Reviews Critical in Oncology/Hematology 50 (2004) 322
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. 2. 3. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isoforms, domains, ligand binding characteristics and expression of ER and ER ................... 4 4
Transcriptional activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.1. Coregulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.2. Ligands: E2 , antiestrogens, phytoestrogens and subtype-specic ligands . . . . . . . . . . . . . . . . . . . . . . . . 7 3.3. Insight into the molecular basis for ER agonism and antagonism from crystal structures . . . . . . . . . . 7 3.4. ER phosphorylation and ligand-independent transcriptional activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.5. Non-classical pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Normal physiological roles of ER and ER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Tissue distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Knockout mouse studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1. Reproductive phenotypes in ER knockout mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2. Reproductive phenotypes in ER knockout mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3. Reproductive phenotypes in ER /ER knockout mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4. Bone density and cardioprotection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Summary of ndings with ER knockouts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Consequences of an ER mutation in a man . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The role of ER and ER in cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1. ER and breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2. Tamoxifen resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.3. ER and breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Prostate cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Colon cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4. Ovarian cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Current status of the ER and future research directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 10 11 11 12 12 12 12 13 13 13 13 14 14 15 15 15 16 17 21
4.
5.
6.
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract The temporal and tissue-specic actions of estrogen are mediated by estrogen receptors and . The ERs are steroid hormone receptors that modulate the transcription of target genes when bound to ligand. The activity of these transcription factors is regulated by a variety of factors, including ligand binding, phosphorylation, coregulators, and the effector pathway (ERE, AP1, SP1). The end result of target gene
Corresponding author. Tel.: [phone]; fax: [phone]. E-mail address: [e-mail] (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) 322
transcription is to modulate physiological processes, such as reproductive organ development and function, bone density, and unfortunately contribute to the growth and development of breast and endometrial cancer. The complex biological effects mediated by ER and ER involve communication between many proteins and signaling pathways. An ultimate goal of current research is to enhance the value of the separate estrogen 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, and physiology of the reproductive process through the estrogen receptor (ER). E2 also affects other tissues, such as bone, liver, brain and the cardiovascular system. Because of the functional diversity displayed by estrogens through the ER, much of the current interest in understanding the basis of ER actions at the molecular level is focused towards the goal of therapeutic intervention [1,2]. One of the earliest studies reporting a relationship between breast cancer and ovarian hormones described breast tumor regression after removal of the ovaries [3], the major site of estrogen production in premenopausal women. However, only one in three women respond to oophorectomy [4]. The explanation for these observations became clear when the ER was discovered [5]. In the late 1960s and early 1970s, the ER was initially used as a predictor of breast cancer response to endocrine ablation. Tumors that were ER rich were more likely to respond to endocrine therapy than if the tumor was ER poor [6,7]. In the mid 1970s, before adjuvant therapy became the standard of care, the ER was viewed as a prognostic indicator after surgery, with ER-positive patients responding better than ER negative patients [8]. From the 1970s to the present day, the ER has evolved to be the most effective target for breast cancer therapy. Interactions between E2 and the ER can be blocked using a variety of agents. Selective estrogen receptor modulators (SERMs) such as tamoxifen and raloxifene, are competitive inhibitors of E2 at the ER and display agonist or antagonist behavior depending on the tissue [9]. Pure antiestrogens, exemplied by fulvestrant (ICI 182,780), only produce antagonist effects and are proving to be useful in treating advanced breast cancer [10,11]. Aromatase inhibitors, such as anastrozole, that block the conversion of androstenedione or testosterone to estrone and estradiol, respectively, are a particularly interesting new approach to breast cancer treatment as the compounds appear to increase efcacy and reduce side effects compared with tamoxifen [1215]. The optimal combinations and sequential orders of treatment continue to be investigated in clinical trials. Although the primary focus of research for the rst 30 years (19601990) has been on the role of steroid receptors in reproductive functions and breast cancer, there is reason to believe that there are opportunities to design new molecules targeted to novel sites dominated by one ER or the other. This is especially true since the publication of the Womens
Health Initiative did not demonstrate an overall health benet for women taking hormone replacement therapy (HRT) [16]. Positive aspects of HRT include a decrease in the rate of bone density loss, a decrease in total and LDL cholesterol, and a protective effect against colon cancer. However, the risk of breast cancer is increased in HRT users [17]. The challenge now is to dissect the individual roles of ER and ER as transcription factors that participate in normal and abberant physiological processes. Clearly, the goal will be a menu of multifunctional medicines that can be used singly or in combination to treat and prevent a range of diseases associated with menopause or reproductive function.
2. Isoforms, domains, ligand binding characteristics and expression of ER and ER The therapeutic targets estrogen receptors (ER ) and (ER ) are members of the nuclear receptor superfamily of transcription factors. Other members of this family include thyroid receptor, Vitamin D receptor, retinoic acid receptor, and other steroid receptors such as the glucocorticoid receptor, androgen receptor, progesterone receptor and mineralocorticoid receptor. ER was the rst estrogen receptor cloned and it was isolated from MCF-7 human breast cancer cells in the late 1980s [1820]. In accordance with its role as a transcription factor, this 66 kDa ER localizes primarily to the nucleus. A 46 kDa isoform (hER 46) that lacks the rst 173 amino acids of the 66 kDa form of ER has also been preliminarily characterized [21]. In addition, several ER splicing variants have been described [22,23], but whether they are expressed as proteins that have a biological function remains unknown. Ten years later, ER was cloned from rat prostate using degenerate PCR primers [24]. Mouse [25] and human [2628] forms of ER have also been cloned. The human ER gene is located on chromosome 6 and the ER gene is on chromosome 14, demonstrating that they are in fact encoded by separate genes and are distinct [27]. A variety of ER mRNA isoforms have been described in humans, primates, rats and mice [29], but the 530 amino acid form of ER [28] is considered to be the wild type, full length human ER . Because the functional signicance and expression of the various ER isoforms are unclear, the reader is referred to recent reviews [29,30]. Human ER and ER (long form) share common structural domains, which are designated AF (Fig. 1). The A/B
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322 Table 1 The relative binding afnity of various ligands for ER and ER Compound Relative binding afnitya ER E2 4-OHT 1C1164,384 DES Genistein 100 178 85 468 5 ER 100 339 166 295 36
5
Fig. 1. Structure and homology between human ER and the long form of ER . The domains A-F are depicted, as well as the percent identity between the individual domains at the amino acid level. Activation functions 1 (AF1) and 2 (AF2) are also indicated. Adapted from [28].
a The relative binding afnity was calculated as a ratio of concentrations of E2 or competitor required to reduce the specic radioligand binding by 50%. Adapted from [32].
domain contains activation function 1 (AF1), a constitutive activation function contributing to the transcriptional activity of the ER. This domain is one of the least conserved domains between ER and ER , exhibiting only a 30% identity. Based on functional studies, ER has been shown to lack AF1 activity [31]. The DNA binding domain, or C domain, is the most highly conserved region between ER and ER , with 96% identity. This allows both receptors to bind to similar target sites. The D domain, or hinge region, is not well conserved (30%) between the receptors and it contains the nuclear localization signal. Finally, the E/F region encompasses the ligand binding domain (LBD), a coregulator binding surface, the dimerization domain, a
second nuclear localization signal, and activation function 2 (AF2). In contrast to AF1, AF2 is a ligand-dependent activation function. The E/F domains of ER and ER exhibit a sequence identity of 53%. Despite the 53% sequence identity between ER and ER in the ligand binding domain, the two receptors exhibit subtle differences in ligand binding specicity. A KD of 0.6 nM for ER and 0.24 nM for ER for the ligand 16 -iodo-E2 was determined using saturation binding assays, which is similar to the range of E2 binding to ERs (0.11 nM) in various systems [32]. Additional studies using E2 showed that the KD for ER was 0.05 nM and for ER was 0.07 nM [33]. Despite the slight differences in the
Fig. 2. Distribution of ER and ER in the human body. Adapted from [34].
6
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
afnity of ER and ER for E2 , both receptors are considered to have a similar afnity for E2 . However, differences were observed for other ligands, such as antiestrogens and phytoestrogens. A wide variety of structurally distinct compounds bind to the ER with differing afnities. Certain ligands act as ER agonists, and these include the natural ligand E2 , as well as the synthetic estrogen diethylstilbestrol (DES). Certain phytoestrogens, which are environmental compounds produced by plants, can also be estrogenic. Genistein, present in soya beans and soy products, is a widely utilized phytoestrogen. Other compounds, such as ICI 182,780, are receptor antagonists. A nal group of mixed agonists and antagonists are comprised of the SERMs and examples include tamoxifen and raloxifene. The relative binding afnity (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 binding afnities: 4-OHT(339) > DES(295) > ICI 164, 384(166) > E2 (100) > genistein (36) [32]. These differences could result in functional consequences for a receptor subtype. ER ligands interact with ER subtypes in various parts of the human body (Fig. 2). The abundance and distribution of the receptors will, in part, determine whether a ligand will have a particular effect. Using RT-PCR, Northern blot analysis, immunohistochemistry and in situ hybridization techniques, ER and ER are known to be localized in the breast, brain, cardiovascular system, urogenital tract and bone [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 may also localize to distinct cellular subtypes within each tissue. For example, within the ovary, ER is largely present in the thecal and interstitial cells, whereas ER is predominantly in the granulosa cells [36,37]. In the prostate, ER localizes to the epithelium, whereas ER localizes to the stroma [38].
