Selective Modulation of Estrogen Receptor Action

Yoel Sadovsky and Stuart Adler

Department of Obstetrics and Gynecology Washington University School of Medicine St. Louis, MO 63110

Address correspondence and requests for reprints to: Yoel Sadovsky, M.D., Department of Obstetrics and Gynecology, Washington University School of Medicine, Box 8064, St. Louis, Missouri 63110.


    Introduction
 Top
 Introduction
 References
 
In this volume of JCEM (see page 6), Baker, et al. (1) present data on a clinical trial of raloxifene in healthy, ovulatory women. Raloxifene is a new ligand that was developed by Eli Lilly and Company. Although not yet approved by the FDA, it has shown promise as a selective estrogen for postmenopausal preservation of bone mass, without uterotropic effects. Raloxifene is likely to join tamoxifen and clomiphene as a therapeutic compound acting as a ligand for estrogen receptors (ER), but with a distinctly different spectrum of activities from the natural hormone, 17-ß-estradiol.

The clinical usefulness of anti-estrogens in the treatment of estrogen-dependent tumors has posed a long-standing enigma. Tamoxifen and other estrogen-antagonists block the action of estrogens in several tissues, most notably the breast. Paradoxically, the same compounds exhibit an estrogen-like effect in other tissues, particularly the uterus. Clearly, these opposite effects of a single drug may hamper its clinical usefulness, as treatment of breast cancer may predispose the patient to endometrial hyperplasia, as well as the development of uterine tumors. Recent exciting discoveries related to the structure and function of ER, along with the development of new classes of selective estrogen receptor modulators (SERM), have shed new light on this enigma. The goal has been to find the ideal SERM, one that lowers the risk of breast and endometrial carcinoma via an antiestrogenic effect, yet exhibits an estrogen-like effect on hepatic and cardiovascular functions and maintains bone density as well.

Several classes of anti-estrogens were originally designed to inhibit ER activity. However, each of these compounds displays a different effect on estrogen responsive tissues. Two representatives of these drugs, tamoxifen and clomiphene, are commonly used in clinical practice and illustrate the heterogeneous effects as SERMs on target tissues (2). Tamoxifen exhibits an anti-estrogen activity on mammary tissues in postmenopausal women, but enhances endometrial cell proliferation and has an estrogen-like effect on the vaginal mucosa. In addition, tamoxifen increases the incidence of hot flashes and preserves bone density. In menstruating women, tamoxifen exhibits an anti-estrogen effect on the neuroendocrine axis and leads to elevated estradiol levels. Like tamoxifen, clomiphene acts centrally as an anti-estrogen, leading to ovulation induction. Furthermore, clomiphene diminishes endometrial proliferation and, like tamoxifen, it exhibits anti-estrogenic influence on the breast, but at a lower potency (2). Tamoxifen and clomiphene achieve levels of active compound that can produce distinct effects in both postmenopausal, hypoestrogenic populations and in fully estrogenized, menstruating women. Utilizing their properties as antagonists of estrogen action, both compounds have a role in clinical practice. Nevertheless, their agonist activities and potential actions on nonestrogenic pathways may also, for better or worse, affect women who are treated with these compounds.

Raloxifene, a nonsteroidal, benzothiophene derivative, appears to be a promising SERM; it acts as an estrogen-antagonist in the breast and uterus, but as an estrogen-agonist on the bone and liver. The beneficial effects of raloxifene as an estrogen replacement therapy in model systems of estrogen-deficient animal, as well as in postmenopausal women, have already been demonstrated (3, 4). Given its selective influence on estrogen responsive tissues, it is conceivable that raloxifene may benefit women of reproductive age who require an antiestrogen for the treatment of endometriosis or for suppression of benign uterine tumors such as leiomyomata. With this in mind, Baker, et al. (1) in this edition of JCEM sought to characterize the reproductive effects of raloxifene in healthy menstruating women. Their study was conducted in two parts. In the first part, raloxifene was administered to women for 5 days during the follicular, periovulatory, or luteal phases of the cycle. In the second part, two different doses of raloxifene were administered throughout the cycle. Raloxifene neither affected ovulation nor altered the length of the menstrual cycle. Raloxifene appeared to increase the production of FSH, whether administered for 5 days during the follicular phase or throughout the cycle. In contrast, raloxifene administration during the periovulatory period caused a small decrease in LH levels. Interestingly, while the general pattern of estradiol and progesterone levels during the menstrual cycle was unchanged, raloxifene treatment caused an overall increase in estradiol production during the entire menstrual cycle. The mechanism of this increase was not dissected in the present study, but may reflect an anti-estrogen effect of raloxifene at the hypothalamic/pituitary level or a direct effect on the ovary. Importantly, examination of the uterus revealed that, despite this alteration in estradiol production, the number of mitoses in endometrial glands was lower. Consistent with this finding, sonographic evaluation of endometrial thickness revealed a somewhat thinner endometrium in response to raloxifene. The authors acknowledge that the small numbers of patients in each of the groups might have precluded the identification of additional effects by raloxifene. It is also clear that the key premise set forth by the authors was not fulfilled, and it is unlikely that raloxifene may become useful for the treatment of endometriosis or leiomyomata. Nevertheless, this study provides further support to the selective action of raloxifene, previously observed in postmenopausal women. It also demonstrates that raloxifene administration, at the doses tested, neither inhibits nor stimulates ovulation, a fact that may be reassuring for women who may be treated with raloxifene for future indications.

