Advertisement

The Multifaceted Mechanisms of Estradiol and Estrogen Receptor Signaling*

  • Julie M. Hall
    Correspondence
    To whom correspondence should be addressed. E-mail: [email protected] niehs.nih.gov
    Affiliations
    From the Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
    Search for articles by this author
  • John F. Couse
    Affiliations
    From the Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
    Search for articles by this author
  • Kenneth S. Korach
    Affiliations
    From the Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
    Search for articles by this author
  • Author Footnotes
    * This minireview will be reprinted in the 2001 Minireview Compendium, which will be available in December, 2001. This is the second article of five in the “Nuclear Receptor Minireview Series.”
Open AccessPublished:October 05, 2001DOI:https://doi.org/10.1074/jbc.R100029200
      ER
      estrogen receptor
      ERE
      estrogen response element
      AF
      activation function
      PR
      progesterone receptor
      EGF
      epidermal growth factor
      IGF
      insulin-like growth factor
      MAPK
      mitogen-activated protein kinase
      SERM
      selective ER modulator
      The steroid hormone 17β-estradiol (E2) is a key regulator of growth, differentiation, and function in a wide array of target tissues, including the male and female reproductive tracts, mammary gland, and skeletal and cardiovascular systems. The predominant biological effects of E2 are mediated through two distinct intracellular receptors, ERα1 and ERβ, each encoded by unique genes (
      • Giguere V.
      • Tremblay A.
      • Tremblay G.B.
      ) but possessing the hallmark modular structure of functional domains characteristic of the steroid/thyroid hormone superfamily of nuclear receptors (introduced in the Minireview Prologue (
      • Olefsky J.M.
      )). Certain functional domains of the ERα and ERβ exhibit a high degree of homology, namely the DNA- and ligand-binding domains, at 97 and 60%, respectively, whereas considerable divergence is apparent in the N terminus (18% homology). Hence, ERα and ERβ interact with identical DNA response elements and exhibit a similar binding affinity profile for an array of endogenous, synthetic, and naturally occurring estrogens when assayed in vitro (
      • Kuiper G.G.J.M.
      • Carlsson B.
      • Grandien K.
      • Enmark E.
      • Haggblad J.
      • Nilsson S.
      • Gustafsson J.A.
      ).In vitro studies also suggest the two receptors may play redundant roles in estrogen signaling; however, tissue localization studies have revealed distinct expression patterns for each receptor that suggest otherwise. Whereas ERα is the predominant subtype expressed in the breast, uterus, cervix, vagina, and several additional target organs, ERβ exhibits a more limited expression pattern and is primarily detected in the ovary, prostate, testis, spleen, lung, hypothalamus, and thymus (
      • Couse J.F.
      • Lindzey J.
      • Grandien K.
      • Gustafsson J.A.
      • Korach K.S.
      ). Regional expression differences of the two receptors have been identified in the brain (
      • Shughrue P.J.
      • Lane M.V.
      • Merchenthaler I.
      ). Further evidence of distinct biological functions for the ERs is revealed by the contrasting phenotypes observed in the individual lines of ER knockout mice, the αERKO and βERKO, which exhibit phenotypes that generally mirror the respective ER expression patterns (
      • Couse J.F.
      • Korach K.S.
      ). The most striking phenotypes in the female αERKO mice include estrogen insensitivity (leading to hypoplasia) in the reproductive tract, hypergonadotropic hypergonadism, lack of pubertal mammary gland development, and excess adipose tissue, whereas in the male, testicular degeneration and epididymal dysfunction are major factors (
      • Couse J.F.
      • Korach K.S.
      ). These phenotypes combined with severe deficits in sexual behavior result in complete infertility in both sexes of the αERKO. In contrast, βERKO males are fertile and to date show no obvious phenotypes; however, βERKO females exhibit inefficient ovarian function and subfertility. Interestingly, compound knockout mice (αβERKO) exhibit phenotypes that most heavily resemble those of the αERKO, with the exception of the ovarian phenotype, characterized by progressive germ cell loss accompanied by redifferentiation of the surrounding somatic cells, suggesting a requisite role for both ER forms in this tissue (
      • Couse J.F.
      • Curtis Hewitt S.
      • Bunch D.O.
      • Sar M.
      • Walker V.R.
      • Davis B.J.
      • Korach K.S.
      ,
      • Dupont S.
      • Krust A.
      • Gansmuller A.
      • Dierich A.
      • Chambon P.
      • Mark M.
      ).
      With the cloning of the first ER cDNA 15 years ago has come an immense appreciation of the complex molecular mechanisms underlying the diverse physiological actions of E2 and the multitude of synthetic ER ligands. This minireview will discuss the recent progress toward our understanding of the molecular mechanisms of estrogen signaling, focusing on the following four pathways: 1) classical ligand-dependent; 2) ligand-independent; 3) DNA binding-independent; and 4) cell-surface (nongenomic) signaling (Fig.1). Insights into these cellular/molecular mechanisms and the role of each ER subtype as revealed by ER-null animal models will be incorporated. A discussion of the mechanisms of action of several relevant selective ER modulator (SERM) ligands and their activities through the different ER signaling pathways will follow.
      Figure thumbnail gr1
      Figure 1The multifaceted mechanisms of estradiol and estrogen receptor signaling. The biological effects of estradiol (E2) are mediated through at least four ER pathways.1, classical ligand-dependent, E2-ER complexes bind to EREs in target promoters leading to an up- or down-regulation of gene transcription and subsequent tissue responses. 2, ligand-independent. Growth factors (GF) or cyclic adenosine monophosphate (not shown) activate intracellular kinase pathways, leading to phosphorylation (P) and activation of ER at ERE-containing promoters in a ligand-independent manner. 3,ERE-independent, E2-ER complexes alter transcription of genes containing alternative response elements such as AP-1 through association with other DNA-bound transcription factors (Fos/Jun), which tether the activated ER to DNA, resulting in an up-regulation of gene expression. 4, Cell-surface (nongenomic) signaling, E2 activates a putative membrane-associated binding site, possibly a form of ER linked to intracellular signal transduction pathways that generate rapid tissue responses. The roles of coactivators and corepressors in ER signaling (not shown above) are discussed in the first minireview of this series (
      • Rosenfeld M.G.
      • Glass C.K.
      ).

