Receptor aComplex · [CANCER RESEARCH 61, 3632–3639, May 1, 2001] Silencing and Reactivation of the Selective Estrogen Receptor Modulator-Estrogen Receptor aComplex1 Hong Liu, Eun-Sook

[CANCER RESEARCH 61, 3632–3639, May 1, 2001]

Silencing and Reactivation of the Selective Estrogen Receptor Modulator-EstrogenReceptor a Complex1

Hong Liu, Eun-Sook Lee, Alexander De Los Reyes, James W. Zapf, and V. Craig Jordan2

Robert H. Lurie Comprehensive Cancer Center, Northwestern University Medical School, Chicago, Illinois 60611 [H. L., E-S. L., A. D. L. R., V. C. J.]; Center for Breast Cancer,National Cancer Center, Koyang City Kyunggi-do, 4-11-351, Korea [E-S. L.]; and Signal Pharmaceuticals, San Diego, California 92121 [J. W. Z.]

ABSTRACT

4-Hydroxytamoxifen (4-OHT), a selective estrogen receptor modulator,is an agonist at a transforming growth factor-a (TGF-a) target gene insitu in MDA-MB-231 human breast cancer cells stably transfected withwild-type human ERa. In contrast, raloxifene (Ral) is a complete anties-trogen silencing activation function (AF) 1 and AF2 in this system. Anatural mutation D351YERa enhances 4-OHT agonist activity andchanges Ral-like compounds from antagonists to partial agonists. Wereasoned that: either the conformation of the Ral-D351YERa is altered,thereby reactivating AF2 in the ligand binding domain, or the change atamino acid 351 allosterically reactivates AF1 in the Ral-D351YERa com-plex. Unlike the estradiol-ERa complex, agonist activity of 4-OHT andraloxifene through ERa and D351YERa were not attributed to coactiva-tor (such as SRC-1, AIB1) binding to the ligand binding domain. Weconclude that the classic AF2 is not responsible for the agonist activities of4-OHT-ERa, 4-OHT-D351YERa, and Ral-D351YERa. To address therole of AF1, stable transfectants of ERa or D351YERa with an AF1deletion (D351DAF1, D351YDAF1) were generated in MDA-MB-231 cells.Additionally, D538A/E542A/D545A triple mutations within helix 12(D351–3m, D351Y3m) or the COOH-terminal 537 deletion (D351D537)were tested. The agonist activities of 4-OHT and raloxifene were lost inthese stable transfectants, but antiestrogenic action was retained. Thereactivation of an estrogen-like property of the Ral-ERa complex throughAF1 with the D351Y mutation illustrates a novel allosteric mechanism forthe selective estrogen receptor modulator ERa complex.

INTRODUCTION

ERa3 is a member of the nuclear hormone receptor superfamily ofligand-dependent transcriptional factors (1) and an important targetfor the treatment and prevention of breast cancer. Like all members inthis superfamily, ERa has A to F domains from the NH2 terminus tothe COOH terminus (Fig. 1), containing AF1 and AF2 (2–4). AF1,which is localized in the NH2-terminal A/B region, is believed to beconstitutive in a cell- and promoter-specific manner and responsiblefor the partial agonist activity of tamoxifen (5, 6). AF2 resides in theCOOH-terminal LBD (region E) and exerts estrogen-dependent tran-scriptional activity by recruiting coactivators such as ERAP160/140(7), RIP140 (8), SRC-1 (9), TIF2/GRIP1 (10), and AIB1 (11). Afterbinding to estrogens, ERa forms a homodimer (12) and binds viaregion C (DNA binding domain) to EREs in the promoter region of anestrogen responsive gene such as TGF-a (13, 14) to regulate geneexpression.

Tamoxifen exhibits a wide range of estrogen-like and antiestrogenactions based on the target tissue being studied (15). The nonsteroidalcompounds of this class are now referred to as SERMs (16). Acompound related to tamoxifen, Ral, is used clinically for the preven-tion of osteoporosis (17, 18) and is being tested against tamoxifen forthe prevention of breast cancer in high-risk women (19, 20). Ral andits analogues are less estrogen-like than tamoxifen in the rodent uterus(21–24) and are more inhibitory than tamoxifen and 4-OHT on thegrowth of breast cancer cells in culture (24). Recent studies demon-strate that Ral is virtually nonestrogenic in the human uterus (17, 25,26). Our previous studies showed that 4-OHT is an agonist and Ral isa complete antagonist in ERa stably transfected MDA-MB-231 hu-man breast cancer cells (27, 28). However, there is currently noadequate molecular mechanism to explain the differences in theestrogenic and antiestrogenic activities of the Ral- and 4-OHT-ERacomplexes.

X-ray crystallographic structures of antiestrogen occupied LBD ofERa has provided valuable insights into the mechanism of antiestro-gen action (29, 30). After 4-OHT or Ral binds to the receptor, helix 12is repositioned to a hydrophobic groove to block the AF2 coactivator(such as GRIP1 and SRC-1) binding. Thus, both 4-OHT and Ralsilence AF2. However, there is a distinctive difference between4-OHT-LBD and Ral-LBD. It appears that the basic amines of 4-OHTand Ral display different relationships (a salt bridgeversusa hydrogenbond) with amino acid D351 on helix 3 of the LBD. We suggest thatthe relationship of the antiestrogenic side chain and the charge atamino acid 351 is critical for estrogen-like actions of SERM-ERacomplexes.

We have described previously an allosteric mechanism that silencesAF-1 activity in the 4-OHT-ERa complex (31). We found that AF-1activity in ERa is actually controlled by the correct positioning ofresidual charge at amino acid 351 aspartate in the LBD (31, 32). Thereportedly constitutively AF-1 activity of the 4-OHT-ERa complex(5) can be silenced allosterically by substituting glycine for aspartatein ERa (31). Alternatively, if the antiestrogenic side chain of 4-OHTis changed from diethylaminoethoxy to an allylcarboxylic acid, thisagain allosterically silences AF-1 activity by repelling the surfaceaspartate at 351 and displacing the surface charge (32). In either case,the complexes lose estrogen-like properties, but antiestrogenic effectsare retained. Clearly, the amino acid at 351 is an important regulatorof the estrogen-like properties of SERMs, and is, therefore, a valuabletarget to probe the molecular mechanism of ERa.

