HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (14) Citing Articles via Google Scholar Google Scholar Articles by Murphy, L C Articles by Shiu, R Search for Related Content PubMed PubMed Citation Articles by Murphy, L C Articles by Shiu, R

¹ Department of Biochemistry and Medical Genetics, University of Manitoba, Winnipeg, Manitoba, Canada
² Manitoba Institute of Cell Biology, University of Manitoba, Winnipeg, Mannitoba, Canada
³ Department of Physiology, University of Manitoba, Winnipeg, Manitoba, Canada

(Requests for offprints should be addressed to L C Murphy, Manitoba Institute of Cell Biology, University of Manitoba, 675 McDermot Ave., Winnipeg, Manitoba R3E 0 V9, Canada; Email: lcmurph{at}cc.umanitoba.ca)

	Abstract

Top Abstract Introduction Materials and methods Results Discussion References

 
To investigate the effect of altered oestrogen receptor (ER) and ERß expression on oestrogen and anti-oestrogen action in breast cancer, we have stably expressed an inducible ERß1 in MCF7 breast cancer cells. Stably expressing clones were isolated and over-expression of ERß1 correlated with increased levels of specific radiolabelled oestradiol (E2) binding. Increased ERß1 did not affect endogenous levels of ER but increased progesterone receptor (PR) levels. Over-expression of ERß1 reduced growth responses to E2 in contrast to little if any effect of over-expression of ER. In oestrogen-replete conditions, over-expression of ERß1 but not ER reduced proliferation. Over-expression of ERß1 did not result in anti-oestrogen resistance but was associated with increased sensitivity to 4-hydroxytamoxifen. Our results suggested that over-expression of ERß1 in the presence of an endogenously expressed ER was associated with tamoxifen sensitivity but may negatively modulate ER-mediated growth. However, not all ER activities were inhibited since endogenous PR expression was increased by both ER and ERß1 over-expression. These data paralleled those seen in some in vivo studies showing a relationship between PR and ERß expression as well as ERß expression and tamoxifen sensitivity of ER-positive breast cancer patients. These models are relevant and will be useful for dissecting the role of ERß1 expression in ER-positive breast cancer.

	Introduction

Top Abstract Introduction Materials and methods Results Discussion References

 
Oestrogens are considered to be major driving forces in breast tumourigenesis and breast cancer progression (Colditz 1998, Clemons & Goss 2001). Current evidence suggests that oestrogen action is primarily mediated through two receptors, oestrogen receptor (ER) and ERß (Kuiper et al. 1996). These structurally related receptors are generally considered to be ligand-regulated transcription factors which classically modulate target gene transcription by binding to oestrogen responsive sequences (ERE) in target gene promoters (McDonnell & Norris 2002). There is evidence that ER and ERß can be expressed together in some cell types and independently expressed in others (Dotzlaw et al. 1997, Enmark et al. 1997). Based on studies in which ERß expression was determined at the protein level by immunohistochemistry (IHC) (Jarvinen et al. 2000, Jensen et al. 2001, Mann et al. 2001, Omoto et al. 2002, Saunders et al. 2002) and ER was determined by IHC or by enzyme-linked immunoassay, an estimated frequency of ER and ERß status in breast tumors was obtained (Murphy et al. 2003). The most frequently occurring status is ER positive/ERß positive (59%) with similar frequencies of between 11 and 17% for the other three ER phenotypes. It is important to note that there are two groups of ERß-expressing breast tumors, those with coexpression of ER, and those that express ERß alone. The former occurs most frequently and, under experimental conditions when ER and ERß are coexpressed, they form heterodimers preferentially over homodimerization (Cowley et al. 1997). It is likely that homodimers of ER, homodimers of ERß and ER/ERß heterodimers will differentially affect gene expression. Further, transient coexpression of ER and ERß in cell lines results in ERß-induced reduction of ER activity at low ligand concentrations, as measured using ERE-regulated reporters (Hall & McDonnell 1999), suggesting that ERß can directly modulate ER activity in an inhibitory fashion. This may be a mechanism for differential oestrogen sensitivity and some reports have documented the differential expression of these two receptors under conditions of altered oestrogen sensitivity (Leygue et al. 1998a, Weihua et al. 2000). Interestingly, ERß expression is downregulated and ER expression upregulated during human breast tumourigenesis, suggesting that the ability of ERß to modulate ER may be altered during breast tumourigenesis (Leygue et al. 1998a, Roger et al. 2001). This correlates with altered oestrogen sensitivity and activity occurring during breast tumourigenesis (Murphy & Watson 2002). In addition, altered oestrogen action can occur during breast cancer progression and this is thought to underlie the development of tamoxifen resistance (Clarke et al. 2001). Increased expression and/or activity of ERß has been suggested as a mechanism of tamoxifen resistance, but the data are controversial (Paech et al. 1997, Speirs et al. 1999, Mann et al. 2001, Murphy et al. 2002).