      Classical Mechanism of Estrogen Action: Ligand-dependent

      The ligand-dependent mechanism of ER action is the defining element of the Class I members of the nuclear steroid/thyroid receptor superfamily (
      • Mangelsdorf D.J.
      • Thummel C.
      • Beato M.
      • Herrlich P.
      • Schutz G.
      • Umesono K.
      • Blumberg B.
      • Kastner P.
      • Mark M.
      • Chambon P.
      • Evans R.M.
      ), of which the ERα and -β are members (Fig.1). This model states that in the absence of hormone, the receptor is sequestered in a multiprotein inhibitory complex within the nuclei of target cells. The binding of ligand induces an activating conformational change within the ER and promotes homodimerization and high affinity binding to specific DNA response elements (EREs), which are cis-acting enhancers located within the regulatory regions of target genes. The DNA-bound receptors contact the general transcription apparatus either directly or indirectly via cofactor proteins (
      • McKenna N.J.
      • Lanz R.B.
      • O'Malley B.W.
      ), of which several have been identified, including SRC-1, GRIP1, AIB1, CBP/p300, TRAP220, PGC-1, p68 RNA helicase, and SRA (discussed in the first Minireview of this series (
      • Rosenfeld M.G.
      • Glass C.K.
      )). It is generally accepted that the ER-coactivator interactions stabilize the formation of a transcription preinitiation complex and facilitate the necessary disruption of chromatin at the ERE. Depending on the cell and promoter context, the DNA-bound receptor exerts either a positive or negative effect on expression of the downstream target gene.
      The ligand-dependent transcriptional activity of ERα is mediated by two separate nonacidic activation domains, a constitutive activation function-1 (AF-1) located within the N terminus (A/B domain) and a hormone-dependent AF-2, located in the ligand-binding domain (
      • Tora L.
      • White J.
      • Brou C.
      • Tasset D.
      • Webster N.
      • Scheer E.
      • Chambon P.
      ). The AF-2 functional domain includes a highly conserved amphipathic α-helix (H12) that is essential for ligand-dependent transcriptional activity and interaction with members of the SRC family of coactivators (
      • McKenna N.J.
      • Lanz R.B.
      • O'Malley B.W.
      ,
      • Heery D.M.
      • Kalkhoven E.
      • Hoare S.
      • Parker M.G.
      ). AF-1 activity is also dependent on recruitment of coactivators, both similar and unique from those utilized by AF-2 (
      • McKenna N.J.
      • Lanz R.B.
      • O'Malley B.W.
      ,
      • Webb P.
      • Nguyen P.
      • Shinsako J.
      • Anderson C.
      • Feng W.
      • Nguyen M.P.
      • Chen D.
      • Huang S.M.
      • Subramanian S.
      • McInerney E.
      • Katzenellenbogen B.S.
      • Stallcup M.R.
      • Kushner P.J.
      ). The AFs function in a synergistic manner in most ligand-activated mechanisms but may also function independently in certain cell and promoter contexts (
      • Tzukerman M.T.
      • Esty A.
      • Santiso-Mere D.
      • Danielian P.
      • Parker M.G.
      • Stein R.B.
      • Pike J.W.
      • McDonnell D.P.
      ). Whereas the activation domains of ERα are well characterized, it is not yet clear how the homologous regions of ERβ contribute to the transcriptional activity of the receptor. Sequence homology andin vitro studies do indicate that ERβ also contains an AF-2 domain within the C terminus (
      • Cowley S.M.
      • Parker M.G.
      ,
      • Hall J.M.
      • McDonnell D.P.
      ,
      • McInerney E.M.
      • Weis K.E.
      • Sun J.
      • Mosselman S.
      • Katzenellenbogen B.S.
      ); however, reports that this region functions independently within ERβ (
      • Hall J.M.
      • McDonnell D.P.
      ) imply the AF-2 domains of the two ER subtypes play different roles. The high degree of sequence divergence in the N terminus, including the lack of a functional AF-1 in the human ERβ subtype, also suggests this region may function differently (
      • Cowley S.M.
      • Parker M.G.
      ,
      • Hall J.M.
      • McDonnell D.P.
      ). Furthermore, murine ERβ contains an AF-1 region very similar to that of ERα in both sequence and function (
      • Tremblay A.
      • Tremblay G.B.
      • Labrie F.
      • Giguere V.
      ), indicating that extrapolations of findings between species may be difficult.
      The ERKO mice have proven useful in confirming the ligand-dependent ERE-mediated mechanisms of ER signaling that were postulated from previous target gene promoter analyses andin vitro studies. Perhaps the best illustrative examples of classical E2-ER-ERE-mediated gene induction are those exhibited by the genes encoding the progesterone receptor (PR), another Class I member of the steroid receptor superfamily, and lactoferrin, a secretory protein. The proximal promoter of the PR gene of multiple species possesses a functional ERE-like sequence able to bind ER and confer estrogen responsiveness when complexed to various reporter constructs (
      • Kraus W.L.
      • Montano M.M.
      • Katzenellenbogen B.S.
      ). Similarly, the promoter of the mouse lactoferrin gene possesses a single palindromic ERE sequence located slightly less than 350 base pairs upstream of the transcription initiation site and has also been shown to confer estrogen responsiveness when complexed to various reporter constructs (
      • Liu Y.
      • Teng C.T.
      ). Prior to these studies, there were numerous reports of increased PR and lactoferrin expression in the uterus following E2 exposure (
      • Pentecost B.T.
      • Teng C.T.
      ,
      • Aronica S.M.
      • Katzenellenbogen B.S.
      ). The obligatory role of ERα in this mechanism is illustrated by the complete lack of E2 induction of the PR and lactoferrin genes in the uteri of αERKO females (
      • Couse J.F.
      • Korach K.S.
      ). Because estrogen-stimulated increases in PR are localized in the stroma and myometrium, whereas increases in lactoferrin are isolated to the luminal and glandular epithelium (
      • Tibbetts T.A.
      • Mendoza-Meneses M.
      • O'Malley B.W.
      • Conneely O.M.
      ), it can be stated that ERα gene disruption results in estrogen insensitivity in multiple compartments of the uterus. However, the maintenance of constitutive levels of PR and lactoferrin expression in the αERKO uteri suggests the existence of ERα-independent pathways and perhaps a role for ERβ (
      • Couse J.F.
      • Korach K.S.
      ).
      Given the high degree of homology in the DNA-binding domains of the ERs, it was not initially clear whether ERβ represented a level of redundancy in estrogen signaling or in fact the receptor was necessary to mediate unique biological activities. Both receptor subtypes are positive regulators of ERE-complexed reporter constructs in mammalian cell culture studies, although ERα usually exhibits higher activation (
      • Cowley S.M.
      • Parker M.G.
      ,
      • Hall J.M.
      • McDonnell D.P.
      ). The preservation of ERα-mediated E2 actions in the αERKO mouse, such as the induction of PR expression in the medial preoptic nucleus of the brain and certain male sexual behaviors, indicates the ability of ERβ to compensate for the loss of ERα in some pathways (Ref.
      • Couse J.F.
      • Korach K.S.
      , and references therein). However, it has been interesting to note that when coexpressed in vitro the two receptors preferentially form heterodimers (
      • Cowley S.M.
      • Hoarse S.
      • Mosselman S.
      • Parker M.G.
      ), indicating that there may also be direct convergence between the ERα/ERβ pathways. Furthermore, ERβ has been shown to interact in a constitutive,i.e. ligand-independent, manner with the EREs of target promoters and can attenuate the ligand-activated transcriptional activity of ERα (
      • Hall J.M.
      • McDonnell D.P.
      ). Thus, it has been proposed that in cells where both receptors are expressed, overall estrogen responsiveness may be determined by the ERα:ERβ ratio (
      • Hall J.M.
      • McDonnell D.P.
      ). The recent discovery of additional ERβ isoforms, ERβcx and ERβ2, that preferentially heterodimerize and inhibit ERα activity suggests a modulatory role of ERβ (
      • Ogawa S.
      • Inoue S.
      • Watanabe T.
      • Orimo A.
      • Hosoi T.
      • Ouchi Y.
      • Muramatsu M.
      ,
      • Hanstein B.
      • Liu H.
      • Yancisin M.C.
      • Brown M.
      ). A recent report by Gustafsson and co-workers (
      • Weihua Z.
      • Saji S.
      • Makinen S.
      • Cheng G.
      • Jensen E.V.
      • Warner M.
      • Gustafsson J.A.
      ) described an enhanced response to E2 in the uteri of βERKO mice, further suggesting a modulatory role of ERβ on ERα-mediated transcriptional activity.