A D351YERa mutant (33) enhances agonist activity of 4-OHT andalters the pharmacology of Ral from an antiestrogen to a partialagonist (34, 35). In this study, we addressed two possible mechanismsthat could explain the enhanced estrogen-like activity of the D351YERa Ral complex: either (a) helix 12 of the Ral-D351YERa complexis now repositioned to reactivate AF2; or (b) Ral-D351YERa be-comes 4-OHT-ERa-like and allosterically activates AF-1. We stablytransfected cDNAs of mutated ERa (Fig. 1) with truncated AF1 ormutated AF2 into MDA-MB-231 ERa-negative breast cancer cells.Biological activities were assessed by measuring endogenous TGF-amRNA levels induced by estrogen or antiestrogens. We compared andcontrasted the effects of Ral and 4-OHT on TGF-a mRNA levels to

Received 10/19/00; accepted 3/1/01.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by the United States Army Medical Research and MaterielCommand Breast Cancer Research Program DAMD17-96-16169 (to H. L.), the LynnSage Breast Cancer Foundation of Northwestern Memorial Hospital, the Avon ProductsFoundation and SPORE in breast cancer CA89018-01 (to V. C. J.).

2 To whom requests for reprints should be addressed, at Robert H. Lurie Comprehen-sive Cancer Center, Northwestern University Medical School, Olson Pavilion 8258, 303East Chicago Avenue, Chicago, IL 60611. Phone: (312) 908-5250; Fax: (312) 908-1372;E-mail: [email protected].

3 The abbreviations used are: ERa, estrogen receptora; AF, activating function; LBD,ligand binding domain; SERM, selective estrogen receptor modulator; TGF, transforminggrowth factor; 4-OHT, 4-hydroxytamoxifen; Ral, raloxifene; E2, estradiol; IC50, 50%inhibitory concentration; GST, glutathioneS-transferase.

3632

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

support the concept that both the 4-OHT-ERa and Ral-D351YERacomplexes display agonist activity without reactivating AF2. Wepropose that the Ral-D351YERa complex displays agonist activity byreactivating ER activity through a triple point mechanism that requiresAF1, an intact helix 12, and an appropriate amino acid at 351.

MATERIALS AND METHODS

Plasmid Construction. pSG5HEGO, a wild-type ERa expression vector,was a kind gift from Professor Pierre Chambon. pSG5HETO, a D351YERaexpression vector, was described previously (36). To make pSG5D351DAF1and pSG5D351YDAF1 expression vectors, the fragments of amino acid 181–595 of ERa or D351YERa were generated from pSG5HEGO or pSG5HETO,respectively, by PCR. The PCR products were inserted into the pSG5 vector atthe EcoRI site. pSG5D351D537 expression vector was constructed using thesame strategy. pSG5D351-G3m and pSG5D351Y3m expression vectors weregenerated based on pSG5HEGO or pSG5HETO, respectively, using the Quick-Change site-directed mutagenesis kit (Stratagene, La Jolla, CA). pGEX-HBD3,an expression vector for a GST fusion protein containing the LBD of thehuman wild-type ERa (GST-HBD3), was provided generously by Dr. MylesBrown (Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA;Ref. 7). pGEX-HBD(D351Y), an expression vector for GST-HBD(D351Y) con-taining the D351Y mutation, was constructed by replacing theHindIII/EcoRIfragment (778 bp) of pGEX-HBD3 with theHindIII/EcoRI fragment ofpSG5HETO. The sequences of all plasmids were confirmed by sequencinganalysis (ABI automated sequencer).

Cell Culture. MDA-MB-231 ERa-negative human breast cancer cellswere obtained originally from American Type Culture Collection (Rockville,MD). MDA-MB-231, wild-type ERa, D351YERa, and the other stable trans-fectants generated for this study were maintained as described previously (34).

GST Pull-down Assay. GST pull-down assays were performed as de-scribed previously (7, 37).35S-labeled SRC-1 and AIB1 were made frompBK-CMV-SRC-1 (kindly provided by Dr. B. W. O’Malley, Baylor College ofMedicine, Houston, TX) and pcDNA3.1-AIB1 (kindly provided by P. Meltzer,NIH, Bethesda, MD), respectively, using anin vitro transcription-coupledtranslation system (Promega Corp., Madison, WI).

Stable Transfection. MDA-MB-231 cells were electroporated with 10mgof pSG5D351–3m, pSG5D351D537, pSG5D351DAF1, pSG5D351-Y3m, orpSG5D351YDAF1 expression vectors and 0.5mg of pBK-CMV (Stratagene,La Jolla, CA) to generate D351–3m, D351D537, D351DAF1, D351Y3m, orD351YDAF1 (Fig. 1), respectively, as described elsewhere (36). Neomycin-resistant clones (two to five clones/stable transfectants) were screened and

characterized for ERa expression using Northern and Western blot analysesand hormone binding assays. Clones with comparable levels of ERa werechosen for further study.

Western Blot Analysis. Twenty-five mg of whole cell lysate were sepa-rated on a 7.5% SDS-PAGE. Anti-ERa polyclonal antibody G20 was fromSanta Cruz Biotechnology (Santa Cruz, CA). Antirabbit IgG conjugated withhorseradish peroxidase (Sigma Chemical Co., St. Louis, MO) was used tovisualize bands using an ECL kit (Amersham Corp., Arlington Heights, IL).

Ligand Binding Assay. 4-OHT and ICI 182,780 were generous gifts fromDr. Alan E. Wakeling (AstraZeneca, Maccelsfield, United Kingdom), and Ral(formerly known as keoxifene) was a gift from Eli Lilly (Indianapolis, IN).

Ligand binding assays were performed following a modified procedure(38). Briefly, for saturation binding assays, the stable transfectants wereincubated with increasing concentrations of [3H]E2 (46 Ci/mmol, AmershamCorp., Arlington Heights, IL) for 2 h at room temperature to obtain totalbinding. To determine nonspecific binding, each concentration of [3H]E2 wascompeted with 400-fold excess of radioactive inert diethylstilbestrol. Thespecific binding was obtained by subtracting the nonspecific binding from thetotal binding. For competition binding assays, the stable transfectants wereincubated with 1 nM [3H]E2 with increasing concentrations of different ligandsincluding 4-OHT, Ral, or ICI 182,780 for 2 h at room temperature. Eachbinding assay was repeated at least three times and theKds for E2, and IC50 forantiestrogens were calculated using GraphPad Prism (GraphPad Prism Soft-ware, Inc., San Diego, CA).

Northern Blot Analysis. TGF-a mRNA levels were assessed by Northernblot analysis as described previously (34).b-Actin mRNA levels were detectedas the loading controls. The band densities were quantitated using ImageQuant(Molecular Dynamics, Sunnyvale, CA). The induction of TGF-a mRNA levelswere standardized byb-actin mRNA levels and expressed as fold of induction(set the control as 1).

Statistics Analysis. The data from ligand binding assays and Northern blotanalyses were analyzed by ANOVA, followed byt test using StatMost (SaltLake City, UT).