ERligand complexes and translate this information into the coordinated regulation of gene transcription. 3.1. Coregulators The initiation of transcription is complex and requires the interaction of many proteins at a target gene promoter. Transcriptional activation by the ER requires the recruitment of transcriptional regulators, such as general transcription factors, coactivators, corepressors, cointegrators, histone acetyltransferases, and histone deacetylases. (reviewed in [4749]). All of these regulators interact to affect transcription and the accessibility of target gene promoters. Coactivators are proteins that enhance transcription. The contact between coactivators and the ER is made through the LXXLL motif present in the coactivator [50], although the site on the ER required for this interaction varies. Coactivators include steroid receptor coactivator 1 (SRC-1), SRC-2 and SRC-3, which are members of the p160 family. p300 and CREB-binding protein (CBP) are cointegrators, in that they do not themselves bind DNA, but are recruited to promoters by other transcription factors, such as SRC-1. Corepressors decrease transcription and include nuclear receptor corepressor (NCoR) and silencing mediator for retinoid and thyroid hormone receptor (SMRT). Local chromatin structure is remodeled to allow for gene transcription [4749]. Chromatin remodeling factors include ATP-dependent nucleosome remodeling complexes and proteins that contain acetyltransferase activity. Histone acetylation correlates with transcription, whereas deacetylation correlates with gene repression. p300/CBP-associated factor (PCAF), p300/CBP, SRC-1 and SRC-3 contain intrinsic acetyltransferase activity. In contrast, corepressors do not contain histone deacetylase (HDAC) activity, but they recruit other proteins that have HDAC activity. Although the majority of coregulators interact with multiple members of the nuclear receptor family, there are examples of coregulators that interact exclusively with the ER. In a screen to identify proteins that interact with the E/F domain of the dominant negative L540Q mutant ER, the protein repressor of estrogen receptor activity (REA) was isolated [51]. REA is a selective ER and ER coregulator that enhances the inhibitory effectiveness of the L540Q ER and of antiestrogens. In contrast to other corepressors that interact with the unliganded receptor, REA preferentially interacts with the liganded ER. In addition, REA binds to the L540Q mutant ER and antiestrogen-liganded ER more than the wild type ER. REA also competes with the coactivator SRC-1 for modulation of ER transcriptional activity. These studies suggest that when cellular REA levels are high, antiestrogen action will be enhanced [52]. Coactivators can also interact preferentially with a particular activation function region. For example, p68 RNA helicase is a coactivator specic for the ER AF1 region [53]. p68 binds CBP, so p68 may serve as a bridge to associate AF1 with AF2 coactivators. p68 enhanced the transcriptional
3. Transcriptional activity The transcriptional activity of the ER is mediated by AF1 and AF2 (Fig. 1) [3942] and these regions were largely delineated using mutational studies. The activity of AF1 and AF2 differs depending on the cellular environment and promoter context [43]. In some cells, either AF1 or AF2 is dominant, and in others, both activation functions synergize [44]. In addition, AF1 and AF2 are differentially regulated by ligand. E2 is an agonist regardless of whether AF1 or AF2 is dominant. The pure antiestrogen ICI 164,384 blocks both AF1 and AF2, affects dimerization [45] and targets the ER for degradation [46]. Tamoxifen acts by blocking AF2 activity so it is an antagonist in cells where AF2 is dominant and a partial agonist where AF1 is dominant. ER activity is mediated by AF2, since ER does not contain AF1. Cells therefore have the ability to distinguish between different
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
7
activity of the 4-OHTER complex and the phosphorylation of S118 of ER was required for the ability of p68 to enhance transcription. In addition to interacting with both ER and ER or a particular activation function, coactivators can interact selectively with ER or ER . For example, SRC-3 enhances ER and progesterone receptor (PR) stimulated transcription, but has no effect on ER -mediated transcription [54]. Therefore, coregulators provide an additional layer of specicity and regulation to the transcriptional activity of the ER. In addition to being a general ER coregulator, this could be accomplished by targeting ER or ER specically, or interacting with AF1 or AF2. The discovery of new coregulators will continue to dene the selective role of either ER or ER . 3.2. Ligands: E2 , antiestrogens, phytoestrogens and subtype-specic ligands Initial transcriptional activation studies using E2 were performed using ER cloned from rat prostate. These studies showed that an ERE-luciferase reporter was activated in the presence of E2 in CHO cells transfected with ER [24]. Studies in HepG2 and Hela cells showed that both AF1 and AF2 contribute to the activity of ER , but the individual contributions depend on the cell context [31]. In contrast, AF2 mediates the transcriptional activity of ER to E2 , since AF1 is inactive and inhibits the transcriptional activity mediated by AF2. In the presence of sub-saturating amounts of E2 , ER activity was actually suppressed by ER , suggesting that the relative amounts of ER and ER can affect E2 activity [31]. The agonist activity of the antiestrogen tamoxifen has been shown to be dependent on cell type, promoter context, and ER subtype [55]. The agonist activity of tamoxifen appears to be mediated through ER , because no agonism is observed with ER [31,55]. In fact, the addition of ER to ER and tamoxifen inhibited the agonist activity of tamoxifen [31]. The lack of tamoxifen agonism at ER is likely to result from differences in the A/B region, and AF1 in particular, between ER and ER . Replacing the A/B domain of ER with the A/B domain of ER (ER / ) increased the transcriptional activity of the chimera in response to E2 compared to that observed with ER alone [56]. In addition, the ER / chimera displayed a transcriptional response to tamoxifen, which was not observed with ER . These studies further suggested the importance of ER AF1 in mediating the agonist effect of tamoxifen. Therefore, differences in the A/B region between ER and ER are responsible for the cell, promoter and ligand specicity displayed by the estrogen receptors. Phytoestrogens are plant-derived compounds that are consumed in the diet. They contain inherent estrogenic activity or are converted to estrogenic compounds by bacteria in the gut. The rst class of phytoestrogens is the isoavonoids, which are present in soybean products, some fruits and veg-
etables, and red clover. Genistein, dadzein and glycitein are the main dietary-derived isoavones [57]. Many of these compounds have a greater afnity for one ER subtype. For example, the isoavone genistein has a 20-fold greater binding afnity for ER compared to ER , but it activates transcription through ER and ER [33]. The second class of phytoestrogens is the lignans, such as enterodiol and enterolactone, which are present in whole grain cereals, seeds, berries and nuts [58]. Interest in phytoestrogens and their potential role in the prevention of breast, prostate and colon cancer is a result of a higher incidence of these cancers in the western world compared to Asian populations [59]. The Asian diet is largely vegetarian or semivegetarian, which has a higher proportion of phytoestrogens compared to western diets that include more animal protein and fat. In general, phytoestrogens may provide some protection against breast, prostate and colon cancer (reviewed in [57,58]), but more human studies are needed before a denitive conclusion can be drawn. All of the ER ligands described to date can bind to both ER and ER , but with differing afnities. The development of agonists and antagonists that are selective for either ER or ER would be useful tools to analyze the individual role of ER or ER . For example, the R,R-enantiomer of 5,11-cis-diethyl-5,6,11,12-tetrahydrochrysene-2,8-diol (THC) acts as an ER agonist and an ER antagonist and has a sixfold greater afnity for ER over ER [60]. Another example is methyl-piperidino-pyrazole (MPP), a basic side chain pyrazole that has a 200-fold binding selectivity for ER over ER and is an ER selective antagonist [61]. The structure function relationships of ER and ER -specic ligands are reviewed in detail elsewhere [1,2]. 3.3. Insight into the molecular basis for ER agonism and antagonism from crystal structures The crystal structures of ligands complexed with the ER have provided invaluable insight into the structural basis of receptor agonism and antagonism (Fig. 3). The rst studies of the human ER LBD complexed with E2 and raloxifene indicated that although E2 and raloxifene bind at the same site within ER , structural differences resulted. Helix 12 (Fig. 3A, green) is positioned over the ligand binding pocket in the ER E2 complex [62,63], thereby generating a functional AF2 that is able to interact with coactivators. In contrast, because the side chain of raloxifene is too long to t within the binding pocket, it displaces helix 12 (Fig. 3C, compare arrows in 3A and 3C) [62,63]. This prevents the formation of a competent AF2 region. Subsequent studies directly analyzed the interaction of DES (Fig. 3B), 4-hydroxytamoxifen (4-OHT, the active metabolite of tamoxifen [64]) and the coactivator GRIP1 with human ER [65]. Like raloxifene, the bulky side chain of 4-OHT displaces helix 12 (Fig. 3D, arrow) and places it in a position where it mimics bound coactivator.