      Ligand-independent Activation of ER: Cross-talk with Peptide Growth Factors

      In addition to hormone-mediated activation, it is now well accepted that ER function can be modulated by extracellular signals in the absence of E2 (Fig. 1). These findings focus primarily on the ability of polypeptide growth factors such as epidermal growth factor (EGF) and insulin-like growth factor-1 (IGF-1) as well as the intracellular effector analog 8-bromo-cyclic adenosine monophosphate to activate ER and increase the expression of ER target genes (
      • Smith C.L.
      ). Many of these findings have been corroborated with in vivostudies, such as the ability of EGF to mimic the effect of E2 on the female reproductive tract (
      • Curtis S.W.
      • Washburn T.
      • Sewall C.
      • DiAugustine R.
      • Lindzey J.
      • Couse J.F.
      • Korach K.S.
      ). Although the molecular mechanisms involved in ligand-independent activation of ER are somewhat characterized, the biological role of these processes is not yet clear. It is possible that hormone-independent pathways allow ER activation in the presence of low E2 levels, as found in males. Alternatively, this phenomenon may serve as a mechanism to amplify growth factor pathways and thereby enhance mitogenesis within ER-positive tissues.
      The mechanisms by which the ER and growth factor pathways converge are not entirely clear. However, studies do indicate that each pathway may be dependent on the other for full manifestation of the respective ligand-mediated response. In the mouse uterus, cotreatment with anti-EGF antibodies was able to attenuate the uterine response to E2; in turn, administration of the ER antagonist ICI 164,384 reduced the uterine response to EGF (
      • Ignar-Trowbridge D.M.
      • Nelson K.G.
      • Bidwell M.C.
      • Curtis S.W.
      • Washburn T.F.
      • McLachlan J.A.
      • Korach K.S.
      ). These studies together with similar observations made in the mammary gland (
      • Ankrapp D.P.
      • Bennett J.M.
      • Haslam S.Z.
      ) led to the proposed model that the mitogenic action of E2 in these tissues is at least partially mediated by EGF; conversely, the mitogenic effects of EGF require the presence of ERα. More definitive evidence comes from studies of the uteri of αERKO females, which despite expressing wild-type levels of functional EGF and EGF receptor and showing evidence of EGF signaling, remain unresponsive to the mitogenic actions of EGF (
      • Curtis S.W.
      • Washburn T.
      • Sewall C.
      • DiAugustine R.
      • Lindzey J.
      • Couse J.F.
      • Korach K.S.
      ).
      Specific receptor domains of the ER are critical to E2-independent activation. Specifically, the effects of elevated intracellular cAMP are mediated through AF-2, whereas growth factor activation of ER requires the N-terminal AF-1 domain of the receptor (
      • El-Tanani M.K.K.
      • Green C.D.
      ). The majority of evidence indicates that modification of the phosphorylation state of the ER by cellular kinases may serve as an important mechanism of ligand-independent activation. The serine 118 residue of the human ERα AF-1 is phosphorylated by the mitogen-activated protein kinase (MAPK) pathways following treatment with EGF or IGF, enabling the receptor to interact with the ERα-specific coactivator p68 RNA helicase and activate target gene transcription (
      • Kato S.S.
      ). Interestingly, the MAPK pathway also enhances the activity of the murine ERβ through stimulating the recruitment of SRC-1 to the N terminus (
      • Tremblay A.
      • Tremblay G.B.
      • Labrie F.
      • Giguere V.
      ). Recent evidence has emerged that coactivators may also serve as points of convergence between ER and growth factor signaling pathways, as it was shown that SRC-1 and AIB1 are phosphorylated by MAPK, an event thought to enhance their transcriptional activities (
      • Font de Mora J.
      • Brown M.
      ,
      • Rowan B.G.
      • Weigel N.L.
      • O'Malley B.W.
      ).

      ERE-independent Genomic Actions of ER

      The ER signaling mechanisms discussed thus far provide an explanation for the regulation of genes in which a functional ERE-like sequence can be documented within the promoter. However, concurrent studies reporting E2-ER induction of genes in which no ERE-like sequence was apparent led to the discovery that agonist-bound ER can indeed lead to gene regulation in the absence of direct DNA binding (Fig. 1). Notably, ERα activation of IGF-1 and collagenase expression is mediated through the interaction of receptor with Fos and Jun at AP-1 binding sites, whereas several genes containing GC-rich promoter sequences are activated via an ERα-Sp1 complex. The molecular mechanism of ER action at these alternative sites is becoming increasingly clear. Studies show that E2-ERα activation of AP-1-responsive elements requires both AF-1 and AF-2 domains of the receptor, which bind and enhance the activity of the p160 components (e.g. SRC-1, GRIP1) of the coactivator complex recruited to the site by Fos/Jun. Interestingly, human ERβ, which lacks a functional AF-1, is unable to activate transcription of AP-1-regulated genes when bound with ER agonists, indicating the possibility of distinct physiological actions of the two ERs via the regulation of unique subsets of genes (Ref.
      • Kushner P.J.
      • Agard D.A.
      • Greene G.L.
      • Scanlan T.S.
      • Shiau A.K.
      • Uht R.M.
      • Webb P.
      , and references therein). Although demonstration of interaction between ER and AP-1 pathways in vivo has been more difficult, the ERKO models should be valuable tools in which to investigate the contribution of this pathway to estrogen signaling.