RESULTS

Agonist Activity of 4-OHT-ER a Complex Required Both AF1and an Intact Helix 12. AF1 has constitutive transcriptional activityin a cell type- and promoter-specific manner and is believed to beresponsible for the partial agonist activity of 4-OHT (5, 6). However,these early studies were done using transient transfection systems withan artificial reporter gene. 4-OHT is a potent agonist in MDA-MD-231 cells stably transfected with wild-type ERa using theTGF-a geneas anin situ reporter (34). To dissect the roles that AF1 and helix 12play in the estrogenicity of 4-OHT-ERa, we generated stable trans-fectants in MDA-MB-231 cells using D351–3m, D351D537, orD351DAF1 expression vectors (Fig. 1). D351–3m has a triple muta-tion (D538A/E542A/D545A; Ref. 6) that abolishes coactivator bind-ing as measured by a GST-DSRC-1 pull-down assay (data not shown).D351D537 has a COOH-terminal truncation from amino acid 537.

Fig. 2. Expression levels of wild-type and mutant ERa in different ERa stabletransfectants measured by Western blot analysis. Equal protein loading was assured byb-actin level (data not shown).

Fig. 1. Schematic diagram of wild-type and mutant human ERa used in this study.ppp,D538/E542/D545 were mutated to alanines (6). Names on theleft are the abbreviationsused in the text to identify the wild-type or mutant ERa.

3633

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

D351DAF1 has an A/B domain truncation. These mutant receptorswere expressed at similar levels (Fig. 2).

We determined estrogenic activities of E2 and antiestrogens in thesestable transfectants by measuring TGF-a mRNA levels (Fig. 3). E2induced TGF-a mRNA levels in a concentration-dependent mannerwith wild-type ERa (Fig. 3). E2 inducedTGF-a gene expression at aconcentration as low as 0.01 nM. 4-OHT also enhanced TGF-a mRNAlevels in a concentration-dependent manner (data not shown). Thesestudies confirmed a previous report (34), and 1mM 4-OHT, whichsignificantly increased the TGF-a mRNA level (P, 0.05 comparedwith the value in the control group; Fig. 3), was used as our repro-ducible standard. 4-OHT, therefore, was a full agonist in this assay,whereas Ral and ICI 182,780 did not increase TGF-a mRNA levelsbut inhibited E2-induced expression, as complete antagonists. InD351DAF1 cells, which contain an AF1 truncated ERa, E2 enhancedTGF-a mRNA levels, but a higher concentration (1 nM) was neces-sary. 4-OHT did not induce TGF-a gene expression, confirming thatAF1 is required for 4-OHT agonist activity (5). In D351–3m cells, E2still activatedTGF-a gene expression in a concentration-dependentmanner; however, a higher concentration (1 nM) of E2 was againrequired to start stimulatingTGF-a gene expression than that inwild-type ERa cells (0.01 nM). These data support the view that AF1and AF2 have a collaborative effect at low concentrations of E2.Surprisingly, 4-OHT lost its agonist activity and inhibited the E2-induced TGF-a mRNA level in D351–3m cells. 4-OHT became acomplete antagonist. Ral and ICI 182,780 remained antiestrogenic inD351–3m cells. D351D537 failed to mediateTGF-a gene expressionwith E2 or 4-OHT (data not shown). Taken together, these resultsindicate that both AF1 and an intact helix 12 are required for 4-OHTagonist activity.

Lower Ligand Binding Affinity or Lower ER a Expression AreNot the Reason for Loss of 4-OHT Agonist Activity in CellsContaining Mutant Receptor. Loss of agonist activity of 4-OHTcould be the result of: (a) changed ligand binding affinities; (b)changed mutant receptor expression levels induced by different ligandtreatment; or (c) changed transcriptional activities of the mutantreceptors. To address these questions, we first performed saturationbinding assays to determine theKd for E2 and competition binding

assays to establish the IC50 for antiestrogens of the mutant receptors.The results are shown in Table 1. D351–3m and D351DAF1 hadsimilar binding affinities for E2, 4-OHT, Ral, or ICI 182,780 aswild-type ERa. D351D537 had significantly higher binding affinitiesfor 4-OHT, Ral, and ICI 182,780, whereas the truncation did notaffect the binding affinity for E2. We also measured the receptorexpression levels in these stable transfectants after treatment withethanol vehicle, 1 nM E2, 1 mM 4-OHT, 1mM Ral, or 1mM ICI 182,780for 24 h (Fig. 4). Clearly, 4-OHT did not have inhibitory effects on theprotein levels of the mutant receptors. It is interesting to note, how-ever, that truncation (D351D537) or mutation of helix 12 (D351–3m)prevents the anticipated loss of ER protein usually observed with ICI182,780. E2 did not decrease the receptor levels in D351–3m andD351D537, which is consistent with a recent publication by Lonardetal. (39) that intact AF2 is required for down-regulation of ERa by E2.In addition, the receptor level in D351DAF1 cells were not down-regulated by E2 either, suggesting that AF1 also plays a role in thereceptor stability.

The results from the ligand binding assays and Western blot anal-yses suggest that loss of agonist activity of 4-OHT in D351–3m andD351DAF1 cells are likely attributable to changes in the transcrip-tional activities of the complexes rather than lower ligand bindingaffinities of the receptors or loss of receptor protein.

Agonist Activity of Ral-D351Y ERa Is Not Attributable toActivating AF2 by Recruiting AF2 Coactivators. A single pointmutation at D351Y changes the pharmacology of Ral from an antag-onist to a partial agonist, and 4-OHT remained as a potent agonist (27,34). On the basis of the Ral-LBD crystallographic structure (29), it ispossible that the D351Y point mutation caused a conformationalchange of Ral-LBD, which results in repositioning of helix 12 andfacilitating coactivator binding in AF2. To address the possibility ofactivation of AF2 by 4-OHT or Ral, we conducted GST pull-downassays using GST-HBD3 (7) and GST-HBD(D351Y) to determine theinteraction of coactivators (SRC-1 and AIB1) with the LBDs of ERaor D351Y ERa. The results from the pull-down assays are illustratedin Fig. 5. 35S-labeled AIB1 only bound to GST-HBD3 or GST-HBD(D351Y) in the presence of E2. Antiestrogens did not induce theassociation of [35S]AIB1 with GST-HBD3 or GST-HBD(D351Y) (Fig.5, top panel). Ral (shown), 4-OHT, and ICI 182,780 (not shown)inhibited E2-induced binding of [

35S]AIB1 to GST-HBD3 and GST-HBD(D351Y). Similar results were obtained when

35S-labeled SRC-1was used (Fig. 5,bottom panel). In addition, 4-OHT and Ral did notinduce any proteins specifically binding to the GST-HBD3 or GST-HBD(D351Y) when [

35S]methionine metabolically labeled whole cellextracts from MDA-MB-231 and MCF-7 cells were used (data notshown). Therefore, we concluded that agonist activities of 4-OHT-ERa, 4-OHT-D351Y ERa, and Ral-D351Y ERa were not attributedto recruiting AF2 coactivators to the receptors. We believe it isunlikely that the D351Y mutation facilitates reorientation of helix 12to seal 4-OHT or Ral in the LBD.