8
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
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 been studied in complex with ICI 164,384 [67] and raloxifene [66]. The individual helices are numbered and the ligand is depicted as a space lling model. For each 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 structure of the ligands is also shown. Reprinted from [63,65,67] with permission from Elsevier and from [66] with permission from Oxford University Press.
The rst structural description of ER was in complex with raloxifene or genistein [66]. The structure of the ER LBD was similar to the previously reported ER LBD structure. In addition, the human ER -genistein (Fig. 3E) and the rat ER raloxifene (Fig. 3G) complexes are quite similar. As in the ER -4-OHT structure, the side chain of raloxifene displaces helix 12 and prevents it from sealing the ligand in the ligand binding pocket (Fig. 3G, arrow). In contrast, genistein binds to ER in a manner similar to E2 . The main difference is that instead of helix 12 assuming the typical agonist position that E2 induces, helix 12 is in a more antagonist position (Fig. 3E, arrow). Genistein is a partial agonist
at ER and coactivators must displace helix 12 into a more agonist conformation before activating transcription. The only crystal structure of the pure antiestrogen ICI 164,384 with an estrogen receptor is with the rat ER LBD (Fig. 3F) [67]. The side chain of ICI 164,384 is longer than that of raloxifene. For the side chain of ICI 164,384 to achieve the same position that the side chain of raloxifene occupies, the steroidal core is ipped 180o around its longest (hydroxy-to-hydroxyl) axis. The position of the side chain exposes a large hydrophobic patch on the surface because the side chain of ICI 164,384 binds along the coactivator recruitment site. This abolishes the interaction of helix 12
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
9
with the LBD such that it is disordered. Therefore, helix 12 cannot move into position to form a coactivator recruitment site. This overall conformation could favor the recruitment of corepressors and resemble misfolded or denatured proteins, thereby targeting the ER for degradation. In summary, when an agonist binds to the ER, a conformational change occurs that forms the AF2 coactivator binding site. However, the large side chain of antagonists such as tamoxifen, raloxifene, or ICI 164,384 protrudes from the binding cavity and displaces helix 12, thereby disrupting AF2. The molecular events resulting from ligand binding subsequently translate into agonism or antagonism at the ER (for review, see [63]). However, it is important to note that only the ER LBD has been used to generate these crystal structures, so the potential structural importance and interaction of the remaining part of the ER with the LBD has yet to be determined. In addition, a second binding site has been identied in the LBD of ER and ER [68]. Although multiple tetrahydrochrysene derivatives were docked into this second site, the classical steroid binding site was still the preferred site. Future studies could determine whether these multiple binding sites have consequences for the activity of the ER. 3.4. ER phosphorylation and ligand-independent transcriptional activity In contrast to the ligand-mediated transcriptional activity or genomic effects described above, the activity of the ER can also involve ligand-independent or non-genomic effects. This is based on reports showing that many effects induced by E2 occur within a short time frame of seconds to minutes, which is faster than transcriptional events. These rapid effects may be mediated in part by plasma membrane associated forms of the ER [6971]. Interestingly, ER and ER can both localize to the membrane [72]. The ligand-independent activity of the ER is a result of phosphorylation of the ER (Fig. 4) and creates cross-talk between the ER and other signaling pathways [73,74]. Phosphorylation of the ER is largely on serine residues in the
AF1 region of ER . The rst reports showed that E2 induced phosphorylation only on serine residues in MCF-7 cells [75]. In addition, using a transient transfection system in COS-1 cells, the ER was shown to be phosphorylated only on serine residues by E2 , as well as other ligands such as 4-OHT and ICI 164,384 [76]. More specically, E2 , 4-OHT and ICI 164,384 induced ER phosphorylation at S118 [77]. Further evidence indicated that S118 was a major site of phosphorylation 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 phosphorylates S118 and the possibility exists that multiple kinases carry out this phosphorylation. S118 is a target of MAP kinase in vitro and in response to EGF or insulin-like growth factor (IGF) treatment in vivo [80]. Other reports have shown that MAP kinase does not phosphorylate S118 in MCF-7 cells in response to E2 , but EGF and PMA treatment result in phosphorylation of S118 via MAP kinase [81]. Further information suggests that MAP kinase phosphorylates S118 independent of ligand, whereas Cdk7 phosphorylates S118 in response to E2 in COS-1 cells [82]. Serine residues in ER other than S118 are also phosphorylated. In response to E2 , S167 is the major phosphorylation site in recombinant ER expressed in Sf9 insect cells and MCF-7 cells and it is phosphorylated by casein kinase II [83]. Upon EGF or phorbol myristate acetate stimulation, S167 is a target of pp90rsk1 , which is phosphorylated by MAP kinase [84]. Phosphorylation of S167 has also been implicated in the phosphatidylinositol-3-OH kinase (PI(3)K)/Akt pathway. E2 stimulates ER binding to the 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 associated with p85 because this interaction is not affected by E2 [86]. The end result is that Akt phosphorylates S167 of ER and the consensus site is not present in ER [87]. The phosphorylation of S104 and S106 is mediated by the cyclin A-CDK2 complex in U-2 OS human osteosarcoma
Fig. 4. Kinases and ligands induce ER phosphorylation at specic sites. Phosphorylation of the ER occurs within AF1 at S104/S106, S118 and S167 by a variety of ligands and kinases as indicated. S236, in the DNA binding domain, is phosphorylated by PKA and Y537 is phosphorylated by src family kinases.
10
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
cells [88]. S236 in the DNA binding domain, is phosphorylated by protein kinase A and this phosphorylation regulates dimerization [89]. Tyrosine phosphorylation has been detected at Y537 in MCF-7 and Sf9 cells [90] and p60c-src and p56lck mediate this phosphorylation event. Y537 is not phosphorylated by E2 treatment, indicating that it is a basal phosphorylation site. Although it is clear that many studies have focused on the role of phosphorylation of ER , few have studied ER . The role of phosphorylation in the interaction between ER and coactivators indicated that the interaction between mouse ER and SRC-1 increased in the presence of E2 [25]. Phosphorylation of the unliganded receptor by MAP kinase at S106 and S124 (S118 in human) in AF1 enhanced the recruitment of SRC-1 and therefore the activity of ER [91]. The role of phosphorylation in the activity of ER remains to be elucidated. Activation of ER via phosphorylation at multiple sites (S104, S106, S118, S167, S236, Y537) by multiple kinases is important because of the interaction between growth factor signaling and the ER. Increased growth factor signaling may account for the loss of E2 dependence, thereby producing antiestrogen resistant tumors. In addition, an association has been observed between elevated MAP kinase phosphorylation/activity and a poor response to endocrine therapy in breast cancer patients [92]. Although the precise relationship between ER phosphorylation and clinical outcome remains to be elucidated, the ER phosphorylation state has the potential to be a predictive biomarker and intervention target. 3.5. Non-classical pathways The classical scheme of ER action involves ligand binding to the ER, dissociation of heat shock proteins from the ER and receptor dimerization. The ER dimer then interacts with coregulatory proteins, binds to DNA sequences termed estrogen response elements (ERE) that are located in the regulatory regions of responsive target genes, and transcription is activated. However, the ER can also mediate transcription via tethered interactions through proteinprotein interactions at AP1 (reviewed in [93]) and Sp1 sites. In Hela cells transfected with ER and an AP1 reporter, E2 , DES, raloxifene, tamoxifen and ICI 164,384 stimulated reporter activity to varying degrees [94]. The amount of stimulation varied depending on the cell type [95]. In contrast, ER activated the AP1 reporter in the presence of raloxifene, tamoxifen and ICI 164,384, but not with E2 and DES. ER and ER therefore respond differently to estrogens and antiestrogens at AP1 sites. The regions of the ER that are required for the stimulation of AP1-mediated transcription vary depending on the cell type and ligand [9597]. The ER also activates transcription of target genes through ERSp1 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 and transforming growth factor [98]. ER and ER have been shown to bind to the C-terminal domain of the Sp1 protein [99]. Transient transfections of MCF-7, Hela and MDA-MB-231 cells with a Sp1 reporter and ER or ER showed varying patterns of activation by estrogen and antiestrogens. In MCF-7 and MDA-MB-231 cells, E2 , 4-OHT and ICI 182,780 activated Sp1 through ER . In contrast, no changes were observed in Hela cells with any ligand. 4-OHT activated Sp1 through ER in MCF-7 cells, but no changes were observed by any ligand in MDA-MB-231 cells [99]. All of the ligands decreased Sp1 reporter activity in Hela cells in the presence of ER . Additional results [99] suggested that the relative amounts of ER and ER present in a cell inuence Sp1 activity. Amino acids 79117 in AF1 are important for the interaction of ER with Sp1, and this region could mediate an association with other proteins that are important for the ER /Sp1 mechanism. In summary, cell-specic regulation occurs as a result of multiple factors, including coregulator expression and recruitment, the ratio of ER to ER , the nature of the ligand, ER phosphorylation and the pathway activated (ERE, Sp1 or AP1). The activity of ER and ER is also complicated by the fact that they can form functional homo and heterodimers [28,31,100,101], which may activate different target genes. The inherent structural differences in the A/B domain between ER and ER , where ER lacks a functional AF1 domain, may result in a large effect on the activation proles of target genes, especially when coregulators that interact preferentially with the AF1 domain are considered.