To investigate further the effect of altered relative expression of ER and ERß on oestrogen and anti-oestrogen action in human breast cancer cells, we have stably expressed a tetracycline-inducible (Teton) human ERß1 in MCF7 human breast cancer cells, which endogenously express high ER but low ERß levels (Dotzlaw et al. 1997, Fuqua et al. 1999). The effect of altering the relative expression of these two ERs on oestrogen- and anti-oestrogen-regulated proliferation, as well as on endogenous markers of ER activity such as progesterone receptor (PR) (Osborne 1998), has been determined in this model system.

	Materials and methods

Top Abstract Introduction Materials and methods Results Discussion References

 
Plasmids

Epitope-tagged tetracycline-inducible human ERß1 (full-length) expression plasmids were generated. Human ERß1 long form of 530 amino acids (Leygue et al. 1998b, Ogawa et al. 1998b) was tagged at its N-terminus with a polyhistidine and an Xpress epitope tag using the pcDNA4/HisMax(A) plasmid (Invitrogen Canada Inc., Burlington, Ontario, Canada) and characterized as previously described (Peng et al. 2003). This cDNA was then subcloned downstream of the tetracycline-inducible promoter in the pTRE2 hyg plasmid (Gossen et al. 1995) (Clontech Laboratories Inc., Palo Alto, CA, USA). This construct contains the selectable hygromycin resistance gene on the same plasmid as the inducible gene.

Hormones and reagents

Oestradiol-17ß (E2) and the tetracycline analogue doxycycline were obtained from Sigma Chemical Co. (Oakville, Ontario, Canada). Hygromycin B was obtained from Clontech and G418 was from Invitrogen Canada Inc.

Cell culture and transfection

The cells were routinely cultured in Dulbecco’s modified Eagles’ medium (DMEM) containing 5% (v/v) foetal calf serum (FBS), 0.4% (w/v) glucose, glutamine and penicillin–streptomycin and 10 nM E2 (5% CM) as previously described (Peng et al. 2003). To obtain oestrogen-depleted cells, the culture medium of stock cells (approximately 50% confluent) was changed to phenol red-free DMEM supplemented with 5% (v/v) FBS that has been treated twice with charcoal/dextran, glucose, glutamine and penicillin-streptomycin (5% CS) and replaced every 2 days. The cells were grown for a least a week and not more than 2 weeks in this oestrogen-depleted medium, and were passaged once during this time. When oestrogen-depleted cells were around 50% confluent, they were harvested and used for experiments. For transfection, MCF7-clone 89 cells (Venditti et al. 2002) were grown in 10% CM in 100 mm dishes until 90% confluent. The cells were then split into three 100 mm dishes and the next day transfected (2–4 µg linearized plasmid per dish) using the Effectene transfection reagent according to the manufacturer’s instruction (QIAGEN, Mississauga, Ontario, Canada). Two days after transfection the medium was changed to 10% CM plus 500 µg/ml G418 and 400 µg/ml hygromycin, and the medium was changed every day for 4 days and then every 3 days. Approximately 2 weeks later, hygromycin resistance colonies began to appear and 120 large, healthy colonies were transferred to 96-well plates and screened after suitable numbers of cells were achieved as described below.

Generation of MCF7 human breast cancer cells with doxycycline-inducible epitope tagged human ERß1

The generation and characterization of MCF7 cells stably expressing reverse tetracycline transactivator ((rTA), clone 89 rTA) have been previously described (Venditti et al. 2002). These cells (clone 89) were transfected and selected as described above. Resistant clones were screened by Western blotting for doxycycline-inducible His-Xpress-ERß1 expression (tagged-ERß1), using anti-Xpress antibody (no. R910–25; Invitrogen Canada Inc.; 1/5000 in 20 mM Tris base, 0.137 M NaCl, pH 7.5 and 0.1% (v/v) Tween-20 (TBST)) as previously described (Peng et al. 2003). Multiple positive clones were isolated and two independent clones were selected for further characterization.

RNA extraction and RT-PCR conditions

Total RNA was extracted from frozen cell pellets using Trizol reagent (Life Technologies, New York, NY, USA) according to the manufacturer’s instructions. One microgram of total RNA was reverse transcribed, as previously described (Miksicek et al. 2002).