Fig. 3. Northern blot analysis of TGF-a mRNA level in the wild-type ERa, D35DAF1,and D351–3m cells. The cells were treated with ethanol (Control), 0.001–10 nM of E2, 1mM 4-OHT, 1 mM Ral, 1 mM ICI 182,780 (ICI), or combination of 10 nM E2 with 1 mMantiestrogen for 24 h. Experiments were repeated three times, and quantitative TGF-aband densities standardized byb-actin were presented as fold of induction (set control as1) and plotted as means;bars,SD. p, compared with the controls, TGF-a mRNA levelswere significantly enhanced (P# 0.05).

Table 1 Ligand binding characteristics of mutant ERa stable transfectants

Transfectants Kd (nM)

IC50 (nM)

4-OHT Ral ICI

Wild-type ERa 0.506 0.08 0.616 0.37 1.016 0.52 3.086 0.61D351-3m 0.316 0.02 1.426 1.11 1.366 0.64 3.716 1.14D351D537 0.696 0.22 0.356 0.20a 0.316 0.12a 1.386 0.47a

D351DAF1 0.556 0.15 1.616 0.17 1.846 0.14 3.886 1.72D351YER 1.626 0.38a 4.396 1.30a 4.566 1.19a 3.486 0.35D351Y3m 0.886 0.26 1.766 0.80 5.206 1.05a 5.676 0.78D351YDAF1 0.646 0.06 3.576 1.07a 9.176 2.23a 4.196 1.22a Compared with the values in wild-type ERa cells,P # 0.05.

3634

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

Both AF1 and an Intact Helix 12 Are Required for TGF-a GeneExpression Induced by Ral with D351YERa. To determinewhether AF1 or AF2 alone is sufficient to mediateTGF-a geneexpression induced by 4-OHT or Ral, we established stable transfec-tants D351Y3m and D351YDAF1 in MDA-MB-231 cells using re-spective expression vectors (Fig. 1), as described in “Materials andMethods.” D351Y3m and D351YDAF1 cells had comparable expres-sion levels of the mutant receptors as that in D351YERa cells (Fig. 2).Although D351YERa had a significantly lower binding affinities forE2, 4-OHT, and Ral than wild-type ERa, D351Y3m and D351YDAF1had similar ligand binding characteristics as D351Y ERa (Table 1).

E2 (Fig. 6), 4-OHT, and Ral (data not shown) induced TGF-amRNA level in a concentration-dependent manner with D351YERa.4-OHT was an agonist, and Ral showed partial agonist activity. Thesedata were reported previously (27) and are our reproducible standard.

D351YDAF1 cells, expressing AF1 truncated D351Y ERa, displayed adecreased E2-induced TGF-a expression. As observed in D351DAF1cells (Fig. 3), 4-OHT became a complete antiestrogen,i.e., 4-OHT didnot induce TGF-a mRNA level and inhibited E2-induced transcriptionalactivity (Fig. 6). Ral also lost its partial agonist activity. These resultsindicated that AF1 is essential for the agonist activities of 4-OHT and Ralmediated by D351Y ERa. We were surprised to find that E2 (up to 10nM), 1 mM 4-OHT, and Ral failed to induceTGF-a mRNA levels inD351Y3m cells. Loss of agonist activities of 4-OHT and Ral inD351Y3m and D351YDAF1 cells were not attributed to decreased levelsof the receptors (Fig. 7). Thus, both AF1 and an intact helix 12 ofD351YERa are necessary for the agonist activities of 4-OHT and Ral.

DISCUSSION

TGF-a gene expression can be regulated by a variety of agentsincluding estrogens. E2 increases the expression of TGF-a mRNA andsecretion of TGF-a protein (13, 14, 40, 41). There are two imperfect13-bp palindromic estrogen response element-like sequences that liebetween2260 and2203 within the human TGF-a 59-flanking region(13, 42). ER-mediatedTGF-a gene expression requires the DNAbinding domain of ERa (43). We have developed an assay system inMDA-MB-231 breast cancer cells stably transfected with cDNA ofERa to study the structure-function relationship of SERM-ERa com-plexes using a TGF-a targetin situ (36, 44, 45). The target system hasthe advantage of being able to distinguish between SERM-ERa com-plexes in the context of a breast cancer cell.

The action of E2 on TGF-a gene expression can be mediated byAF2 at the COOH terminus or/and AF1 at the NH2 terminus of theERa. When estrogens are present, the hinge chain (NVVPY) betweenhelix 11 and helix 12 is closer to helix 3, and helix 12 is positionedover the ligand binding pocket (Fig. 8A; Refs. 29 and 30). A hydro-phobic cleft is formed for AF2 coactivator binding (46). Thus, E2activates AF2 by recruiting AF2 coactivators. E2 also activatesTGF-agene expression through AF1 (Fig. 3, D351–3m cells). However, AF1or AF2 alone was not sufficient to maximize induction at lowerconcentrations of E2. Previous reports (4, 6, 47) suggest that AF1 andAF2 have a synergistic effect through an interaction with SRC-1 (48).

4-OHT is believed to exert antiestrogenic activity by silencing thetranscriptional activity of AF2 by repositioning helix 12 to block thecoactivator binding site (Fig. 8B). The agonist activity of 4-OHT isbelieved to be mediated through AF1 in a cell- and promoter-depen-dent manner (5, 6). 4-OHT inducedTGF-a gene expression in wild-type ERa cells (Fig. 3) without inducing coactivator binding to

Fig. 4. Representative Western blot analysis of ERa protein levels in wild-type ERa,D351DAF1, D351–3m, and D351D537 cells. The cells were treated with ethanol (Con-trol), 10 nM E2, 1 mM 4-OHT, 1 mM Ral or 1 mM ICI 182,780 (ICI) for 24 h. Rabbitpolyclonal anti-ERa antibody G20 was used, and the Western blot analysis was performedas described in “Materials and Methods.” The experiments were repeated three times.

Fig. 5. Interaction between LBDs of wild-typeERa (HBD3) and D351Y ERa (HDB(D351Y)) withSRC-1 or AIB1 in vitro. 35S-labeled AIB1 orSRC-1 was incubated with GST-HBD3 or GST-HBD(D351Y) in the presence of ethanol (Control),10 nM E2, 1 mM 4-OHT, 1 mM Ral, 1 mM ICI182,780 (ICI), or combination of 10 nM E2 and1 mM raloxifene (Ral1E2). The input lane repre-sents 10% of the total amount of35S-labeled AIB1or SRC-1 used in the pull-down assay.