4. Normal physiological roles of ER and ER 4.1. Tissue distribution Analysis of the tissue distribution of ER and ER provides insight into the potential for targeting specic tissues. The relative distribution of ER and ER mRNA was initially determined in rat tissues using RT-PCR [32]. ER mRNA was highly expressed in epididymis, testis, pituitary gland, ovary, uterus, kidney and adrenal. Moderate amounts were also present in the prostate gland, bladder, liver, thymus and heart. Highest amounts of ER mRNA were detected 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 levels of ER , whereas low but detectable levels of ER were observed in epididymis, the pituitary gland, thymus, various brain sections and spinal cord. Target tissues with higher expression levels of ERs are predicted to be more affected by ER ligands. The expression in rat is similar to that observed in humans (see Section 2) ([35] and references therein).
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
11
4.2. Knockout mouse studies Transgenic mice are valuable experimental models to elucidate gene functions. Knockout mice, where genes of interest have been deleted, provide a basic insight into the normal functions of genes during development and at maturity. The classical scheme of reproductive organ development is that female reproductive tract development is the default pathway and that estrogens are not required for the initial differentiation and development of the female reproductive system. In contrast, testosterone is essential for the proper development of male structures. The potential role of estrogen in male development has remained unclear. Knockout mouse models have been utilized to analyze the role of ER and ER in the general development and physiology of the mouse (Table 2). These models include ER knockout mice ( ERKO), ER knockout mice ( ERKO) and both ER and ER knockout mice ( ERKO) (reviewed in [102105]). It is interesting to note that a loss of either of ER and/or ER is not lethal, and the mice survive to adulthood. 4.2.1. Reproductive phenotypes in ER knockout mice ER knockout mice show no abnormal external phenotypes. However, the most striking phenotypes occur in the tissues that predominantly express ER , such as the uterus and mammary gland, but defects are also observed in the ovary and in sexual behavior. One role of the ovary is to act as an endocrine organ by providing sex steroids to the female and is essential for proper reproductive functions. The ovaries in ERKO mice contained cystic and hemorrhagic follicles that contain no corpora lutea and few granulosa
Table 2 Phenotypes of ER knockout mice ERKO no ER , ER dominant Female Reproductive tract Infertile Hypoplastic uterus; cystic and hemorrhagic follicles; no corpora lutea; few granulosa cells in ovary Develop to a newborn structure; no pubertal development No lordosis; no receptivity Male Reduced fertility Low sperm count; low sperm motility; low testis weight
cells [106]. Adult females have hypoplastic uteri and showed no responses to estrogen, such as increases in uterine wet weight, hyperemia, or the alteration of vaginal epithelial cell morphology [106]. Females exhibit little sexual behavior in that they display no lordosis posture or receptivity, indicating a lack of estrogen responsiveness in the central nervous system [106]. As a result of these phenotypes, female mice lacking ER are infertile. The mammary glands of the adult ERKO look essentially like those of a female mouse before puberty, indicating that rudimentary mammary glands can develop independent of estrogen and ER , but a fully differentiated gland requires ER [107]. ERKO mice were further used to study the potential protective effect of dietary genistein on mammary tumor development. The rationale for this study is the suggestion that genistein could be protective against breast tumors. Because genistein has a greater binding afnity for ER when compared to ER , genistein is predicted to act through an ER pathway. ERKO mice or mice containing wild type ER were fed genistein and treated with dimethylbenz[a]anthracene. Tumors formed in the mice with wild type ER , whereas no tumors were present in the ERKO mice [108]. This indicated that genistein did not provide protection against tumor formation in the wild type mice and that it could actually result in tumor formation. In situations where 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 high expression levels of ER in the mammary gland, point toward ER being the main mediator of estrogen action in the mammary gland.
ERKO no ER , ER dominant Female Reduced fertility Normal uterus; many early atretic follicles; Fewer corpora lutea in ovary Male Fertile Epithelial hyperplasia in collecting duct of prostate and bladder wall
ERKO no ER or ER Female Infertile Hypoplastic uterus; develop structures similar to male seminiferous tubules (sex reversal) in ovary Male Infertile Low sperm number; low sperm motility
Mammary glands
N/A
Normal
N/A
N/A
Sexual behavior
Bone
Decreased density, diameter and length
Normal mounts; reduced intro-missions; rarely ejaculates Decreased density, diameter and length
Normal
Normal
No mounts, intro-missions or ejaculation
Normal
Normal
N/A: not applicable.
12
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
Male ERKO mice show decreased sperm motility [109], a low sperm count, and low testis weight, resulting in reduced male fertility [106]. In terms of sexual behavior, male ERKO mice show normal levels of mounts, but also reduced intromissions, and rare ejaculations [110], which also contribute to the reduction in fertility. 4.2.2. Reproductive phenotypes in ER knockout mice ER knockout mice have also provided additional information as to the function of ERs in the mouse. The most obvious phenotype occurs in the ovary, which is a tissue that contains the greatest expression of ER . Ovaries in females lacking ER have more early atretic follicles and fewer corpora lutea when compared to wild type females [111]. Further experiments indicate a partial arrest of follicular development and a decrease in the frequency of follicular maturation [111]. These female mice show normal sexual behavior [112] and normal mammary gland structure. In addition, ER knockout females have fewer litters and fewer pups per litter. Therefore, ER knockout mice show reduced fertility. In contrast to ER knockout females, ER knockout males are fertile and show normal sexual behavior [112]. However, older males show epithelial hyperplasia in the collecting duct of the prostate and the bladder wall [111]. ERKO mice were developed by another group that exhibited many of the same phenotypes, except that no prostate or bladder hyperplasia was observed [113]. The reason for these differences is unknown. ERKO mice also display abnormalities in the brain, such as regional neuronal hypocellularity, which progressed to a degeneration of neuronal cell bodies with age [114]. 4.2.3. Reproductive phenotypes in ER/ER knockout mice To create a complete picture of ER function in male and female development, mice with both ER and ER knockouts ( ERKO) were developed. As with the ER and ER single knockouts, ERKO mice survive to adulthood and exhibit no abnormal external phenotypes [115]. Young ERKO females show proper differentiation of the uterus, vagina and cervix. However, once the mice reach 2.57 months of age, uterine hypoplasia is observed. 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 occurs in the adult ERKO ovary, where structures reminiscent of male seminiferous tubules of the testis were observed [115]. These structures were not present in the prepubertal ovaries. This is the rst example of sex reversal in an adult mouse gonad because the female ovarian cells are able to re-differentiate to a male Sertoli cell phenotype. ERKO males are infertile and show an 80% reduction in epididymal sperm number and a 5% decrease in sperm motility [115]. They also show no components of sexual behaviors [116].