The primers used consisted of PR-U primer (5'-CACAAAACCTGACACCTCCAGTTC-3'; sense; located in PR 2290–2313) and PR-L primer (5'-GCA AAATACAGCATCTGCCCAC-3'; antisense; located in PR 2490–2511). Nucleotide positions given correspond to sequences of the human PR cDNA (accession number M15716 [GenBank] ). PCR amplifications were performed and PCR products analyzed as previously described (Miksicek et al. 2002). Each PCR consisted of 28 cycles (30 s at 94 °C, 30 s at 55 °C and 30 s at 72 °C). PCR products were then resolved by electrophoresis on 1.5% agarose gels. Following electrophoresis, the gels were stained with ethidium bromide, as previously described (Miksicek et al. 2002). Amplification of the ubiquitously expressed glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was performed in parallel and PCR products were visualized on agarose gels stained with ethidium bromide as previously described (Miksicek et al. 2002). Three independent PCRs for PR and GAPDH were performed for each experiment. Signals visualized by u.v. irradiation on a GelDoc 2000/ChemiDoc System (Bio-Rad, Mississauga, Ontario, Canada) were quantified by densitometry using the Quantity One software (version 4.2; Bio-Rad). PR was normalized to GAPDH in each experiment as previously described (Miksicek et al. 2002).

Whole cell ligand-binding assays

ER ligand-binding assays were done using the whole cell binding method as previously described (Coutts et al. 1996) following oestrogen depletion of the cells by two passages in 5% CS as previously described (Coutts et al. 1996).

Western blotting

Western blotting was carried out essentially as previously described (Adeyinka et al. 2002, Peng et al. 2003) with minor modifications. ER was detected using 1/1000 dilution of an ER mouse monoclonal antibody (NCL-ER-6F11; Novacastra Labs Ltd, Newcastle on Tyne, Tyne and Wear, UK) in TBST buffer, PR was detected using 1/1000 dilution of a PR mouse monoclonal antibody (NCL-PGR-AB; Novocastra Labs Ltd) in TBST and -tubulin was detected using 1/12 000 dilution of tubulin- antibody 2 (mouse monoclonal, clone DM1A; Neomarkers, Fremont, CA, USA). All membranes were incubated for 1 h at room temperature in 0.2% I-block (Tropix, Bedford, MA, USA) in Tris-buffered Saline (TBS). The blot was washed for 10 min at room temperature in TBST and then incubated with the appropriate dilution of primary antibody overnight at 4 °C. This incubation was followed by five 5-min washes in TBST, and then incubation with secondary antibody for 1 h at room temperature (goat anti-mouse IgG-HRP; Jackson Immuno Research Labs Inc., West Grove, PA, USA; 1/5000 in TBST). The blot was washed five times for 5 min with TBST and the signal was visualized using Supersignal West Dura Extended Duration Substrate kit (Pierce, Rockford, IL, USA) according to the manufacturer’s instructions. Signal detection and documentation used the ChemiDoc Gel Documentation System 2000 (Biorad Labs Canada, Mississauga, Ontario, Canada).

Immunofluorescence

Eleven days prior to slide preparation, stock cells were placed in medium containing 5% tetracycline-free FBS (BD Bioscience Clontech, Palo Alto, CA, USA), and the medium was changed every 3 days. These cells were then harvested and 1 x 10⁴ cells in 1 ml medium was added to each well of 4-chambered Falcon culture slides (Becton Dickinson, Franklin Lakes, NJ, USA). For every cell line used, one well was treated with vehicle alone and the duplicate well was treated with 1 µg/ml doxycycline for 48 h. Each chamber slide consisted of two wells of control MCF7-clone 89 cells and two wells of a tagged-ERß1 subclone of cells. Following incubation, the medium was removed by aspiration and the slides were washed with 1x phosphate-buffered saline (PBS) for 2 min. The PBS was removed and the cells were fixed and permeabilized with paraformaldehyde solution (3.7% formaldehyde and 0.18% Triton-X-100 in PBS) for 30 min at 37 °C. The slides were then washed in 1 x PBS for 5 min at room temperature, with agitation. Slides were then removed from the washing solution and, while still damp, they were coated with a hydrophobic barrier around the outside edge using an ImmEdge pen (Vector Laboratories Inc., Burlingame, CA, USA). Inside the barrier the cells were covered in 1 ml blocking solution (1% (w/v) BSA and 1% (v/v) normal goat serum in PBS) and incubated for 30 min at 37°C in a humidified chamber to prevent evaporation. The blocking solution was removed and the cells were incubated with mouse anti-Xpress antibody (1/750 in blocking solution; Invitrogen Canada Inc.) for 1 h at room temperature. The slides were then washed three times with PBS for 5 min with agitation. Slides were then moved to a dark room and then incubated with CY3-labelled goat anti-mouse secondary antibody (1/10 000 in blocking solution) for 1 h at room temperature. This was followed by three 5-min washes at room temperature with PBS with agitation. The cells were stained with Hoechst dye (no. 33258; Sigma Chemical Co., catalogue number B2883;1/10 000 in blocking solution) for 20 min at room temperature, followed by two washes in PBS for 6 min at room temperature. Slides were allowed to stand in PBS for a further 20 min at room temperature, then drained and one drop of Fluorsave reagent (Calbiochem, Hornby, Ontario, Canada) per well was applied. A coverslip was then added and sealed with clear nail polish. The slides were stored in the dark overnight at 4°C and fluorescent images obtained using a Nikon Eclipse E1000 microscope (Nikon, Mississauga, Ontario, Canada) using ACT-1 software (Nikon). All exposure times were for 1 s.