3635

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

GST-HBD3 (Fig. 5), suggesting that 4-OHT can activateTGF-a geneexpression through the AF1 domain. This was confirmed because4-OHT did not induce TGF-a mRNA when liganded to D351DAF1,which does not contain the AF1 domain (Fig. 3). However, 4-OHTalso failed to enhance the TGF-a mRNA level in D351–3m cells (Fig.3), which is consistent with a previous report that AF2 is required foragonist activity of 4-OHT (49). These results indicated that the agonistactivity of 4-OHT was not simply estrogen like, because E2 caninduce the TGF-a mRNA level through both D351DAF1 andD351–3m (Fig. 3). Most importantly, 4-OHT inhibited E2-inducedTGF-a mRNA and behaved as a complete antagonist in D351–3mcells. 4-OHT also lost its agonist activity in D351D537 cells thatcontain a COOH-terminal truncated ERa with intact AF1. Thus, AF1alone is not sufficient to mediateTGF-a gene expression with 4-OHTin the breast cancer cell context. These results demonstrated that bothhelix 12 and AF1 are essential for 4-OHT agonist activity with respect

to TGF-a gene expression in MDA-MB-231 cells and that helix 12and AF1 must have a collaborative interaction to form a transcrip-tional unit for coactivators to bind and to initiate gene expression inthe presence of 4-OHT. Because the crystallographic structure offull-length ERa has not yet been resolved, we do not know the precise

Fig. 6. Northern blot analysis of TGF-a mRNA level in the D351YERa, D351YDAF1,and D351Y3m cells. The experiments were done and presented in the same way as inFig. 3.

Fig. 7. Representative Western blot analysis of ERa protein levels in D351YERa,D351YDAF1, and D351Y3m cells. The experiments were performed as described inFig. 4.

Fig. 8. Crystallographic structures of diethylstilbestrol-bound (A), 4-OHT-bound (B),and Ral-bound (C) LBD of ERa (29, 30). The diagrams were generated using Insight II(Molecular Simulations, Inc., San Diego, CA).

3636

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

position of AF1 relative to displaced helix 12 in the presence ofdifferent ligands. Nevertheless, it is useful to propose a workingmodel to explain the observed agonist activities of the antiestrogen-ERa complexes so that additional structure-function studies can chal-lenge the veracity of the concept.

Coactivators SRC-1 (49), GRIP1, RAC3, and CBP/p300 (50) arereported to enhance agonist activity of 4-OHT in transient transfectionassays. SRC-1, GRIP1, RAC3, CBP/p300, and pCAF bind to the A/Bdomain of ERa in vitro or in yeast or the mammalian two-hybridsystem (49–51). It is logical to conclude that these coactivatorsaugment agonist activity of 4-OHT through the AF1 domain of ERa.However, SRC family coactivators do not bind to the full length ofERa in a 4-OHT-dependent mannerin vitro (52), although the COOHterminus of GRIP1 binds to ERa in a hormone-independent manner(50). Moreover, AF2 is also required for the effect of SRC-1 (49). Itis therefore reasonable to hypothesize that novel coactivators might be

involved to bridge AF1 and AF2 of ERa in combination with the SRCfamily and CBP/p300 coactivators for agonist activity of 4-OHT.

Ral competes with E2 for ERa and causes a conformational changeof the receptor that prevents AF2 activation (Ref. 29; Fig. 8C). Theantiestrogen side chain of Ral extends out of the ERa complex andinteracts with amino acid 351. Helix 12 is repositioned and silencesAF2 activity by preventing coactivator binding (Ref. 29; Fig. 5).However, unlike 4-OHT, Ral did not induce TGF-a mRNA and wasa complete antagonist in MDA-MB-231 cells with wild-type ERa(Fig. 3). Although crystal structures of 4-OHT-LBD and Ral-LBD(29, 30) are very similar, there are significant differences in the twostructures. The piperidine group of Ral is forced to project outwardfrom the receptor surface to shield D351 (Fig. 8C), whereas 4-OHTdoes not have an equivalent bulky group to do so (Fig. 8B). Thesedifferences in the structures might affect the relationship between AF1and AF2 and disrupt the surface for the binding of coactivator(s). Our

Fig. 9. Molecular modeling of the surface struc-tures of 4-OHT-LBD(wild type) (A) or raloxifene-LBD(wild type) (B) (29, 30) and Ral-LBD(D351Y) (C).D351 replaced with Y351 in Ral-bound ERa LBD.To avoid steric clashes, Y351 is placed in a rotomerthat projects the side chain upward. The side chainof Y351 is out of reach of the Ral side chain.Tyrosine residues typically lay down on the surfaceof proteins. In the ERR.pdb structure (29), smallrearrangements in structure around Y351 are re-quired to sterically accommodate the side chain. Ifthis happens, the phenolic side chain would beoriented in rotomer #2. On the right panel for eachis a schematic diagram for a proposed mechanismfor agonist activity of 4-OHT and Ral.A, when4-OHT binds, AF2 and AF1 form a proper dockingsite for coactivator(s) X. X bridges 4-OHT-ERawith the general transcriptional factors (GTFs) toinitiate TGFa gene expression;B, the piperidineside chain of Ral shields the charge of D351 anddisturbs the local charge available for binding co-activators. As a result, AF1 and AF2 cannot col-laborate properly, and TGF-a is silenced.C, thetyrosine at amino acid 351 changed the local chargeavailable for coactivator binding because the piper-idine can no longer shield the charge. Conforma-tion of Ral-D351Y ERa to be 4-OHT-ERa-likeandTGFa gene is switched on.

3637

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

hypothesis is that the close fit of Ral and aspartic acid 351 neutralizesany charge so that a coactivator cannot bind. In contrast, 4-OHTcannot neutralize the charge and binds to a putative coactivator. Thismodel might explain why Ral had no agonist activity with wild-typeERa. We have shown previously that Ral, unlike the pure antiestrogenICI 182,780, does not have a significant effect on ERa levels (45).Therefore, rapid destruction of ERa cannot be the mechanism forminimal estrogen-like activity of the Ral-ERa complex. As a result,we conclude that Ral inhibits transcriptional activity in MDA-MB-231 cells by silencing both AF1 and AF2 of ERa.

When D351 was replaced with a tyrosine, 4-OHT remained a potentagonist, whereas Ral was converted from an antagonist to a partialagonist. A single point mutation at the codon for amino acid 351 thatconverts an aspartic acid to a tyrosine results in lower binding affin-ities for ligands including E2, 4-OHT, and Ral (Table 1), which mightbe attributable to the bulkier side chain of tyrosine (-CH2-C6H4OHversus-CH2COOH). However, the agonist activity 4-OHT, measuredby induction of TGF-a mRNA, was enhanced significantly (P , 0.05compared with that in wild-type ERa cells), and the Ral-ERa com-plex became estrogenic (Fig. 6). Our pull-down data (Fig. 5) showedthat neither 4-OHT nor Ral induced the LBD of D351YERa to bindto SRC-1 or AIB1, suggesting that 4-OHT or Ral did not activate AF2of D351Y ERa by repositioning helix 12 to the E2-LBD conformation(Fig. 8A; Ref. 29). Although it is reported that TIF-2.1 enhancesagonist activity of 4-OHT-D351YERa complex in a transient trans-fection assay, 4-OHT does not induce TIF-2.1 binding to D351YERain vitro (52), suggesting that some other factors in cells might alsoinvolve agonist activity of 4-OHT-D351YERa. The second possibil-ity was that AF1 alone or a combination of AF1 and an intact helix 12contributed to the estrogenic activity of D351YERa. Fig. 6 shows thatneither 4-OHT nor Ral had estrogenic activities in D351Y3m orD351YDAF1 cells. Thus, D351YERa needs both AF1 and an intacthelix 12 for the induction of TGF-a mRNA in the presence of 4-OHTor Ral. We concluded that the behavior of Ral-D351YERa is con-sistent with the 4-OHT-ERa complex. Introducing a negativelycharged amino acid at 351 that is out of reach of the influence of theantiestrogen side chain of Ral (Fig. 9C) facilitated an allosteric acti-vation of AF-1 in the Ral-D351YERa complex.