      Nongenomic Effects of Estrogens

      The observed rapid biological effects of E2 in the bone, breast, vasculature, and nervous system suggest that estrogens may also elicit nongenomic effects, possibly through cell-surface ER forms that are linked to intracellular signal transduction proteins (Fig. 1). It is now clear that ER and membrane-coupled tyrosine kinase pathways are integrally linked, as E2 has recently been shown to activate the MAPK signaling pathway in a variety of cell types. Importantly, there is increasing evidence that some of the vascular protective effects of E2 through ERα are mediated by a nongenomic mechanism involving a biphasic activation of endothelial nitric oxide synthase by estrogen through the MAPK and phosphatidylinositol 3-kinase/Akt pathways (
      • Mendelsohn M.E.
      ,
      • Simoncini T.
      • Hafezi-Moghadam A.
      • Brazil D.P.
      • Ley K.
      • Chin W.W.
      • Liao J.K.
      ). In osteoblasts and osteoclasts, MAPK is also rapidly activated by E2, which may be involved in the proliferative and antiapoptotic effects of the hormone, perhaps providing one mechanism by which estrogen is bone protective (
      • Kousteni S.
      • Bellido T.
      • Plotkin L.I.
      • O'Brien C.A.
      • Bodenner D.L.
      • Han L.
      • Han K.
      • DiGregorio G.B.
      • Katzenellenbogen J.A.
      • Katzenellenbogen B.S.
      • Roberson P.K.
      • Weinstein R.S.
      • Jilka R.L.
      • Manolagas S.C.
      ).
      It is still controversial whether the putative membrane ER is similar to one or both of the intracellular forms. The extensive data gathered on the structures of the two known nuclear ER forms clearly indicate that neither is a transmembrane protein. However, Razandi et al. (
      • Razandi M.
      • Pedram A.
      • Levin E.R.
      ) reported the detection of membrane ERα and ERβ receptors in Chinese hamster ovary cells transfected with an expression vector of the respective receptor cDNA, indicating that the membrane and nuclear forms of each ER originate from the same transcript and exhibit similar affinities for E2. These studies further demonstrated that the membrane-bound ERs were G protein-linked and able to elicit a variety of signal transduction events, including the induction of cell proliferation. In contrast, the preservation of rapid E2 actions on kainate-induced neuronal currents in the hippocampus of the αERKO mouse or wild-type mice treated with ICI 182,780 suggests the existence of a functional membrane ER that is distinct from the intracellular nuclear forms (
      • Gu Q.
      • Korach K.S.
      • Moss R.L.
      ). This hypothesis is also supported by studies demonstrating MAPK activation by the transcriptionally inactive stereoisomer 17α-estradiol in the cerebral cortex of the αERKO, an activity that was not blocked by ICI 182,780 (
      • Singh M.
      • Setalo G.
      • Guan X.
      • Frail D.E.
      • Toran-Allerand C.D.
      ). Furthermore, Bentenet al. (
      • Benten W.P.M.
      • Stephan C.
      • Lieberherr M.
      • Wunderlich F.
      ) have recently reported transcription-independent estrogen signaling in mouse macrophages that is mediated through a novel G protein-coupled membrane ER. In light of these data, it will be important to define the precise nature of the estrogen binding protein(s) involved in these pathways so that pharmaceuticals can be developed which target both the nongenomic and nuclear effects of estrogen. The ERKO mutant mice provide an appropriate in vivo model to not only study the role of the nuclear receptors but also further the investigations of steroid hormone actions that may be nuclear receptor independent.