Half-life determination of tagged-ERß1

To determine the effect of various ligands on the half-life of tagged-ERß1, cells were set-up in 5% CS and left overnight. The media were changed and then the cells were treated for 48 h with 2 µg/ml doxycycline. The next day the cells were washed three times with 5% CS without doxycycline, and then treated with ethanol alone, E2 (10 nM), 4-hydroxy-tamoxifen (4-OH-tamoxifen; 500 nM) or ICI 182780 (500 nM; a gift kindly provided by AstraZeneca, Macclesfield, Cheshire, UK). Cells were harvested at 0, 2, 3, 4, 6, 8, 10, 12, 15 and 24 h after treatment. Cell extracts were analyzed by Western blot for tagged-ERß1 and tubulin expression. Tagged-ERß1 normalized to the tubulin signals from the same blot was quantified by densitometry using the Quantity One software (version 4.2; Bio-Rad). The half-life of tagged-ERß1 when bound to various selective ER modulators (SERMs) was estimated using the line of best fit on a semi-log plot from two independent experiments.

Cell proliferation assay

Cells depleted of oestrogen for at least a week by passage in 5% CS were set-up at 5 x 10⁴ cells per well in 6-well plates in 5% CS. The cells were left overnight and then treated with 2 µg/ml doxycycline or vehicle alone for 48 h. Cells maintained in the presence or absence of doxycycline then were either treated with E2 (10 nM) or vehicle alone. Triplicate wells were counted electronically (Beckman Coulter Canada Inc., Mississauga, Ontario, Canada) at the times indicated after the start of E2 treatment. Medium was changed every 2 days.

Cells grown in 5% CM were set-up at 10⁴ cells per well in 6-well plates in 5% CM, left overnight and then treated with 2 µg/ml doxycycline or vehicle alone for 48 h. Cells were maintained in the presence or absence of doxycycline and treated with varying concentrations of 4-OH-tamoxifen or vehicle alone. Triplicate wells were counted electronically (Beckman Coulter Canada Inc.) at the indicated times after the start of treatment. Medium was changed every 2 days. Doubling times were calculated from the initial (n_i) and final (n_t) cell numbers from the equation: doubling time (DT)=tlog2/log(n_i)/(n_t), where t is time in days between n_i and n_t. Proliferation rate=DT (control)/DT(treated) x 100%.

Statistical analysis

ANOVA followed by t-test analyses was used as appropriate.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

 
Identification of doxycycline-inducible tagged-ERß1-expressing MCF7 cell lines

Two clones, a low-expressor (ß-low) and a high expressor (ß-high) shown to express doxycycline-inducible tagged-ERß1 by Western blot analysis were chosen for further characterization. Controls were the original previously described MCF7 cell line stably expressing rTA (Venditti et al. 2002) and these MCF7 cells transfected with the empty pTRE2 hyg plasmid alone (V). As an additional control we included an MCF7 cell line stably transfected with a doxycycline-inducible GFP-tagged ERß expression vector (ß-high) (Htun et al. 1999, Zhao et al. 2002). Preliminary experiments showed that induction of the tagged-ERß1 was dependent on the dose of doxycycline used, and maximum induction in any one clone was achieved by treating the cells for 48 h with 1–2 µg/ml doxycycline. A Western blot of the cell lines grown in 5% CS+doxycycline treatment for 48 h is shown in Fig. 1A. A single band of approximately 62 kDa was detected only in the doxycycline-treated tagged-ERß1-transfected lines but not in control cell lines, MCF7, V and -high. The level of tagged-ERß1 expression in each of the clones was different with ß-high > ß-low. When these cell lines were grown on glass slides, fixed and the tagged-ERß1 visualized using anti-Xpress antibody, significant fluorescence was only found over the nucleus (determined by Hoechst staining) of doxycycline-treated tagged-ERß1 clones, with little, if any, signal detected in doxycycline-treated MCF7 cells. An example of the data obtained is shown in Fig. 2. These data were consistent with a general localization of the tagged-ERß1 product in or associated with the nucleus, although we cannot exclude localization of minor levels of tagged-ERß to other subcellular compartments such as the mitochondria (Yang et al. 2004).