On the basis of the present observations, the crystallographic in-formation (29, 30), and our previous studies with D351GERa (31)and the 4-OHT analogue, GW7064 (32), we propose a novel inter-pretation of the data that provides an explanation for the agonistactivities of the antiestrogen-ERa complex. Norriset al. (53) havesuggested that the 4-OHT ERa complex binds to a series of peptidesspecifically at distinctly different places on ERa from that for theknown AF2 coactivators. In the context of the MDA-MB-231 cells,both 4-OHT and E2 ERa complexes initiate a powerful induction oftheTGF-a gene through a direct signal transduction pathway (34). Itis clear that E2 activates AF2 by recruiting coactivators (7–9, 11, 46),but the alternate site for coactivator binding on the 4-OHT-ERacomplex based on phage display is unknown (53). We suggest that ourcurrent studies provide insight to resolve the problem. Amino acid 351(aspartic acid) is on the surface of the LBD and is exposed whenantiestrogens move helix 12 to block AF2 (Fig. 8,B andC). We haveshown in the present study that replacement of aspartate 351 by atyrosine enhances the estrogenic activity of Ral without reactivatingAF2 and repositioning helix 12. The negative charge is out of reach ofthe influence of the antiestrogenic side chain of Ral (Fig. 9C) and isavailable for coactivator binding. This view is supported by the recentreports (31, 52) that an uncharged amino acid at 351 does notinfluence antiestrogenic activity but silences the estrogen-like prop-erties of the 4-OHT-ERa complex. Additionally, redistribution of the

surface charge at D351 with the tamoxifen analogue GW7064 resultsin the loss of estrogen-like properties in the ER complex (32).

However, the estrogen-like properties of the 4-OHT-ERa or Ral-D351YERa complexes are not just dependent upon the exposednegative charge of the amino acid at 351 (31). The change of three keynegatively charged amino acids (D538, E542, and D545) to alanineson helix 12 suppresses the estrogen-like actions of the 4-OHT-ERacomplex without affecting antiestrogenic actions of the complex (Fig.3A). This principle also applies for the estrogen-like action of the4-OHT-D351YERa and Ral-D351YERa complexes. The AF1 regionof ERa had been thought previously to be constitutive and ligandindependent (4, 5). We now show that AF-1 activation is controlledallosterically by the 4-OHT-ERa, 4-OHT-D351YERa, and the Ral-D351YERa complexes.

In summary, we have advanced previous studies (31, 32) to con-solidate a working model of the SERM-ERa complex that incorpo-rates a novel concept of allosteric activation and silencing of AF-1.We envision a triple-point coactivator complex interaction that en-ables an antiestrogenic compound such as 4-OHT [or Ral and EM652(45) with D351YERa] to initiate TGF-a gene transcription (Fig. 9).The coactivators in MDA-MB-231 cells require AF1, an intact helix12, and an appropriately positioned negative charge at amino acid 351to enhance the estrogen-like actions of a nonsteroidal antiestrogen.

ACKNOWLEDGMENTS

We thank Drs. Debra A. Tonetti and Anait S. Levenson for valuablediscussions and Henry Muenzner for excellent technical assistance.

REFERENCES

1. McKenna, N. J., Lanz, R. B., and O’Malley, B. W. Nuclear receptor coregulators:cellular and molecular biology. Endocr. Rev.,20: 321–344, 1999.

2. Kumar, V., Green, S., and Staub, A. Localisation of the oestradiol-binding andputative DNA-binding domains of the human oestrogen receptor. EMBO J.,5:2231–2236, 1986.

3. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J. R., and Chambon, P. Functionaldomains of the human estrogen receptor. Cell,51: 941–951, 1987.

4. Tora, L., White, J., Brou, C., Tasset, D., Webster, N., Scheer, E., and Chambon, P.The human estrogen receptor has two independent nonacidic transcriptional activationfunctions. Cell,59: 477–487, 1989.

5. Berry, M., Metzger, D., and Chambon, P. Role of the two activating domains of theoestrogen receptor in the cell-type and promoter-context dependent agonistic activityof the anti-oestrogen 4-hydroxytamoxifen. EMBO J.,9: 2811–2818, 1990.

6. Tzukerman, M. T., Esty, A., Santiso-Mere, D., Danielian, P., Parker, M. G., Stein,R. B., Pike, J., and McDonnell, D. P. Human estrogen receptor transactivationalcapacity is determined by both cellular and promoter context and mediated by twofunctionally distinct intramolecular regions. Mol. Endocrinol.,94: 21–30, 1994.

7. Halachmi, S., Marden, E., Martin, G., MacKay, H., Abbondanza, C., and Brown, M.Estrogen receptor-associated proteins: possible mediators of hormone-induced tran-scription. Science (Wash. DC),264: 1455–1458, 1994.

8. Cavailles, V., Dauvois, S., Danielian, P. S., and Parker, M. G. Interaction of proteinswith transcriptionally active estrogen receptors. Proc. Natl. Acad. Sci. USA,91:10009–10013, 1994.

9. Onate, S. A., Tsai, S. Y., Tsai, M. J., and O’Malley, B. W. Sequence and character-ization of a coactivator for the steroid hormone receptor superfamily. Science (Wash.DC), 270: 1354–1357, 1995.

10. Hong, H., Kohli, K., Garabedian, M. J., and Stallcup, M. R. GRIP1, a transcriptionalcoactivator for the AF-2 transactivation domain of steroid, thyroid, retinoid, andvitamin D receptors. Mol. Cell. Biol.,17: 2735–2744, 1997.

11. Anzick, S. L., Kononen, J., Walker, R. L., Azorsa, D. O., Tanner, M. M., Guan, X. Y.,Sauter, G., Kallioniemi, O. P., Trent, J. M., and Meltzer, P. S. AIB1, a steroid receptorcoactivator amplified in breast and ovarian cancer. Science (Wash. DC),277: 965–968, 1997.

12. Kumar, V., and Chambon, P. The estrogen receptor binds tightly to its responsiveelement as a ligand-induced homodimer. Cell,55: 145–156, 1988.