4.2.4. Bone density and cardioprotection An established association exists between the declining levels of estrogen at menopause and the development of osteoporosis, implicating estrogen in the maintenance of bone mass. ERKO females and males showed decreases in 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 after menopause. As a result of the suggestion that estrogen might be cardioprotective, there has been an interest in studying the role of the ERs in the cardiovascular system. Indeed, the publication of the negative results of the Womens Health Initiative [16] has increased interest in this aspect of female physiology to discover mechanisms of physiological benet that can be separated from disadvantageous side effects such as fatalities from cardiovascular complications. Models of carotid injury have been developed in knockout mice to study the individual contribution of ER and ER in cardioprotection. In wild type and ERKO mice, E2 treatment inhibited the increase in carotid medial vessel wall area and vascular smooth muscle cell proliferation normally observed in injured carotid arteries [117]. Because E2 inhibits markers of vascular injury in the ERKO mice, ER is clearly not required for this process. Similar results were observed in the ERKO mice [118], suggesting that ER and ER are individually redundant in mediating the vasoprotective effects of E2 . In contrast, E2 did not protect against the increase in carotid medial wall area after injury in ERKO mice, but did inhibit vascular smooth muscle cell proliferation [119]. Unfortunately, the specic batch of ERKO mice exhibited uterine weight increases in the presence of E2 , suggesting residual ER activity. In fact, it has been suggested that the original ERKO mice [106] express a smaller ER transcript [120] that may have some functional activity [119,121]. As a result, these data [119] should be interpreted with caution. Nevertheless, overall studies in ER 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 in maintaining the vasculature by mediating increases in nitric oxide synthase, which ultimately results in vasodilation. Wild type male mice have increased basal levels of endothelial nitric oxide in the aorta when compared with male ERKO mice [122]. ERKO mice also display increases in L-type Ca2+ channels in the heart, which could lead to abnormalities in cardiac excitability [123]. These studies suggest that ER is involved in cardiac modulation through the regulation of nitric oxide synthesis as well as cardiac Ca2+ channel expression. 4.3. Summary of ndings with ER knockouts Overall, these transgenic mice studies provide insight into the role of estrogen receptors in the development and function of reproductive structures. In the case of the male, ERs
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
13
are not necessary for the development of reproductive structures. However, ER is required for male fertility, because ER loss results in a low sperm count and a decrease in sperm motility. As would be expected, ER loss has more of an impact on the female. Surprisingly, early differentiation of the reproductive tract occurs in the absence of ER and ER . Later in life, both ERs are necessary for proper ovarian function, whereas ER is critical for uterine physiology. ER , but not ER , is required for the proper development of the mammary gland as well as the maintenance of bone. In addition, ER and ER may play redundant roles in E2 -mediated cardioprotection. 4.4. Consequences of an ER mutation in a man Although only one example of an ER germ-line defect in humans has been described thus far [124], comparisons can be made to the phenotypes observed in knockout mice. A man was shown to have a cytosine to thymine substitution at codon 157 in ER , resulting in a premature stop codon. The translated protein would therefore be truncated and lack the DNA and hormone binding regions. The patient presented with osteoporosis, unfused epiphyses and elevated serum estrogen, among other phenotypes. These results provide evidence for the critical role of estrogen in bone development and mineralization as well as epiphyseal maturation in males. Cardiovascular function was also impaired, in that signs of early atherosclerosis were detected [125] and ow-induced dilation of the brachial artery was impaired [126].
Table 3 ER and PR status and response to endocrine therapy Status Response Number ER+ ER+PR ER+PR+ ER PR+ ERPRERPR+ PR Adapted from [128]. 303/571 80/248 223/323 32/239 233/354 22/208 10/31 102/456 (%) 53 32 69 13 66 11 32 22
5. The role of ER and ER in cancer The analysis of knockout mice has provided a framework in which to study the potential functions of ER and ER in human target tissues. Phenotypes of ERKO mice have pointed toward the importance of ER in the uterus and mammary gland of females. In addition, ERKO mice have suggested an important function for ER in the ovary in females and in the prostate gland in males. The laboratory studies in mice naturally advance the study of the complex role of the individual ERs in human cancer. 5.1. Breast cancer 5.1.1. ER and breast cancer The ER is an important target to develop drugs for the treatment and prevention of breast cancer [127]. The interaction of estrogen with the ER can result in increased proliferation of target cells so the rationale for endocrine therapy is to block the interaction of estrogen with the ER. This goal can be accomplished by blocking the production of estrogen by ovariectomy, or inhibiting the conversion of steroidal precursors to estrogen using aromatase inhibitors. The ER can also be targeted directly using SERMs such as tamox-
ifen and raloxifene as competitive inhibitors of estrogen action, or by the removal and degradation of the ER by pure antiestrogens such as ICI 182,780 (fulvestrant). Endocrine manipulations are among the least toxic and most effective therapies for the treatment of hormone responsive breast cancers. In the clinic, factors such as ER, PR and nodal status have historically predicted response to endocrine therapy (Table 3). Patients that are ER+ show a 53% objective response rate to endocrine therapy, and this can be 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 ERPR+ and 11% of the ERPR group exhibiting a response. 66% of patients with PR+ tumors will respond, versus 22% with PR tumors. The measurement of receptor status has changed from the ligand binding assay to immunohistochemical methods. Nodal positivity is another predictor of response to endocrine therapy. Tumors that contain positive nodes correlate with a lower disease free survival (DFS) [129]. If a tumor population is node positive or node negative, the presence of the ER correlates with better DFS. Current treatment strategies have shown that 5 years of adjuvant tamoxifen treatment is benecial in pre- and postmenopausal women with ER-positive tumors [130]. In addition, tamoxifen can be used for the prevention of breast cancer [131]. SERMs act as estrogens in select target tissues but act as antiestrogens in other target tissues [132]. The ideal SERM would be an estrogen agonist in bone, liver, the cardiovascular system and brain and an estrogen antagonist in the breast and uterus. The activity of tamoxifen is dependent on circulating levels of E2 , which are high in premenopausal women and low in postmenopausal women. Tamoxifen is an antiestrogen in the breast, and decreases low density lipoprotein cholesterol levels in postmenopausal women. Tamoxifen treatment decreases bone density in premenopausal women but increases bone density in postmenopausal women [133]. Estrogenic activity of tamoxifen is observed in the uterus, which results in an increased incidence of endometrial cancer in postmenopausal women [131]. The molecular basis for the tissue specicity of tamoxifen is not well understood, but a possible explanation involves the interaction of the ERtamoxifen complex with coregulators. In certain
14
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
breast cancer cell lines where tamoxifen is an antagonist, tamoxifen recruits corepressors; however, in endometrial cells where tamoxifen is an agonist, tamoxifen recruits coactivators [134]. SRC-1 is necessary for the agonist activity of tamoxifen in endometrial cells [134]. These ndings suggest that differences in coregulator recruitment to a promoter can determine the functionality of the ER in different tissues. Changes in coregulator levels affect target gene expression and may change the gene activation prole to one supporting a proliferative phenotype. This molecular mosaic appears to correlate with clinical outcomes. Studies analyzing SRC-1 and tamoxifen showed that high SRC-1 levels may correlate with a favorable response to tamoxifen treatment in women with recurrent breast cancer [135]. Gene amplication and overexpression of amplied in breast cancer-1 (AIB1/SRC-3) has been documented in breast and ovarian cancer cell lines and breast cancer biopsies [136,137]. Patients with ER-positive breast tumors expressing high levels of SRC-3 and HER2 display a poor outcome with tamoxifen therapy [138]. In contrast, high SRC-3 levels are associated with better prognosis in patients not receiving adjuvant tamoxifen. The authors hypothesis is that increased HER2 signaling results in activation of SRC-3 and the via ER phosphorylation, which results in tamoxifen resistance. Clearly, the strategic goal of establishing the importance of both ER and ER in the outcomes of endocrine therapy should be addressed and resolved using the tissue resources of the ATAC trial of tamoxifen versus anastrozole [12]. 5.1.2. Tamoxifen resistance Despite over 30 years of clinical experience with tamoxifen, for most patients, tumors that initially regress with tamoxifen will eventually recur and require alternate treatment. The mechanisms of cellular resistance to tamoxifen are under investigation (reviewed in [139142]). Most resistant tumors continue to retain a functional ER [143], so loss of the ER is not sufcient to explain resistance. Because a functional ER is present, the cells have changed how the ERtamoxifen complex is perceived and how it signals, or there is altered expression of genes to counteract tamoxifen signaling. During the development of resistance, tamoxifen could become an estrogen agonist by inducing E2 -specic genes [144,145]. Tamoxifen resistance could also be explained by ER mutations, coregulator expression and recruitment, or interactions with other signaling pathways. Additionally, a number of nonspecic mechanisms may contribute to the tamoxifen resistant phenotype. The mechanisms may include events that limit the intracellular availability of tamoxifen, such as binding to other proteins, partitioning into lipophilic membrane domains, altered transport into or out of the cell, the development of oxidative stress or the conversion of tamoxifen to other metabolites. Other mechanisms include overexpression of growth factors, increased angiogenesis or heterogeneity in the tumor cell population (reviewed in [146]).