      Tissue-specific Activation of ER: Selective Estrogen Receptor Modulators

      Although it has become increasing clear that the activated ERα and ERβ are involved in diverse cellular pathways, additional complexity in ER signaling became apparent with the identification of synthetic compounds with mixed agonist/antagonist activities on the ER. A major challenge to the pharmaceutical industry continues to be the development of ER ligands that retain the beneficial effects of estrogen in the targeted tissue, e.g. bone, brain, and cardiovascular tissues, but lack the mitogenic and perhaps carcinogenic actions in the breast and uterus. The classical receptor theory that agonists function as molecular switches, converting ER from an inactive to an active form whereas antiestrogens competitively inhibit agonist binding and lock the receptor in a latent state, could not explain the mechanism of such a SERM. However, as put forth above, the classic model of ER action was oversimplified, as it does not account for the molecular pharmacology of several known antiestrogens. The first evidence that the activities of synthetic ER ligands were dependent on the target tissue came from clinical studies in which patients administered tamoxifen as adjuvant therapy for estrogen-dependent breast tumors exhibited estrogen-like positive effects in bone (
      • Love R.R.
      • Mazess R.B.
      • Barden H.S.
      • Epstein S.
      • Newcomb P.A.
      • Jordan V.C.
      • Carbone P.P.
      • DeMets D.L.
      ). Laboratory evidence that antiestrogens play an active role in modifying ER structure was first provided by protease digestion analysis of the human ERα, which demonstrated that antagonists induce unique conformational changes within the ER that were clearly distinct from that of the apo- and E2-occupied receptors (
      • McDonnell D.P.
      • Clemm D.L.
      • Hermann T.
      • Goldman M.E.
      • Pike J.W.
      ). Recent studies of the crystal structure of the ERα and ERβ ligand-binding domain when complexed with different ligands have confirmed that agonists and antagonists induce distinct alterations in ER structure (Ref.
      • Pike A.C.
      • Brzozowski A.M.
      • Walton J.
      • Hubbard R.E.
      • Thorsell A.
      • Li Y.
      • Gustafsson J.A.
      • Carlquist M.
      , and references therein). Collectively, these data indicate that antiestrogens differentially modulate receptor structure and thereby confer distinct changes in receptor function.
      At least three distinct classes of antiestrogens are currently recognized, based both on structural data and biological activity (
      • Jordan V.C.
      • Morrow M.
      ). Type I antiestrogens, represented by ICI 182,780, function as pure ER antagonists and oppose estrogen activity in all contexts both in vitro and in vivo. Type II and type III antiestrogens, represented by the benzothiophene-derived raloxifene and triphenylethylene tamoxifen, respectively, are recognized as SERMs (
      • McDonnell D.P.
      ). These agents differ from pure antiestrogens in that they mimic the biological effects of estrogen in selected tissues but oppose estrogen action in others (Table I). Many of these agents are currently in clinical use; raloxifene has recently received FDA approval for the prevention and treatment of osteoporosis (
      • McDonnell D.P.
      ). Furthermore, the notable antagonist effects of tamoxifen in the breast make it a first line therapy for the treatment of pre- and postmenopausal women with ER-positive advanced (Stage IV) breast cancer. Unfortunately, the estrogenic activity of tamoxifen provides for its undesirable uterotrophic effect and may also contribute to the phenomenon by which certain breast cancers alter their biology and recognize the compound as an agonist for growth (
      • Jordan V.C.
      • Morrow M.
      ). However, recent observations that tamoxifen-resistant breast xenographs are not cross-resistant to GW5638 (a tamoxifen-derived compound) indicate the still unexploited potential to develop tissue-targeted SERMs that maintain the beneficial effects of estrogen in some tissues while functioning as antagonists in the breast and uterus (
      • Connor C.E.
      • Norris J.D.
      • Broadwater G.
      • Willson T.M.
      • Gottardis M.M.
      • Dewhirst M.W.
      • McDonnell D.P.
      ).
      Table IBiological activities of ER ligands in selected target tissues
      Bone
      + denotes agonist activity; − denotes antagonist activity.
      BreastCardiovasculatureUterus
      17β-Estradiol++++
      ICI 182,780
      Tamoxifen+++
      Raloxifene++
      GW5638++
      1-a + denotes agonist activity; − denotes antagonist activity.
      The demonstration of SERMs prompted a reexamination of the pharmacology of ER ligands and more specifically the role of the AF domains in ER function through the classical ERE-mediated pathway. As discussed, pure agonists such as E2 are active in all cell and promoter contexts regardless of which AF function is dominant (