Studies in recent years have expanded our classical understanding of ER action. A ligand-bound receptor dissociates from a heat-shock protein complex, the receptor homo-dimer binds to its cognate palindromic response element in target gene promoters, and its subsequent interactions with the cellular transcriptional machinery result in gene activation. The mechanism of action of pure anti-estrogens like ICI 164,384 appears to reflect an increase in receptor turnover or the occupation of the receptor’s ligand binding domain by a nonfunctional ligand, rendering the receptor incapable of dimerization (5, 6). However, this is clearly not the case for the SERMs. For SERM compounds like tamoxifen, clomiphene, and raloxifene, it has been difficult to explain how the actions of ER occupied by these nonsteroidal ligands can vary and be selective for distinct actions. It is now clear that ER, like other members of the steroid receptor superfamily, binds auxiliary proteins that either abrogate its activity (termed corepressors) or potentiate it (termed coactivators) (7). The transcriptional integrator, CBP/p300, participates as a part of the transcription complex in regulation of many genes and with many classes of transcription factors (8). CBP/p300 has been associated with histone acetylase and thus provides a link between transcription factors and chromatin structural changes ultimately associated with active gene transcription (9). For ER, the interaction with CBP/p300 has been shown in vitro to be dependent on the presence of both a suitable agonist ligand, such as 17-ß-estradiol, and a coregulator. The ligand-dependent interaction of ER with these proteins thus may provide a more detailed view of how ER causes changes in gene transcription and what determines the tissue selective effects of SERMs.

The resolution of the crystal structure of ER’s ligand binding domain, bound to either estradiol or raloxifene, may provide the molecular basis for interaction of ligand-bound ER with selective coregulators (10). While the ligand binding pocket of ER absolutely requires that its ligand possesses an aromatic ring, the relatively large size of this pocket can accommodate a variety of hydrophobic side groups. When bound by estradiol, discrete helices within the ligand binding domain are packed in a manner that directs the charged activation function domain of ER away from the body of the protein, allowing it to interact with coactivators. In contrast, the hydrophobic side-chain of raloxifene prevents this alignment of the activation domain, hindering its correct alignment with other regions within the ligand-binding domain of ER, thereby impeding interaction with particular coactivators. These data are important because they support the notion that individual ligands for ER do not all exhibit the same effects.

In addition to their effect on coactivators, SERMs can direct the binding of ER to distinct promoter sequences that differ from the consensus palindromic estrogen response element. This has been clearly demonstrated for raloxifene, which is capable of activating transcription via an adenine/guanine-rich sequence within the gene of transforming growth factor-ß3, an important modulator of bone remodeling (11, 12). Raloxifene is more potent than estradiol in stimulating this promoter, and this effect of raloxifene does not require the DNA-binding domain of ER. Thus, within a single cell type, the presence of either the consensus estrogen response element or the newly defined raloxifene response element dictates either a raloxifene-induced repression (by the former element) or activation (by the latter element) of gene expression (11). The generality of these novel types of response elements remains to be established.

Additional elements in the promoters of estradiol-responsive genes can play a key role in determining the effect of SERMs on gene activation. The role of the promoter element, AP-1, a binding site for proteins from the Jun/Fos family was recently defined (13). In the presence of an estrogen response element, tamoxifen acts as an estrogen antagonist. In contrast, when the estrogen response element is replaced by an AP-1 site, tamoxifen is converted to an estrogen agonist in cell lines of diverse tissue origin, excluding breast cells. This mechanism involves direct interaction of ER with AP-1 proteins and utilizes the AP-1 DNA response element. As also demonstrated for the raloxifene response element, in some cell lines activation from these sites does not require the DNA binding domain of ER. A new dimension to the interaction of ER with AP-1 proteins was unveiled with the discovery of ERß (14). Paech, et al. (15) reported that the interaction of AP-1 with ER{alpha}, as described above, was diametrically opposite to its interaction with ERß: while estradiol stimulated activity from an ER{alpha} complexed with AP-1, estradiol diminished the transcriptional activity of AP-1 associated ERß. Importantly, raloxifene was a partial agonist on an AP-1 coupled ER{alpha}, yet a complete agonist on an AP-1 coupled ERß. Together, these findings indicate that the cellular response to SERMs is determined by the expression of ER subtypes, by a repertoire of accessory DNA-bound proteins (such as AP-1), and by coactivators and corepressors, all interacting in a promoter-specific manner. The relative contribution of each of these component to the cellular response to SERM administration in vivo remains to be determined.