Mutations in ER have been demonstrated in breast tumors, but there is no reason to believe they play a major role in tamoxifen resistance at this point. Nevertheless, if these mutations develop during the course of tamoxifen therapy, the tumor cells could begin to respond differently to tamoxifen. The K303R mutation results in increased sensitivity to estrogen [147]; however, tamoxifen is still effective in blocking estrogen action. In addition, the Y537N mutation results in constitutive activity of the ER that is unaffected by E2 , tamoxifen and ICI 164,384 [148]. The D351Y mutation was discovered in tamoxifen stimulated MCF-7-derived breast tumor models [149]. Despite the fact that the mutation enhances the estrogen-like action of SERMs [150,151], there is no general enhancement of mutations in tamoxifen resistant disease [152156]. Knowledge of the D351Y mutation in ER and the close association of D351 with the antiestrogenic side chain of SERMs have provided an important insight into the molecular modulation of the SERMER complex [12,157]. The balance between ER and ER may also play a role. ER mRNA was shown to be upregulated in tamoxifen resistant human tumors and cell lines, suggesting that ER is a poor prognostic factor for the development of drug resistance [158]. Since the target of SERM action is the ER, it is clear that increases or decreases in coregulatory molecules will modulate the SERMER complex to be estrogenic or antiestrogenic, respectively. This principle is illustrated by the nding of elevated levels of the coactivator SRC-3 in tumors that fail on tamoxifen, but only in the presence of cell surface signaling [138]. It is known that repression of gene activation by tamoxifen is an active process, where tamoxifen recruits corepressors [159]. If corepressor levels are low in a particular tumor, tamoxifen may be unable to recruit a sufcient amount of corepressors to silence gene transcription, thereby contributing to drug resistance. Alterations in signal transduction pathways are another mechanism of tamoxifen resistance. Examples include resistance to the growth inhibitory effects of TGF- 1 [153], enhanced AP1 signaling [160163] and upregulation of Akt/PI3K [86,87], HER2 [164166], IGF-1 receptor [167], and active MAP kinase [168]. Activation of these pathways can lead to ligand-independent activation of the ER via phosphorylation. Overall, there are numerous potential mechanisms that may contribute singly or in combination to the development of drug resistance. Despite the possibility of drug resistance, there are potential treatments after the development of tamoxifen resistance. These include the use of aromatase inhibitors to block the production of estrogen [12] as well as pure antiestrogens to degrade the ER [11]. 5.1.3. ER and breast cancer The 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 because it is expressed in normal and malignant breast tissue,
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
15
binds 17 -estradiol and ER can heterodimerize with ER [28,31,100,101]. One problem is that most of the ER analyses have been at the RNA level using PCR based techniques. This is because the analysis of ER protein has been difcult with currently available antibodies as they continue to yield inconsistent results. Further, large studies comparing the distribution of ER and ER using multiple antibodies in multiple tissues are required to build on the preliminary ndings of Taylor and Al-Azzawi [35]. The fact that there are many different antibodies of doubtful value could lead to inconsistencies in the literature. For example, one study indicates that ER is a good prognostic indicator for breast cancer. Expression of ER was associated with better survival in patients receiving adjuvant tamoxifen [171]. Another study shows that ER is associated with negative axillary node status and low grade tumors [172]. In addition, ER cases had a better disease free survival rate [173] and levels of ER were decreased in proliferative preinvasive tumors [174]. These studies suggest a protective role for ER in breast cancer. In contrast, evidence also suggests that ER is a poor prognostic indicator. Tumors that expressed both ER and ER were node positive and of a higher grade [175]. ER mRNA levels were also elevated in tumors that displayed tamoxifen resistance [158]. Overall, the majority of studies suggest that the presence of ER is a good prognostic marker for breast cancer. However, the relative amounts of ER and ER must be considered. As normal breast tissue becomes tumorigenic, the amount of ER increases whereas the amount of ER decreases [176]. The majority of ER present in breast tumors is ER so the biological relevance of ER in breast cancer remains a topic of debate. Nevertheless, large studies that can correlate tumor characteristics with precise determinations of ER mRNA and protein are needed to identify specic situations where ER may be a critical player in either carcinogenesis or disease progression. 5.2. Prostate cancer Several lines of evidence suggest that ER could be involved in prostate cancer. Most importantly, ER is expressed at high levels in the prostate [32]. Within the prostate, ER localizes primarily in the epithelium, whereas ER is in the stroma [38]. ER has also been detected in normal and malignant prostate tissue. Because the prostate gland expresses a large amount of ER , it may be more susceptible to the effects of environmental estrogens [177]. Secondly, one report analyzing ER knockout mice showed that these mice displayed prostatic hyperplasia [111], whereas no evidence of hyperplasia was observed in the ERKO mice. This suggests that ER may protect against abnormal growth in the prostate. Although the role of ER in the prostate remains unclear, most ndings support an antiproliferative role. If this scenario is proven correct, ER selective ligands could potentially be used for prostate cancer therapy.
5.3. Colon cancer The distribution of ER and ER has been evaluated in colon cancer cell lines as well as human colon cancer samples. ER is present in the human colon cancer cell lines HCT116, HCT8, DLD-1, LoVo, HT29, Colo320, SW480 and Colo205 but ER is not present [178180]. Studies on human samples showed that ER protein was expressed at extremely low levels in normal and malignant colon tissue compared to ER levels [181]. Although no differences were observed in ER mRNA levels using RT-PCR between normal and malignant colon tissue, a loss of ER protein was observed during malignant transformation [181]. In addition, the localization of ER in normal colon was nuclear, whereas a cytoplasmic localization was also observed in colorectal carcinoma tissue [182]. Therefore, multiple studies support the idea that ER is the primary ER expressed in the colon, and that the loss and change in localization of ER is associated with the progression to cancer. The recent results of the Womens Health Initiative (WHI) trial have provided an interesting insight on the role of estrogen in colon carcinogenesis [16]. The primary goal of the WHI was to evaluate the use of hormone replacement therapy (HRT, estrogen plus medroxyprogesterone acetate) in postmenopausal women aged 5079 for the prevention of cardiovascular disease. Other outcomes were also documented. The HRT group had a 26% increase in breast cancer rates, which was anticipated [17,183]. Colorectal cancer rates were reduced by 37%, which was also suggested in previous studies [184]. The decreased incidence of colon cancer could be mediated by ER . In women, ER mRNA levels have been shown to be decreased in colon tumors compared to normal tissue, whereas ER levels did not change and are much lower than ER levels [185]. This evidence suggests that the activation of ER in the colon by HRT (estrogen) provides protection against colon cancer. This is similar to the situation observed in prostate cancer, where ER is thought to play a protective role. 5.4. Ovarian cancer Although approximately two-thirds of ovarian cancers are ER positive, responses to endocrine therapy are modest [186]. However, contraceptives that combine estrogen and progestins decrease the risk of ovarian cancer, such that 5 years of contraceptive use confers a 50% risk reduction that persists for at least 10 years after the cessation of use [187]. ER is the predominant ER in the ovary, where it is found in the granulosa cells, whereas ER localizes to the thecal and interstitial cells. Knockout mouse studies have shown that ERKO mouse ovaries contained cystic and hemorrhagic follicles that contained no corpora lutea and few granulosa cells [106]. Ovaries in ERKO mice have more early atretic follicles, fewer corpora lutea, a partial arrest of follicular development and a decrease in the frequency of follicular maturation [111].
16
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
Studies of normal and malignant human ovaries have yielded conicting results. One study showed an increase in ER mRNA relative to ER in ovarian cancer compared to normal ovary [188]. Another study showed varying amounts of ERs in normal ovary, lower levels of ER in ovarian epithelial primary tumors, and only ER in metastatic tumors [189]. ER levels were lower than ER levels in ovarian cancer compared to normal tissue [190]. In contrast, a decrease in ER mRNA was observed relative to ER in human ovarian surface epithelial cells [191]. The majority of studies therefore support a scenario in which ER becomes the dominant ER in ovarian cancer. This implies a mechanism that results in ER overexpression or a selective growth advantage for ER -positive cells. Further studies are needed to fully determine the contributions of ER and ER to ovarian cancer. Overall, ER appears to play a protective role against the development of breast, prostate and colon cancer. A variety of mechanisms supporting an antiproliferative role for ER have been proposed to explain these ndings. In ER +/ER colon cancer cell lines, estrogen has no effect on cell growth, but genistein slightly inhibited cell growth [179]. In addition, transfections of the cDNA for ER inhibited the growth, invasion and motility of the MDA-MB-231 breast cancer cell line [192]. These data support a scenario in which activation of ER -mediated pathways is able to suppress cell growth. Another possibility is that the presence of ER could simply antagonize the growth stimulatory effects mediated by ER . This is suggested by a study in which ER inhibited the agonist activity of ER tamoxifen complex [31]. Further studies are needed to dissect the precise interactions between ER and ER in cell growth control.