13. Saeki, T., Cristiano, A., Lynch, M. J., Brattain, M., Kim, N., Normanno, N., Kenney,N., Ciardiello, F., and Salomon, D. S. Regulation by estrogen through the 59-flankingregion of the transforming growth factora gene. Mol. Endocrinol.,5: 1955–1963,1991.

14. El-Ashry, D., Chrysogelos, S. A., Lippman, M. E., and Kern, F. G. Estrogen inductionof TGF-a is mediated by an estrogen response element composed of two imperfectpalindromes. J. Steroid Biochem. Mol. Biol.,59: 261–269, 1996.

15. MacGregor, J. I., and Jordan, V. C. Basic guide to the mechanisms of antiestrogenaction. Pharmacol. Rev.,50: 151–196, 1998.

3638

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

16. Levenson, A. S., and Jordan, V. C. Selective oestrogen receptor modulation: molec-ular pharmacology for the millennium. Eur. J. Cancer,35: 1628–1639, 1999.

17. Delmas, P. D., Bjarnason, N. H., Mitlak, B. H., Ravoux, A. C., Shah, A. S., Huster,W. J., Draper, M., and Christiansen, C. Effects of raloxifene on bone-mineral density,serum-cholesterol concentrations, and uterine endometrium in postmenopausalwomen. N. Engl. J. Med.,337: 1641–1647, 1997.

18. Ettinger, B., Black, D. M., Mitlak, B. H., Knickerbocker, R. K., Nickelsen, T.,Genant, H. K., Christiansen, C., Delmas, P. D., Zanchetta, J. R., Stakkestad, J., Gluer,C. C., Krueger, K., Cohen, F. J., Eckert, S., Ensrud, K. E., Avioli, L. V., Lips, P., andCummings, S. R. Reduction of vertebral fracture risk in postmenopausal women withosteoporosis treated with raloxifene: results from a 3-year randomized clinical trial.Multiple Outcomes of Raloxifene Evaluation (MORE) Investigators [published erra-tum appears in JAMA,282: 2124, 1999]. JAMA,282: 637–645, 1999.

19. Cummings, S. R., Eckert, S., Krueger, K. A., Grady, D., Powles, T. J., Cauley, J. A.,Norton, L., Nickelsen, T., Bjarnason, N. H., Morrow, M., Lippman, M. E., Black, D.,Glusman, J. E., Costa, A., and Jordan, V. C. The effect of raloxifene on risk of breastcancer in postmenopausal women: results from the MORE randomized trial. MultipleOutcomes of Raloxifene Evaluation. JAMA,281: 2189–2197, 1999.

20. Jordan, V. C., and Morrow, M. Tamoxifen, raloxifene, and the prevention of breastcancer. Endocr. Rev.,20: 253–278, 1999.

21. Black, L. J., and Goode, R. L. Evidence for biological action of the antiestrogensLY117018 and tamoxifen by different mechanisms. Endocrinology,109: 987–989,1981.

22. Jordan, V. C., and Gosden, B. Inhibition of the uterotropic activity of estrogens andantiestrogens by the short-acting antiestrogen LY117018. Endocrinology,113: 463–468, 1983.

23. Jones, C. D., Jevnikar, M. G., Pike, A. J., Peters, M. K., Black, L. J., Thompson,A. R., Falcone, J. F., and Clemens, J. A. Antiestrogens. 2. Structure-activitystudies in a series of 3-aroyl-2-arylbenzo[b]thiophene derivatives leading to[6-hydroxy-2-(4-hydroxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]-phenyl]methanone hydrochloride (LY156758), a remarkably effective estrogenantagonist with only minimal intrinsic estrogenicity. J. Med. Chem.,27: 1057–1066, 1984.

24. Grese, T. A., Sluka, J. P., Bryant, H. U., Cullinan, G. J., Glasebrook, A. L., Jones,C. D., Matsumoto, K., Palkowitz, A. D., Sato, M., Termine, J. D., Winter, M. A.,Yang, N. N., and Dodge, J. A. Molecular determinants of tissue selectivity inestrogen-receptor modulators. Proc. Natl. Acad. Sci. USA,94: 14105–14110, 1997.

25. Draper, M. W., Flowers, D. E., Huster, W. J., Neild, J. A., Harper, K. D., and Arnaud,C. A controlled trial of raloxifene (LY139481) HCl: impact on bone turnover andserum lipid profile in healthy postmenopausal women. J. Bone Miner. Res.,11:835–842, 1996.

26. Lufkin, E. G., Whitaker, M. D., Nickelsen, T., Argueta, R., Caplan, R. H.,Knickerbocker, R. K., and Riggs, B. L. Treatment of established postmenopausalosteoporosis with raloxifene: a randomized trial. J. Bone Miner. Res.,13: 1747–1754, 1998.

27. Levenson, A. S., Catherino, W. M., and Jordan, V. C. Estrogenic activity is increasedfor an antiestrogen by a natural mutation of the estrogen receptor. J. Steroid Biochem.Mol. Biol., 60: 261–268, 1997.

28. Levenson, A. S., Svoboda, K. M., Kwaan, H. C., and Jordan, V. C. Agonist activityof antiestrogen-receptor complexes to regulate urokinase plasminogen activator(uPA) and plasminogen activator inhibitor type 1 (PAI-1) endogenous gene expres-sion in breast cancer cells. Cancer Lett.,125: 215–220, 1998.

29. Brzozowski, A. M., Pike, A. C. W., Dauter, Z., Hubbard, R. E., Bonn, T., Engstro¨m,O., Öhman, L., Greene, G. L., Gustafsson, J. Å., and Carlquist, M. Molecular basis ofagonism and antagonism in the oestrogen receptor. Nature (Lond.),389: 753–758,1997.

30. Shiau, A. K., Barstad, D., Loria, P. M., Cheng, L., Kushner, P. J., Agard, D. A., andGreene, G. L. The structural basis of estrogen receptor/co-activator recognition andthe antagonism of this interaction by tamoxifen. Cell,95: 927–937, 1998.

31. MacGregor Schafer, J., Liu, H., Bentrem, D. J., Zapf, J. W., and Jordan, V. C.Allosteric silencing of activating function 1 in the 4-hydroxytamoxifen estrogenreceptor complex is induced by substituting glycine for aspartate at amino acid 351.Cancer Res.,60: 5097–5105, 2000.

32. Bentrem, D. J., Dardes, R. C., Liu, H., MacGregor-Schafer, J., Zapf, J. W., andJordan, V. C. Molecular mechanism of action at estrogen receptora of a newclinically relevant antiestrogen (GW7604) related to tamoxifen. Endocrinology,142:838–846, 2001.

33. Wolf, D. M., and Jordan, V. C. The estrogen receptor from a tamoxifen-stimulatedMCF-7 tumor variant contains a point mutation in the ligand binding domain. BreastCancer Res. Treat.,31: 129–138, 1994.