      • Tzukerman M.T.
      • Esty A.
      • Santiso-Mere D.
      • Danielian P.
      • Parker M.G.
      • Stein R.B.
      • Pike J.W.
      • McDonnell D.P.
      ). In contrast, the pure antiestrogen ICI 182,780 completely inhibits the activity of both AF-1 and AF-2 and thereby blocks the classical activation pathways of ERα and ERβ. By definition, the mixed agonist/antagonist activity of SERMs is dictated by the cell and promoter context, possibly reflecting the type and availability of certain cellular factors. This may be illustrated by tamoxifen, which inhibits AF-2 activity and consequently functions as an antagonist in all environments where AF-2 is required. However, in contexts where AF-1 of ERα is the dominant activator, tamoxifen displays partial agonist activity (
      • Tzukerman M.T.
      • Esty A.
      • Santiso-Mere D.
      • Danielian P.
      • Parker M.G.
      • Stein R.B.
      • Pike J.W.
      • McDonnell D.P.
      ). Therefore, the agonist activities of E2 and tamoxifen appear to be mediated through distinct mechanisms, likely reflecting differential AF-1 cofactor recruitment by the E2and tamoxifen-liganded ERα (
      • Norris J.D.
      • Paige L.A.
      • Christensen D.J.
      • Chang C.Y.
      • Huacani M.R.
      • Fan D.
      • Hamilton P.T.
      • Fowlkes D.M.
      • McDonnell D.P.
      ). Interestingly, coexpression of coactivator or corepressor with ERα alters tamoxifen agonist/antagonist activities (
      • Smith C.L.
      • Nawaz Z.
      • O'Malley B.W.
      ). Thus, one could envision that tamoxifen resistance in ER-positive tumors could occur as a result of alterations in the levels of cellular cofactors, a hypothesis that is currently under investigation by several laboratories.
      The discovery that raloxifene and GW5638 act as estrogen agonists in bone and cardiovascular tissue but do not appear to function as either AF-1 or AF-2 agonists on ERE-containing promoters (
      • McDonnell D.P.
      ) suggests that not all of the biological activities of SERMs are mediated through classical ERα pathways. The recognition that ERβ is expressed throughout the cardiovascular and skeletal systems therefore also provided a potential explanation for the tissue-selective agonist activities of SERMs. Although initial studies showed that many of the known antiestrogens function as antagonists of ERβ transcriptional activity on ERE-containing promoters (
      • Kuiper G.G.J.M.
      • Carlsson B.
      • Grandien K.
      • Enmark E.
      • Haggblad J.
      • Nilsson S.
      • Gustafsson J.A.
      ), in vitro evidence has emerged that SERMs function as potent ERβ agonists through nonclassical pathways at AP-1-responsive elements (
      • Paech K.
      • Webb P.
      • Kuiper G.G.J.M.
      • Nilsson S.
      • Gustafsson J.A.
      • Kushner P.J.
      • Scanlan T.S.
      ). One current hypothesis is that this activity involves titration of repressor proteins by the ERβ DNA-binding domain from the Fos/Jun complex (
      • Kushner P.J.
      • Agard D.A.
      • Greene G.L.
      • Scanlan T.S.
      • Shiau A.K.
      • Uht R.M.
      • Webb P.
      ). However, it remains to be determined whether the tissue-specific agonist activities of SERMs observed in vivoinvolve ERβ action through ERE-independent pathways.
      Recent evidence has emerged indicating the biological activities of SERMs may be influenced by communication between ERα and tyrosine kinase pathways. For example, the observation that the agonist activity of tamoxifen is enhanced when coadministered with cAMP or growth factors suggests that cross-talk between ER and external signals may augment the agonist activity of SERMs and possibly contribute to the onset of hormone resistance in breast tumors (
      • Smith C.L.
      ). Importantly, this latter hypothesis is supported by the finding that treatment of breast cancer cells with growth factors results in activation of ERα through the phosphatidylinositol 3-kinase/Akt pathway and a subsequent inhibition of tamoxifen-induced apoptosis (
      • Campbell R.A.
      • Bhat-Nakshatri P.
      • Patel N.M.
      • Constantinidou D.
      • Ali S.
      • Nakshatri H.
      ). Evidence that tamoxifen, like estrogen, elicits rapid effects through activation of membrane-signaling pathways (
      • Aronica S.M.
      • Kraus W.L.
      • Katzenellenbogen B.S.
      ) may suggest that some of the tissue-specific agonist activities of SERMs are also mediated through nongenomic pathways. Thus, in view of the discovery that E2has rapid antiapoptotic effects in breast cancer cells, it may be useful to develop membrane ER-specific SERMs that are targeted to cancer cells, enabling women on hormone replacement or antiestrogen therapies to retain beneficial effects in bone and the cardiovascular system.