As detailed by Baker, et al. (1), the clinical effectiveness of raloxifene may be limited by competition for ER binding by the physiological levels of the high affinity, natural ligand, 17-ß-estradiol. The ability of low affinity compounds, or compounds present at low concentrations, to alter estrogenic pathways is not only a question for drug design, but is also part of a larger controversy regarding deleterious or beneficial actions of other nonsteroidal compounds including phytoestrogens (e.g. the soy compound, genistein) and environmental/industrial/agricultural compounds (e.g. insecticides, plasticizers, and detergents). Furthermore, for these phytoestrogens and environmental nonsteroidal compounds, our concerns must include not only the high estradiol milieu of the menstruating woman or the developing fetus in pregnancy, but also their actions in low estradiol states including menopause and childhood, and in adult males. Any evaluation of the effects of these compounds must also consider the possibility that they may act as SERMs or have additional effects independent of ER.

Our expanding body of knowledge of the molecular details of estrogenic regulation provides the potential for design or development of other compounds that might possess similar regulatory profiles to raloxifene, but that have a higher affinity for ER or increased bioavailability, thereby enhancing more effective competition with estradiol. This might enable these compounds to exert their actions not only in the postmenopausal age group, but also in the presence of estradiol in premenopausal women. The current study has shown that raloxifene appears to be safe in menstruating, ovulating women. Few and minimal effects specific to raloxifene were observed. A goal for the future remains the development of new, raloxifene-like compounds that can effectively exert their effects even in the presence of normal estradiol levels and thus expand the therapeutic applications of selective estrogen receptor modulators.

Received November 11, 1997.

Accepted November 12, 1997.


    References
 Top
 Introduction
 References
 

  1. Baker VL, Draper M, Paul S, et al. 1998 Reproductive endocrine and endometrial effects of raloxifene hydrochloride, a selective estrogen receptor modulator, in women with regular menstrual cycles. J Clin Endocrinol Metab. 83:6–13.[Abstract/Free Full Text]
  2. Baker VL, Jaffe RB. 1995 Clinical uses of antiestrogens. Obstet Gynecol Survey. 51:45–59.
  3. Black LJ, Sato M, Rowley ER, et al. 1994 Raloxifene (LY139481 HCl) prevents bone loss and reduces serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J Clin Invest. 93:63–69.[Medline]
  4. Draper MW, Flowers DE, Huster WJ, Neild JA, Harper KD, Arnaud C. 1996 A controlled trial of raloxifene (LY139481) HCl: impact on bone turnover and serum lipid profile in healthy postmenopausal women. J Bone Miner Res. 11:835–842.[Medline]
  5. Fawell SE, White R, Hoare S, Sydenham M, Page M, Parker MG. 1990 Inhibition of estrogen receptor-DNA binding by the "pure" antiestrogenic ICI 164,384 appears to be mediated by impaired receptor dimerization. Proc Natl Acad Sci USA. 87:6883–6887.[Abstract]
  6. Dauvois S, Danielian PS, White R, Parker MG. 1992 Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover. Proc Natl Acad Sci USA. 89:4037–4041.[Abstract]
  7. Horwitz KB, Jackson TA, Bain DL, Richer JK, Takimoto GS, Tung L. 1996 Nuclear receptor coactivators and corepressors. Mol Endocrinol. 10:1167–1177.[Abstract]
  8. Kamei Y, Xu L, Heinzel T, et al. 1996 A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell. 85:403–414.[Medline]
  9. Ogryzko VV, Schlitz RL, Russanova V, Howard BH, Nakatani Y. 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell. 87:953–959.[Medline]
  10. Brzozowski AM, Pike ACW, Dauter Z, et al. 1997 Molecular basis of agonism and antagonism in the oestrogen receptor. Nature. 389:753–758.[CrossRef][Medline]
  11. Yang NN, Venugopalan M, Hardikar S, Glasebrook A. 1996 Identification of an estrogen response element activated by metabolites of 17-ß-estradiol and raloxifene. Science. 273:1222–1225.[Abstract]
  12. Yang NN, Venugopalan M, Hardikar S, Glasebrook A. 1997 Correction: Raloxifene response needs more than an element. Science. 275:1249.[Medline]
  13. Webb P, Lopez GN, Uht RM, Kushner PJ. 1995 Tamoxifen activation of the estrogen receptor/AP-1 pathway: potential origin for the cell-specific estrogen-like effects of antiestrogens. Mol Endocrinol. 9:443–456.[Abstract]
  14. Kuiper GGJM, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson J-Å. 1996 Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA. 93:5925–5930.[Abstract/Free Full Text]
  15. Paech K, Webb P, Kuiper GGJM, et al. 1997 Differential ligand activation of estrogen receptors ER{alpha} and ERß at AP1 sites. Science. 277:1508–1510.[Abstract/Free Full Text]