View larger version (23K): [in this window] [in a new window]   Figure 1 (A) Western blot of cell extracts from cell lines grown in 5% CS medium treated (+) or not treated (–) with doxycycline (Dox) for 48 h. Tagged-ERß1 was visualized using anti-Xpress antibody as described in Materials and methods. MCF7 parent cells stably expressing rTA Tet-on transcription factor (Venditti et al. 2002); V=vector alone transfected MCF7 cells; ß-low=Xpress-tagged ERß1 low-expressing MCF7 clone; ß-high=Xpress-tagged ERß1 high-expressing MCF7 clone; -high=GFP-tagged ER high-expressing MCF7 clone. (B) Same blot as in (A), stripped and visualized using anti-ER antibody as described in Materials and methods (note the larger GFP–ER and the smaller approximately 66 kDa endogenously expressed ER). (C) Same blot as in (A), stripped and visualized using anti-tubulin antibody as described in Materials and methods.

View larger version (35K): [in this window] [in a new window]   Figure 2 Immunofluorescent detection of tagged-ERß1 in induced (+doxycycline (Dox)) and uninduced (–Dox) ß-low and control parental MCF7 cells. The magnification was x60 and the white bar in each panel represents 25 µm. Ab, antibody.

Effect of tagged-ERß1 expression on ER expression and E2 ligand-binding assays

Under conditions where the medium was depleted of oestrogen (5% CS as described in Materials and methods), little if any effect of tagged-ERß1 expression on endogenous ER expression was seen (Fig. 1B), and similar amounts of endogenous ER were expressed in control and tagged-ERß1 over-expressing cells. Protein loading was assessed by alpha-tubulin levels on the same blot (Fig. 1C). In order to determine the functional level of tagged-ERß1 over-expression achieved relative to the endogenous level of ER (primarily due to ER since only very low levels of ERß have been identified in these cells (Dotzlaw et al. 1997, Fuqua et al. 2003)) under these conditions, whole cell ligand-binding assays were carried out on cells grown in 5% CS with and without treatment for 48 h with 2 µg/ml doxycycline. The results of several independent experiments are presented in Table 1. A significant increase in specific [³H]E2 binding compared with the non-expressing controls (MCF7 and V) as well as the no doxycycline treatment controls was measured in all doxycycline-treated tagged-ERß1 (ß-high, ß-low) and the -high over-expressing cell lines. In addition, the levels of binding induced by doxycycline correlated with the level of tagged-ERß1 induced by doxycycline as determined by Western blot analysis, ß-high > ß-low (Fig. 1A). Doxycycline induction of tagged-ER in the -high cell line under these same conditions led to an increase in ligand binding similar to that obtained in ß-high tagged-ERß1 cells. These data were consistent with the increased E2 binding being due to the induced expression of tagged-ERß1 or tagged-ER after 48 h doxycycline treatment in 5% CS.

View larger version (28K): [in this window] [in a new window]   Figure 3 (A) Effect of oestrogen and anti-oestrogens on detection of tagged-ERß1 in MCF7 cells in oestrogen-depleted medium. Upper panel visualized with anti-Xpress antibody; lower panel visualized with anti-actin antibody. E2=treated with 10 nM E2; C=vehicle (ethanol)-treated control; T=treated with 500 nM 4-OH-tamoxifen; I=treated with 500 nM ICI 182780. ß-low=low ERß1 expressor+doxycycline (DOX) for 48 h followed by +ligands for 24 h. (B) Effect of E2 and anti-oestrogens on endogenous ER (top panel) and PR (middle panel) expression in the presence of over-expression of ERß1 (+DOX) or no over-expression of ERß1 (–DOX). Bottom panel visualized with anti-tubulin antibody. The same blot used for ER was stripped and visualized with anti-PR antibody and then anti-tubulin antibody as described in Materials and methods. The results shown in this figure are from ß-low treated (+DOX) or not treated (–DOX) for 48 h, followed by treatment+ligand for 24 h, as described above.