      Summary

      In recent years it has become apparent through the use of ER agonists and antagonists that the biological actions of estrogens are multifaceted. Estrogens and antiestrogens mediate their effects through diverse molecular mechanisms. Thus, SERMs can be viewed as a class of compounds that exhibit an expanse of estrogen actions and specificity that capitalize on the preexistence of multiple cellular pathways of ER action in different target tissues. It is anticipated that the continued use of in vitro and animal models will reveal additional mechanisms of ER signaling and tissue response and provide a more clear understanding of the spectrum of estrogen action, which will facilitate the development of novel pharmaceuticals for the treatment of estrogen-associated pathologies.

      REFERENCES

        • Giguere V.
        • Tremblay A.
        • Tremblay G.B.
        Steroids. 1998; 63: 335-339
        • Kuiper G.G.J.M.
        • Carlsson B.
        • Grandien K.
        • Enmark E.
        • Haggblad J.
        • Nilsson S.
        • Gustafsson J.A.
        Endocrinology. 1997; 138: 863-870
        • Couse J.F.
        • Lindzey J.
        • Grandien K.
        • Gustafsson J.A.
        • Korach K.S.
        Endocrinology. 1997; 138: 4613-4621
        • Shughrue P.J.
        • Lane M.V.
        • Merchenthaler I.
        J. Comp. Neurol. 1997; 388: 507-525
        • Couse J.F.
        • Korach K.S.
        Endocr. Rev. 1999; 20: 358-417
        • Couse J.F.
        • Curtis Hewitt S.
        • Bunch D.O.
        • Sar M.
        • Walker V.R.
        • Davis B.J.
        • Korach K.S.
        Science. 1999; 286: 2328-2331
        • Dupont S.
        • Krust A.
        • Gansmuller A.
        • Dierich A.
        • Chambon P.
        • Mark M.
        Development. 2000; 127: 4277-4291
        • Mangelsdorf D.J.
        • Thummel C.
        • Beato M.
        • Herrlich P.
        • Schutz G.
        • Umesono K.
        • Blumberg B.
        • Kastner P.
        • Mark M.
        • Chambon P.
        • Evans R.M.
        Cell. 1995; 83: 835-883
        • McKenna N.J.
        • Lanz R.B.
        • O'Malley B.W.
        Endocr. Rev. 1999; 20: 321-344
        • Tora L.
        • White J.
        • Brou C.
        • Tasset D.
        • Webster N.
        • Scheer E.
        • Chambon P.
        Cell. 1989; 59: 477-487
        • Heery D.M.
        • Kalkhoven E.
        • Hoare S.
        • Parker M.G.
        Nature. 1997; 387: 733-736
        • Webb P.
        • Nguyen P.
        • Shinsako J.
        • Anderson C.
        • Feng W.
        • Nguyen M.P.
        • Chen D.
        • Huang S.M.
        • Subramanian S.
        • McInerney E.
        • Katzenellenbogen B.S.
        • Stallcup M.R.
        • Kushner P.J.
        Mol. Endocrinol. 1998; 12: 1605-1618
        • Tzukerman M.T.
        • Esty A.
        • Santiso-Mere D.
        • Danielian P.
        • Parker M.G.
        • Stein R.B.
        • Pike J.W.
        • McDonnell D.P.
        Mol. Endocrinol. 1994; 8: 21-30
        • Cowley S.M.
        • Parker M.G.
        J. Steroid Biochem. Mol. Biol. 1999; 69: 165-175
        • Hall J.M.
        • McDonnell D.P.
        Endocrinology. 1999; 140: 5566-5578
        • McInerney E.M.
        • Weis K.E.
        • Sun J.
        • Mosselman S.
        • Katzenellenbogen B.S.
        Endocrinology. 1998; 139: 4513-4522
        • Tremblay A.
        • Tremblay G.B.
        • Labrie F.
        • Giguere V.
        Mol. Cell. 1999; 3: 513-519
        • Kraus W.L.
        • Montano M.M.
        • Katzenellenbogen B.S.
        Mol. Endocrinol. 1994; 8: 952-969
        • Liu Y.
        • Teng C.T.
        Mol. Endocrinol. 1992; 6: 355-364
        • Pentecost B.T.
        • Teng C.T.
        J. Biol. Chem. 1987; 262: 10134-10139
        • Aronica S.M.
        • Katzenellenbogen B.S.
        Endocrinology. 1991; 128: 2045-2052
        • Tibbetts T.A.
        • Mendoza-Meneses M.
        • O'Malley B.W.
        • Conneely O.M.
        Biol. Reprod. 1998; 59: 1143-1152
        • Cowley S.M.
        • Hoarse S.
        • Mosselman S.
        • Parker M.G.
        J. Biol. Chem. 1997; 272: 19858-19862
        • Ogawa S.
        • Inoue S.
        • Watanabe T.
        • Orimo A.
        • Hosoi T.
        • Ouchi Y.
        • Muramatsu M.
        Nucleic Acids Res. 1998; 26: 3505-3512
        • Hanstein B.
        • Liu H.
        • Yancisin M.C.
        • Brown M.
        Mol. Endocrinol. 1999; 13: 129-137