6. Current status of the ER and future research directions It is clear that ER and ER are extremely important components of a complex signal transduction pathway that specically regulates the growth and development of target tissues and tumors. At the molecular level, ERs act as transcription factors to target a variety of genes using the classical ERE pathway or tethering mechanisms utilizing AP1 or SP1. Usually, transcriptional activity is in response to endogenous ligands such as steroidal estrogens or other ligands such as antiestrogens or phytoestrogens. The transcriptional activation of the ER results in the activation of target genes that are involved in normal physiological processes such as the maintenance of bone density, proper reproductive organ development, fertility and behavior. Abberant roles for ER have been demonstrated in breast cancer and ER could be involved in prostate cancer and colon cancer. Knowledge of the role of estrogen in physiology and pathology has resulted in the development of effective and relatively safe drugs that target endocrine-related breast cancer, postmenopausal osteoporosis and resulted in re-
cent advances toward the prevention of breast cancer [1,2]. The developing molecular knowledge of estrogen action can be further exploited to design better drugs to target ER and ER selectively. By way of a recent example, 2,3-bis(4-hydroxyphenyl) propionitrile (DPN), an ER selective ligand, has a 70-fold higher relative binding afnity and a 170-fold higher potency for ER over ER in transcription assays [193]. Propyl pyrazole triol (PPT) is an ER selective ligand, with a 400-fold afnity for ER over ER [194]. PPT treated rats exhibited uterine weight gain, increased complement C3 mRNA in the uterus and increased progesterone receptor mRNA in the brain. Therefore, ER and ER selective ligands could be utilized to delineate the specic physiological processes mediated by ER and ER . This knowledge can be further exploited to develop ER subtype-specic therapeutic drugs. The application of subtype-specic drugs for therapeutic use requires consideration of ER subtype expression in target tissues. For example, the cellular environment of breast cancer is comprised of high levels of ER and extremely low levels of ER . Because estrogen is a growth stimulus through ER , the use of aromatase inhibitors or ER -specic antagonists to block this interaction may be optimal. Since evidence [171174] indicates that ER is protective in the breast, the combined use of an ER agonist could provide further benet. A different scenario occurs in the prostate gland and colon, where ER is the dominant ER and little or no ER is present. ER is viewed as protective in these tissues, so the use of an ER agonist may provide the most benet. However, the situation becomes more complicated in tissues that contain high levels of both ER and ER , because the interaction that will occur by activating or blocking both ERs separately must be deciphered. The issue of receptor interaction is extremely difcult to address in tissues throughout the body. Nevertheless, one issue that has important physiological implications is whether even a small amount of ER can cause a signicant response in the presence of a large amount of ER . The question to be answered is how much of an ER is required to elicit a response. In other words, if a large amount of ER is present, can the activation of even a small amount of ER cause a signicant response through alternate signal transduction pathways? If future strategies to design ER subtype-specic drugs are pursued, then the physical chemistry of the agents will be extremely important. It is extremely difcult to control the pharmacokinetics and pharmacodynamics of orally active medicines, especially when the compliance of patients is extremely variable. Treatment strategies involve blocking the activation of a signal transduction pathway, but in the future, a specic ER agonist may be used that inadvertently is detrimental to some unknown target. Thus, the development of a selective ligand with a high afnity for one ER subtype may inadvertently interact with the other ER subtype in some patients if dosing is too high or drug interactions result in inappropriate accumulation.
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322
17
Although the complexities of ER and ER function in target tissues remain to be fully characterized, the knowledge gained to date has provided a solid foundation for further progress. Future discoveries could include the development of the perfect SERM. Additionally, a combination of ER subtype-specic ligands may be able to control cancer in specic target sites or improve current treatment strategies for a variety of debilitating diseases linked to estrogen withdrawal following menopause. Most importantly, the elucidation of the complex signal transduction pathways might open the door to novel therapeutic strategies not previously considered appropriate. Clearly, the close cooperation of laboratory science with clinical outcomes has enhanced our knowledge of the disease process in breast and endometrial cancer with established agents for treatment. The promise for the future is to target the ERs to prevent colon, prostate and ovarian cancer. Indeed, this concept would not have appeared to be reasonable a decade ago and has resulted from advances in molecular biology in the laboratory.
References
[1] Jordan VC. Antiestrogens and selective estrogen receptor modulators as multifunctional medicines. 1. Receptor interactions. J Med Chem 2003;46:883908. [2] Jordan VC. Antiestrogens and selective estrogen receptor modulators as multifunctional medicines. 2. Clinical considerations and new agents. J Med Chem 2003;46:1081111. [3] Beatson G. On the treatment of inoperable cases of the carcinoma of the mamma: suggestions for a new method of treatment, with illustrative cases. The Lancet 1896;2:1047. [4] Boyd S. On oophorectomy in cancer of the breast. Br Med J 1900;2:11617. [5] Jensen EV, Jacobson HI. Basic guides to the mechanism of estrogen action. Recent Prog Hormone Res 1962;18:387414. [6] Jensen EV, Block GE, Smith S, Kyser K, DeSombre ER. Estrogen receptors and breast cancer response to adrenalectomy. Natl Cancer Inst Monogr 1971;34:5570. [7] McGuire WL, Carbone PP, Vollmer EP. In: McGuire, WL, Carbone, PP, Sears, ME, Escher, GC, editors. Estrogen receptors in human breast cancer. New York: Raven Press; 1975. p. 17. [8] Knight WA, Livingston RB, Gregory EJ, McGuire WL. Estrogen receptor as an independent prognostic factor for early recurrence in breast cancer. Cancer Res 1977;37:466971. [9] Jordan VC, Gapstur S, Morrow M. Selective estrogen receptor modulation and reduction in risk of breast cancer, osteoporosis, and coronary heart disease. J Natl Cancer Inst 2001;93:144957. [10] Osborne CK, Pippen J, Jones SE, et al. Double-blind randomized trial comparing the efcacy and tolerability of fulvestrant versus anastrozole in postmenopausal women with advanced breast cancer progressing on prior endocrine therapy: results of a North American trial. J Clin Oncol 2002;20:338695. [11] Howell A, Robertson JF, Quaresma Albano J, et al. Fulvestrant, formerly ICI 182,780, is as effective as anastrozole in postmenopausal women with advanced breast cancer progressing after prior endocrine treatment. J Clin Oncol 2002;20:3396403. [12] Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: rst results of the ATAC randomised trial, Lancet 2002;359:21319.
[13] Bonneterre J, Thurlimann B, Robertson JF, et al. Anastrozole versus tamoxifen as rst-line therapy for advanced breast cancer in 668 postmenopausal women: results of the Tamoxifen or Arimidex Randomized Group Efcacy and Tolerability study. J Clin Oncol 2000;18:374857. [14] Nabholtz JM, Buzdar A, Pollak M, et al. Anastrozole is superior to tamoxifen as rst-line therapy for advanced breast cancer in postmenopausal women: results of a North American multicenter randomized trial. Arimidex Study Group. J Clin Oncol 2000;18:3758 67. [15] Mouridsen H, Gershanovich M, Sun Y, et al. Superior efcacy of letrozole versus tamoxifen as rst-line therapy for postmenopausal women with advanced breast cancer: results of a phase III study of the International Letrozole Breast Cancer Group. J Clin Oncol 2001;19:2596606. [16] Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benets of estrogen plus progestin in healthy postmenopausal women: principal results From the Womens Health Initiative randomized controlled trial. JAMA 2002;288:32133. [17] Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Lancet 1997;350:104759. [18] Walter P, Green S, Greene G, et al. Cloning of the human estrogen receptor cDNA. Proc Natl Acad Sci USA 1985;82:788993. [19] Greene GL, Gilna P, Watereld M, Baker A, Hort Y, Shine J. Sequence and expression of human estrogen receptor complementary DNA. Science 1986;231:11504. [20] Green S, Walter P, Kumar V, et al. Human oestrogen receptor cDNA: sequence. Nature 1986;320:1349. [21] Flouriot G, Brand H, Denger S. Identication of a new isoform of the human estrogen receptor-alpha (hER-alpha) that is encoded by distinct transcripts and that is able to repress hER-alpha activation function 1. EMBO J 2000;19:4688700. [22] Murphy LC, Dotzlaw H, Leygue E, Douglas D, Coutts A, Watson PH. Estrogen receptor variants and mutations. J Steroid Biochem Mol Biol 1997;62:36372. [23] Poola I, Koduri S, Chatra S, Clarke R. Identication of twenty alternatively spliced estrogen receptor alpha mRNAs in breast cancer cell lines and tumors using splice targeted primer approach. J Steroid Biochem Mol Biol 2000;72:24958. [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;93:592530. [25] Tremblay GB, Tremblay A, Copeland NG, et al. Cloning, chromosomal localization, and functional analysis of the murine estrogen receptor beta. Mol Endocrinol 1997;11:35365. [26] Mosselman S, Polman J, Dijkema R. ER beta: identication and characterization of a novel human estrogen receptor. FEBS Lett 1996;392:4953. [27] Enmark E, Pelto-Huikko M, Grandien K, et al. Human estrogen receptor beta-gene structure, chromosomal localization, and expression pattern. J Clin Endocrinol Metab 1997;82:425865. [28] Ogawa S, Inoue S, Watanabe T, et al. The complete primary structure of human estrogen receptor beta (hER beta) and its heterodimerization with ER alpha in vivo and in vitro. Biochem Biophys Res Commun 1998;243:1226. [29] Lewandowski S, Kalita K, Kaczmarek L. Estrogen receptor beta. Potential functional signicance of a variety of mRNA isoforms. FEBS Lett 2002;524:15. [30] Nilsson S, Makela S, Treuter E, et al. Mechanisms of estrogen action. Physiol Rev 2001;81:153565. [31] Hall JM, McDonnell DP. The estrogen receptor beta-isoform (ERbeta) of the human estrogen receptor modulates ERalpha transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology 1999;140:556678.
