received his PhD, DScand a Doctor of Medicine degree honoris causa from LeedsUniversity in the UK.

the biological role of estrogen receptors in cancer · physiology of the reproductive process...

Documents