        • Weihua Z.
        • Saji S.
        • Makinen S.
        • Cheng G.
        • Jensen E.V.
        • Warner M.
        • Gustafsson J.A.
        Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5936-5941
        • Smith C.L.
        Biol. Reprod. 1998; 58: 627-632
        • Curtis S.W.
        • Washburn T.
        • Sewall C.
        • DiAugustine R.
        • Lindzey J.
        • Couse J.F.
        • Korach K.S.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 12626-12630
        • Ignar-Trowbridge D.M.
        • Nelson K.G.
        • Bidwell M.C.
        • Curtis S.W.
        • Washburn T.F.
        • McLachlan J.A.
        • Korach K.S.
        Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 4658-4662
        • Ankrapp D.P.
        • Bennett J.M.
        • Haslam S.Z.
        J. Cell. Physiol. 1998; 174: 251-260
        • El-Tanani M.K.K.
        • Green C.D.
        Mol. Endocrinol. 1997; 11: 928-937
        • Kato S.S.
        Breast Cancer. 2001; 8: 3-9
        • Font de Mora J.
        • Brown M.
        Mol. Cell. Biol. 2000; 20: 5041-5047
        • Rowan B.G.
        • Weigel N.L.
        • O'Malley B.W.
        J. Biol. Chem. 2000; 275: 4475-4483
        • Kushner P.J.
        • Agard D.A.
        • Greene G.L.
        • Scanlan T.S.
        • Shiau A.K.
        • Uht R.M.
        • Webb P.
        J. Steroid Biochem. Mol. Biol. 2000; 74: 311-317
        • Mendelsohn M.E.
        Circ. Res. 2000; 87: 677-682
        • Simoncini T.
        • Hafezi-Moghadam A.
        • Brazil D.P.
        • Ley K.
        • Chin W.W.
        • Liao J.K.
        Nature. 2000; 407: 538-541
        • Kousteni S.
        • Bellido T.
        • Plotkin L.I.
        • O'Brien C.A.
        • Bodenner D.L.
        • Han L.
        • Han K.
        • DiGregorio G.B.
        • Katzenellenbogen J.A.
        • Katzenellenbogen B.S.
        • Roberson P.K.
        • Weinstein R.S.
        • Jilka R.L.
        • Manolagas S.C.
        Cell. 2001; 104: 719-730
        • Razandi M.
        • Pedram A.
        • Levin E.R.
        Mol. Endocrinol. 2000; 14: 1434-1447
        • Gu Q.
        • Korach K.S.
        • Moss R.L.
        Endocrinology. 1999; 140: 660-666
        • Singh M.
        • Setalo G.
        • Guan X.
        • Frail D.E.
        • Toran-Allerand C.D.
        J. Neurosci. 2000; 20: 1694-1700
        • Benten W.P.M.
        • Stephan C.
        • Lieberherr M.
        • Wunderlich F.
        Endocrinology. 2001; 142: 1669-1677
        • Love R.R.
        • Mazess R.B.
        • Barden H.S.
        • Epstein S.
        • Newcomb P.A.
        • Jordan V.C.
        • Carbone P.P.
        • DeMets D.L.
        N. Engl. J. Med. 1992; 326: 852-856
        • McDonnell D.P.
        • Clemm D.L.
        • Hermann T.
        • Goldman M.E.
        • Pike J.W.
        Mol. Endocrinol. 1995; 9: 659-668
        • Pike A.C.
        • Brzozowski A.M.
        • Walton J.
        • Hubbard R.E.
        • Thorsell A.
        • Li Y.
        • Gustafsson J.A.
        • Carlquist M.
        Structure. 2001; 9: 145-153
        • Jordan V.C.
        • Morrow M.
        Endocr. Rev. 1999; 20: 253-278
        • McDonnell D.P.
        Trends Endocrinol. Metab. 1999; 10: 301-311
        • Connor C.E.
        • Norris J.D.
        • Broadwater G.
        • Willson T.M.
        • Gottardis M.M.
        • Dewhirst M.W.
        • McDonnell D.P.
        Cancer Res. 2001; 61: 2917-2922
        • Norris J.D.
        • Paige L.A.
        • Christensen D.J.
        • Chang C.Y.
        • Huacani M.R.
        • Fan D.
        • Hamilton P.T.
        • Fowlkes D.M.
        • McDonnell D.P.
        Science. 1999; 285: 744-746
        • Smith C.L.
        • Nawaz Z.
        • O'Malley B.W.
        Mol. Endocrinol. 1997; 11: 657-666
        • Paech K.
        • Webb P.
        • Kuiper G.G.J.M.
        • Nilsson S.
        • Gustafsson J.A.
        • Kushner P.J.
        • Scanlan T.S.
        Science. 1997; 277: 1508-1510
        • Campbell R.A.
        • Bhat-Nakshatri P.
        • Patel N.M.
        • Constantinidou D.
        • Ali S.
        • Nakshatri H.
        J. Biol. Chem. 2001; 276: 9817-9824
        • Aronica S.M.
        • Kraus W.L.
        • Katzenellenbogen B.S.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 8517-8521
        • Olefsky J.M.
        J. Biol. Chem. 2001; 276: 36863-36864
        • Rosenfeld M.G.
        • Glass C.K.
        J. Biol. Chem. 2001; 276: 36865-36868