18
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322 [54] Suen CS, Berrodin TJ, Mastroeni R, Cheskis BJ, Lyttle CR, Frail DE. A transcriptional coactivator, steroid receptor coactivator-3, selectively augments steroid receptor transcriptional activity. J Biol Chem 1998;273:2764553. [55] Watanabe T, Inoue S, Ogawa S, et al. Agonistic effect of tamoxifen is dependent on cell type, ERE-promoter context, and estrogen receptor subtype: functional difference between estrogen receptors alpha and beta. Biochem Biophys Res Commun 1997;236:1405. [56] McInerney EM, Weis KE, Sun J, Mosselman S, Katzenellenbogen BS. Transcription activation by the human estrogen receptor subtype beta (ER beta) studied with ER beta and ER alpha receptor chimeras. Endocrinology 1998;139:451322. [57] Adlercreutz H. Phyto-oestrogens and cancer. Lancet Oncol 2002;3:36473. [58] Mishra SI, Dickerson V, Najm W. Phytoestrogens and breast cancer prevention: what is the evidence? Am J Obstet Gynecol 2002;188:S6670. [59] Adlercreutz H, Mazur W. Phyto-oestrogens and Western diseases. Ann Med 1997;29:95120. [60] Shiau AK, Barstad D, Radek JT, et al. Structural characterization of a subtype-selective ligand reveals a novel mode of estrogen receptor antagonism. Nat Struct Biol 2002;9:35964. [61] Sun J, Huang YR, Harrington WR, Sheng S, Katzenellenbogen JA, Katzenellenbogen BS. Antagonists selective for estrogen receptor alpha. Endocrinology 2002;143:9417. [62] Brzozowski AM, Pike AC, Dauter Z, et al. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 1997;389:7538. [63] Pike AC, Brzozowski AM, Hubbard RE. A structural biologists view of the oestrogen receptor. J Steroid Biochem Mol Biol 2000;74:2618. [64] Jordan VC, Collins MM, Rowsby L, Prestwich G. A monohydroxylated metabolite of tamoxifen with potent antioestrogenic activity. J Endocrinol 1977;75:30516. [65] Shiau AK, Barstad D, Loria PM, et al. The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 1998;95:92737. [66] Pike AC, Brzozowski AM, Hubbard RE, et al. Structure of the ligand-binding domain of oestrogen receptor beta in the presence of a partial agonist and a full antagonist. EMBO J 1999;18:460818. [67] Pike AC, Brzozowski AM, Walton J, et al. Structural insights into the mode of action of a pure antiestrogen. Structure (Camb) 2001;9:14553. [68] van Hoorn WP. Identication of a second binding site in the estrogen receptor. J Med Chem 2002;45:5849. [69] Zhang Z, Maier B, Santen RJ, Song RX. Membrane association of estrogen receptor alpha mediates estrogen effect on MAPK activation. Biochem Biophys Res Commun 2002;294:92633. [70] Levin ER. Cellular functions of plasma membrane estrogen receptors. Steroids 2002;67:4715. [71] Santen RJ, Song RX, McPherson R, et al. The role of mitogen-activated protein (MAP) kinase in breast cancer. J Steroid Biochem Mol Biol 2002;80:23956. [72] Razandi M, Pedram A, Greene GL, Levin ER. Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: studies of ERalpha and ERbeta expressed in Chinese hamster ovary cells. Mol Endocrinol 1999;13:30719. [73] Lannigan DA. Estrogen receptor phosphorylation. Steroids 2003;68:19. [74] Kato S. Estrogen receptor-mediated cross-talk with growth factor signaling pathways. Breast Cancer 2001;8:39. [75] Denton RR, Koszewski NJ, Notides AC. Estrogen receptor phosphorylation. Hormonal dependence and consequence on specic DNA binding. J Biol Chem 1992;267:72638. [76] Le Goff P, Montano MM, Schodin DJ, Katzenellenbogen BS. Phosphorylation of the human estrogen receptor. Identication of
[32] Kuiper GG, Carlsson B, Grandien K, et al. Comparison of the ligand binding specicity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 1997;138:86370. [33] Kuiper GG, Lemmen JG, Carlsson B, et al. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 1998;139:425263. [34] Gustafsson JA. Estrogen receptor betaa new dimension in estrogen mechanism of action. J Endocrinol 1999;163:37983. [35] Taylor AH, Al-Azzawi F. Immunolocalisation of oestrogen receptor beta in human tissues. J Mol Endocrinol 2000;24:14555. [36] Hiroi H, Inoue S, Watanabe T, et al. Differential immunolocalization of estrogen receptor alpha and beta in rat ovary and uterus. J Mol Endocrinol 1999;22:3744. [37] Sar M, Welsch F. Differential expression of estrogen receptor-beta and estrogen receptor-alpha in the rat ovary. Endocrinology 1999;140:96371. [38] Weihua Z, Warner M, Gustafsson JA. Estrogen receptor beta in the prostate. Mol Cell Endocrinol 2002;193:15. [39] Kumar V, Green S, Stack G, Berry M, Jin JR, Chambon P. Functional domains of the human estrogen receptor. Cell 1987;51:941 51. [40] Tora L, White J, Brou C, et al. The human estrogen receptor has two independent nonacidic transcriptional activation functions. Cell 1989;59:47787. [41] Berry M, Metzger D, Chambon P. Role of the two activating domains of the oestrogen receptor in the cell-type and promoter-context dependent agonistic activity of the anti- oestrogen 4-hydroxytamoxifen. EMBO J 1990;9:28118. [42] Danielian PS, White R, Lees JA, Parker MG. Identication of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J 1992;11:102533. [43] Tzukerman MT, Esty A, Santiso-Mere D, et al. Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol 1994;8:2130. [44] Kraus WL, McInerney EM, Katzenellenbogen BS. Ligand-dependent, transcriptionally productive association of the amino- and carboxyl-terminal regions of a steroid hormone nuclear receptor. Proc Natl Acad Sci USA 1995;92:123148. [45] Fawell SE, White R, Hoare S, Sydenham M, Page M, Parker MG. Inhibition of estrogen receptor-DNA binding by the pure antiestrogen ICI 164,384 appears to be mediated by impaired receptor dimerization. Proc Natl Acad Sci USA 1990;87:68837. [46] Dauvois S, Danielian PS, White R, Parker MG. Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover. Proc Natl Acad Sci USA 1992;89:403741. [47] McKenna NJ, Lanz RB, OMalley BW. Nuclear receptor coregulators: cellular and molecular biology. Endocr Rev 1999;20:32144. [48] Klinge CM. Estrogen receptor interaction with co-activators and co-repressors. Steroids 2000;65:22751. [49] Tremblay GB, Giguere V. Coregulators of estrogen receptor action. Crit Rev Eukaryot Gene Exp 2002;12:122. [50] Heery DM, Kalkhoven E, Hoare S, Parker MG. A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 1997;387:7336. [51] Montano MM, Ekena K, Delage-Mourroux R, Chang W, Martini P, Katzenellenbogen BS. An estrogen receptor-selective coregulator that potentiates the effectiveness of antiestrogens and represses the activity of estrogens. Proc Natl Acad Sci USA 1999;96:694752. [52] Katzenellenbogen BS, Montano MM, Ediger TR, et al. Estrogen receptors: selective ligands, partners, and distinctive pharmacology. Recent Prog Horm Res 2000;55:16393. [53] Endoh H, Maruyama K, Masuhiro Y, et al. Purication and identication of p68 RNA helicase acting as a transcriptional coactivator specic for the activation function 1 of human estrogen receptor alpha. Mol Cell Biol 1999;19:536372.
S.T. Pearce, V.C. Jordan / Critical Reviews in Oncology/Hematology 50 (2004) 322 hormone-regulated sites and examination of their inuence on transcriptional activity. J Biol Chem 1994;269:445866. Ali S, Metzger D, Bornert JM, Chambon P. Modulation of transcriptional activation by ligand-dependent phosphorylation of the human oestrogen receptor A/B region. EMBO J 1993;12:115360. Joel PB, Traish AM, Lannigan DA. Estradiol and phorbol ester cause phosphorylation of serine 118 in the human estrogen receptor. Mol Endocrinol 1995;9:104152. Bunone G, Briand PA, Miksicek RJ, Picard D. Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation. EMBO J 1996;15:217483. Kato S, Endoh H, Masuhiro Y, et al. Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 1995;270:14914. Joel PB, Traish AM, Lannigan DA. Estradiol-induced phosphorylation of serine 118 in the estrogen receptor is independent of p42/p44 mitogen-activated protein kinase. J Biol Chem 1998;273:1331723. Chen D, Washbrook E, Sarwar N, et al. Phosphorylation of human estrogen receptor alpha at serine 118 by two distinct signal transduction pathways revealed by phosphorylation-specic antisera. Oncogene 2002;21:492131. Arnold SF, Obourn JD, Jaffe H, Notides AC. Serine 167 is the major estradiol-induced phosphorylation site on the human estrogen receptor. Mol Endocrinol 1994;8:120814. Joel PB, Smith J, Sturgill TW, Fisher TL, Blenis J, Lannigan DA. pp90rsk1 regulates estrogen receptor-mediated transcription through phosphorylation of S...