To determine if tagged-ERß1 expression affected known oestrogen-responsive endogenous targets, we investigated the level of expression of a known oestrogen-responsive gene product, PR. Under oestrogen-depleted conditions in MCF7 cells, the level of PR expression is markedly reduced and usually not detectable using our Western blot conditions. However, after doxycycline treatment, PR levels were significantly increased in all tagged-ERß1 over-expressing clones (Figs 3B and 4). Furthermore, the increase in level of PR expression was correlated to the level of transgene expression in the different clones, ß-high > ß-low (Fig. 4). When the cells were treated with E2 for 24 h following 48 h of doxycycline, there was a further induction of PR expression in all tagged-ERß1-expressing clones (Fig. 3B for ß-low, data not shown for ß-high) as well as the expected E2-induced PR expression in the non-tagged-ERß1-expressing MCF7 control cells (data not shown). Since PR is regarded as a gold standard downstream marker of ER activity we determined what effects upregulation of tagged-ER (-high) would have on PR expression. As shown in Fig. 4, increased tagged-ER (-high) also resulted in an induction of PR expression found under these conditions. A further increase in PR expression was seen in -high cells with E2 treatment as well (data not shown).

View larger version (21K): [in this window] [in a new window]   Figure 4 (A) Effect of tagged-ER expression (+ doxycycline (Dox) for 48 h) on endogenous PR expression in cells grown under oestrogen-depleted conditions. Western blot visualized with anti-PR antibody. MCF7 parent cells stably expressing rTA Tet-on transcription factor (Venditti et al. 2002); V=vector alone transfected MCF7 cells; ß-low=Xpress-tagged ERß1 low-expressing MCF7 clone; ß-high=tagged-ERß1 high-expressing MCF7 clone; -high=GFP-tagged ER high-expressing MCF7 clone. (B) Same blot as in (A), stripped and visualized with anti-tubulin antibody.

These data suggest that over-expression of ERß1 does not universally inhibit the activity of the endogenous ER, at least at the level of a known marker gene transcription in breast cancer cell lines.

Effect of tagged-ERß1 expression on proliferation in MCF7 human breast cancer cells

When the different cell lines were grown in complete medium containing 5% FBS, induction of tagged-ERß1 but not tagged-ER inhibited the growth of the cells, compared with MCF7, the control parent cell line. Both ß-high (DT+doxycycline, 1.84 ± 0.24 days, means ± S.E.M., n=6; DT – doxycycline, 1.23 ± 0.09 days, n=6) and ß-low (DT+doxycycline, 1.84 ± 0.30 days, n=6; DT – doxycycline, 1.34 ± 0.10 days, n=6) tagged-ERß1 over-expressors showed a significant increase in DT (change in DT shown as the ratio of DT+doxycycline/DT – doxycycline, Fig. 5) in the presence of doxycycline than without (ANOVA, P=0.002, n=6) compared with no significant alterations in the parent (MCF7: (DT+ doxycycline, 1.11 ± 0.13 days, n=6; DT – doxycycline, 1.11 ± 0.13 days) and -high cells (DT+doxycycline, 1.32 ± 0.10 days, n=6; DT – doxycycline, 1.29 ± 0.10 days). Therefore over-expression of ERß1 but not ER can decrease the growth of MCF7 breast cancer cells in culture.

View larger version (22K): [in this window] [in a new window]   Figure 5 Effect of tagged-ER induction on proliferation of cells in 5% CM medium. Uninduced cells were set-up at 104 cells per well in 5% CM, followed by +doxycycline (dox) (2 µg/ml). Media were changed every 2 days. Cells were harvested and counted electronically at various times after treatment (usually day 4 and day 8). DT were calculated as described in Materials and methods. The change in DT between no tagged-ER induction (–dox) and induction of tagged-ER (+dox) was calculated as a ratio. The means±S.E.M. of six independent experiments are shown. ANOVA (P=0.002) followed by t-testing indicates that the increase in DT seen in the ß-low over-expressor and the ß-high over-expressor is significantly different from that of the control parent cell line, MCF7 (P=0.022 and P=0.004 respectively); the over-expressing -high cell line is not significantly different from the control parent cell line, MCF7; the ß-low and the ß-high over-expressors are significantly different from -high (*P=0.03 and **P=0.005).