34. Levenson, A. S., Tonetti, D. A., and Jordan, V. C. The oestrogen-like effect of4-hydroxytamoxifen on induction of transforming growth factora mRNA in MDA-MB-231 breast cancer cells stably expressing the oestrogen receptor. Br. J. Cancer,77: 1812–1819, 1998.

35. Levenson, A. S., and Jordan, V. C. The key to the antiestrogenic mechanism ofraloxifene is amino acid 351 (aspartate) in the estrogen receptor. Cancer Res.,58:1872–1875, 1998.

36. Catherino, W. H., Wolf, D. M., and Jordan, V. C. A naturally occurring estrogenreceptor mutation results in increased estrogenicity of a tamoxifen analog. Mol.Endocrinol.,9: 1053–1063, 1995.

37. Hanstein, B., Eckner, R., DiRenzo, J., Halachmi, S., Liu, H., Searcy, B., Kurokawa,R., and Brown, M. p300 is a component of an estrogen receptor coactivator complex.Proc. Natl. Acad. Sci. USA,93: 11540–11545, 1996.

38. Liu, H., Wormke, M., Safe, S. H., and Bjeldanes, L. F. Indolo[3,2-b]carbazole: adietary-derived factor that exhibits both antiestrogenic and estrogenic activity. J. Natl.Cancer Inst.,86: 1758–1765, 1994.

39. Lonard, D. M., Nawaz, Z., Smith, C. L., and O’Malley, B. W. The 26S proteasomeis required for estrogen receptor-a and coactivator turnover and for efficient estrogenreceptor-a transactivation. Mol. Cell,5: 939–948, 2000.

40. Liu, S. C., Sanfilippo, B., Perroteau, I., Derynck, R., Salomon, D. S., and Kidwell,W. R. Expression of transforming growth factora (TGF a) in differentiated ratmammary tumors: estrogen induction of TGFa production. Mol. Endocrinol.,1:683–692, 1987.

41. Bates, S. E., Davidson, N. E., Valverius, E. M., Freter, C. E., Dickson, R. B., Tam,J. P., Kudlow, J. E., Lippman, M. E., and Salomon, D. S. Expression of transforminggrowth factor a and its messenger ribonucleic acid in human breast cancer: itsregulation by estrogen and its possible functional significance. Mol. Endocrinol.,2:543–555, 1988.

42. Jakobovits, E. B., Schlokat, U., Vannice, J. L., Derynck, R., and Levinson, A. D. Thehuman transforming growth factora promoter directs transcription initiation from asingle site in the absence of a TATA sequence. Mol. Cell. Biol.,8: 5549–5554, 1988.

43. Vyhlidal, C., Samudio, I., Kladde, M. P., and Safe, S. Transcriptional activation oftransforming growth factora by estradiol: requirement for both a GC-rich site and anestrogen response element half-site. J. Mol. Endocrinol.,24: 329–338, 2000.

44. Jiang, S. Y., Langan-Fahey, S. M., Stella, A. L., McCague, R., and Jordan, V. C. Pointmutation of estrogen receptor (ER) in the ligand-binding domain changes the phar-macology of antiestrogens in ER-negative breast cancer cells stably expressingcomplementary DNAs for ER. Mol. Endocrinol.,6: 2167–2174, 1992.

45. MacGregor Schafer, J. I., Liu, H., Tonetti, D. A., and Jordan, V. C. The interactionof raloxifene and the active metabolite of the antiestrogen EM800 (SC5705) with thehuman estrogen receptor (ER). Cancer Res.,59: 4308–4313, 1999.

46. Feng, W., Ribeiro, R. C., Wagner, R. L., Nguyen, H., Apriletti, J. W., Fletterick, R. J.,Baxter, J. D., Kushner, P. J., and West, B. L. Hormone-dependent coactivator bindingto a hydrophobic cleft on nuclear receptors. Science (Wash. DC),280: 1747–1749,1998.

47. Kraus, W. L., McInerney, E. M., and Katzenellenbogen, B. S. Ligand-dependent,transcriptionally productive association of the amino- and carboxyl-terminal regionsof a steroid hormone nuclear receptor. Proc. Natl. Acad. Sci. USA,92: 12314–12318,1995.

48. McInerney, E. M., Tsai, M. J., O’Malley, B. W., and Katzenellenbogen, B. S.Analysis of estrogen receptor transcriptional enhancement by a nuclear hormonereceptor coactivator. Proc. Natl. Acad. Sci. USA,93: 10069–10073, 1996.

49. Smith, C. L., Nawaz, Z., and O’Malley, B. W. Coactivator and corepressor regulationof the agonist/antagonist activity of the mixed antiestrogen, 4-hydroxytamoxifen.Mol. Endocrinol.,11: 657–666, 1997.

50. Webb, P., Nguyen, P., Shinsako, J., Anderson, C., Feng, W., Nguyen, M. P., Chen,D., Huang, S. M., Subramanian, S., McKinerney, E., Katzenellenbogen, B. S.,Stallcup, M. R., and Kushner, P. J. Estrogen receptor activation function 1 works bybinding p160 coactivator proteins. Mol. Endocrinol.,12: 1605–1618, 1998.

51. Onate, S. A., Boonyaratanakornkit, V., Spencer, T. E., Tsai, S. Y., Tsai, M. J.,Edwards, D. P., and O’Malley, B. W. The steroid receptor coactivator-1 containsmultiple receptor interacting and activation domains that cooperatively enhance theactivation function 1 (AF1) and AF2 domains of steroid receptors. J. Biol. Chem.,273: 12101–12108, 1998.

52. Anghel, S., Perly, V., Melancon, G., Barsalou, A., Chagnon, S., Rosenauer, A.,Miller, J., and Mader, S. Aspartate 351 of estrogen receptora is not crucial for theantagonist activity of antiestrogens. J. Biol. Chem.,275: 20867–20872, 2000.

53. Norris, J. D., Paige, L. A., Christensen, D. J., Chang, C. Y., Huacani, M. R., Fan, D.,Hamilton, P. T., Fowlkes, D. M., and McDonnell, D. P. Peptide antagonists of thehuman estrogen receptor. Science (Wash. DC),285: 744–746, 1999.

3639

SILENCING AND REACTIVATION OF ER BY SERMS

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

2001;61:3632-3639. Cancer Res Hong Liu, Eun-Sook Lee, Alexander De Los Reyes, et al.

ComplexαModulator-Estrogen Receptor Silencing and Reactivation of the Selective Estrogen Receptor

Updated version

http://cancerres.aacrjournals.org/content/61/9/3632

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/61/9/3632.full#ref-list-1

This article cites 51 articles, 19 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/61/9/3632.full#related-urls

This article has been cited by 13 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/61/9/3632To request permission to re-use all or part of this article, use this link

on June 5, 2021. © 2001 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from