View larger version (28K): [in this window] [in a new window]   Figure 6 Effect of tagged-ER over-expression on growth of cells in response to oestrogen. Cells were depleted of oestrogen and set-up on day zero at 5x104 cells per well in 5% CS-containing medium, treated for 48 h with doxycycline and then treated with E2 (10 nM) or vehicle alone (ethanol). Cells were harvested and counted at the various times after E2 and medium was changed every 2 days. DT were calculated as described in Materials and methods. The % change in DT due to E2 treatment was calculated as DT-E2/DT-vehicle alone x100. The means±S.E.M. of n=12 (parent MCF7 cells, and ER over-expressing cells= -high) and n=13 (ERß1 low over-expressor=ß-low and ERß1 high over-expressor=ß-high) independent experiments is shown. Significant differences were found for all cell lines in the % decrease DT due to E2 treatment (ANOVA, P < 0.0001, followed by t-testing). In addition, differences between cell lines in % decreased DT due to E2 were found (ANOVA, P=0.012) where ß-low is significantly less than both MCF7 and ß-high (*P=0.033, P=0.049 respectively); similar comparisons for -high were **P=0.002, P=0.003. No significant differences were seen between MCF7 and -high. DT (means±S.E.M. days, n=13–12) in the presence of E2: MCF7, 1.75±0.1 days; -high, 1.62±0.1 days; ß-low, 1.83±0.12 days; ß-high, 2.04±0.12 days. DT (means+S.E.M. days, n=13–12) in the absence of E2: MCF7, 3.73±0.81 days; ß-high, 2.82±0.36 days; -low, 2.33±0.19 days; ß-high, 2.24±0.13 days.

An important hypothesis developed from basic studies on the function of ERß1 using transient transfection and various reporter constructs is that upregulated expression of ERß1 may result in tamoxifen resistance. Therefore in the inducible tagged-ERß1 over-expressing cells the responsiveness to 4-OH-tamoxifen-induced growth inhibition was determined. We determined the effect of various concentrations of 4-OH-tamoxifen on the proliferation rate of ß-high, ß-low, -high and MCF7 parent cell lines. Interestingly, with induction of tagged-ERs (+doxycycline) a biphasic dose–response was observed in the ß-high and ß-low cells, but not in -high or MCF7 control cells (Fig. 7A). At low 4-OH-tamoxifen concentrations (<50 nM), a significant stimulation of proliferation rate was observed (ANOVA, P < 0.0001, n=5). This increase in proliferation rate was higher in ß-high compared with ß-low cells (ANOVA, P < 0.0001, n=5). At higher concentrations (>50 nM), growth was inhibited in all cell lines (Fig. 7A). Comparing the growth inhibition phase, both ß-high and ß-low were significantly more sensitive to 4-OH-tamoxifen than either -high (ANOVA, P < 0.0001) or MCF7 parental cells (ANOVA, P < 0.0001). Both ß-high and ß-low dose–response curves were significantly to the left of both -high and MCF7 curves. However, ß-low was more sensitive than the ß-high so a dose-dependent effect of tagged-ERß1 expression was not observed for growth inhibition. The biphasic growth response to 4-OH-tamoxifen was also not found in MCF7 cells inducibly over-expressing the tagged-C-terminally truncated ERß variant, ERß2/cx (Ogawa et al. 1998a). In contrast to the tagged-ERß1, the tagged-ERß2/cx over-expressing MCF7 cells only showed a growth inhibition in response to 4-OH-tamoxifen (data not shown). Together, these data support the specificity of the biphasic growth response to tagged-ERß1 over-expression in multiple independent clones.

View larger version (10K): [in this window] [in a new window]   Figure 7 (A) Effect of tagged-ER over-expression (+doxycycline) on 4-OH-tamoxifen sensitivity in MCF7 breast cancer cells. Cells were grown in 5% CM and set-up on day zero at 104 cells per well in 5% CM-containing medium, treated for 48 h with doxycycline and then various concentrations of 4-OH-tamoxifen or vehicle alone. Cells were harvested and counted electronically at various times after anti-oestrogen treatment. DT and proliferation rates were calculated as described under Materials and methods, and the means±S.E.M. of n=5 (for 4-OH-tamoxifen concentrations 1–75 nM) and n=3 (for 4-OH-tamoxifen concentrations 100–1000 nM) independent experiments with results expressed as proliferation rate % of control (ethanol alone-treated cells). Medium was changed every 2 days. Solid squares are parental control cells, MCF7 clone 89 (rTA, Tet-on transcription factor expressor only). Open squares are -high over-expressing MCF7 cells. Solid circles are ß-low over-expressing MCF7 cells. Solid triangles are ß-high over-expressing MCF7 cells. (B) 4-OH-tamoxifen sensitivity in MCF 7 breast cancer cell lines in the absence of tagged ER over-expression (-doxycycline). Cells were grown as described above except that they were not treated with doxycycline. Cells were harvested and counted electronically at various times after anti-oestrogen treatment. DT and proliferation rates were calculated as described under Materials and methods, and the means±S.E.M. of n=3 independent experiments with results expressed as proliferation rate % of control (ethanol alone-treated cells). Medium was changed every 2 days. Solid squares are parental control cells, MCF7 (rTA, Tet-on transcription factor expressor only). Open squares are -high over-expressing MCF7 cells. Solid circles are ß-low over-expressing MCF7 cells. Solid triangles are ß-high over-expressing MCF7 cells.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

 
During human breast tumourigenesis and breast cancer progression a marked alteration of oestrogen action occurs. First, the responsiveness of ER+ breast tumour cells to the proliferative effect of oestrogen is direct (Lippman & Bolan 1975, Clarke et al. 1997), in contrast to normal breast epithelial cells where the growth effects of oestrogen appear to be indirect (Anderson et al. 1998). Secondly, the expression of ERs is markedly altered between normal breast tissue and breast tumours (Leygue et al. 1998a, Roger et al. 2001, Murphy & Watson 2002). ER is up-regulated in ER+ tumours relative to normal breast epithelium whereas current data suggest that ERß is downregulated in breast tumours compared with normal breast epithelium (Leygue et al. 1998a, Roger et al. 2001). Since there are both in vitro and in vivo studies (Hall & McDonnell 1999, Weihua et al. 2000, Forster et al. 2002, Peng et al. 2003) suggesting that ERß may be a negative modulator of ER action, it has been speculated that the downregulation of ERß during breast tumourigenesis removes a modulatory factor of ER that is important for maintaining normal sensitivity of breast epithelial cells to oestrogen action. However, despite the general decreased expression of ERß in ER+ breast tumours compared with normal tissue, many breast tumours continue to express ERß and the levels of expression can vary widely amongst tumours (Dotzlaw et al. 1997, Jarvinen et al. 2000). And although still controversial, often when ERß protein is determined, its expression can be correlated with good prognostic markers such as ER and PR expression (Jarvinen et al. 2000, Murphy et al. 2002, Saunders et al. 2002, Skliris et al. 2003). This is also consistent with data suggesting that higher ERß expression is associated with sensitivity to tamoxifen therapy rather than resistance to tamoxifen therapy. However, other data have suggested ERß association with anti-oestrogen resistance and poor prognostic markers (Speirs et al. 1999, Speirs 2002). The development of an inducible tagged-ERß1-expressing MCF7 breast cancer cell line has allowed us to address the functional outcome of altering the relative expression of ERß1 to ER in the ‘gold standard’ model of oestrogen/anti-oestrogen-responsive human breast cancer. The development of such breast cancer cell line models is highly relevant to human breast cancer in vivo, since the majority of human breast tumours (~59%) have been shown to co-express both ER and ERß1 (Murphy et al. 2003).

Consistent with studies in which stable over-expression of ERß1 alone in cell lines was investigated, a predominantly nuclear location of the transfected derivative ERß1 (in our case epitope tagged) was observed (Lazennec et al. 2001). We have isolated various clones of the tetracycline-inducible tagged-ERß1 that under maximal induction conditions express low and high levels of the tagged-ERß1 protein. Using the ligand-binding assay, we have determined that in low ERß clones the tetracy-cline analogue doxycycline increased specific binding of radiolabelled E2 around 200% that of the non-induced (no doxycycline treatment) and/or non-expressing control cells, while doxycycline treatment of the high ERß1-expressing clones resulted in around a 300% increase in specific E2 binding over controls. Since parental MCF 7 cells contain predominantly ER, the control and uninduced levels of ligand binding reflect predominantly endogenous ER levels. After doxycycline treatment, ERß was induced to less than or equimolar levels (estimating from ligand-binding assays) to the endogenous ER in the ß-low clones, suggesting that the majority of the ERß would be in the form of heterodimers with ER (Cowley et al. 1997) in these clones. However, this depends on the kinetics of dimer formation in vivo and hence some homodimers of ER may also exist. In ß-high, the tagged-ERß1 was induced to higher levels than ß-low clones, suggesting that all the ER molecules may be heterodimerized with ERß1, with the possibility that a small population of ERß1 homodimers may also exist. Although in vitro ER–ERß1 heterodimer is apparently favoured (Cowley et al. 1997) it is unclear if this is the case in vivo. However, recently it was demonstrated using Fluorescence resonance energy transfer (FRET) analysis in living cells that ER homodimers of each receptor and ER–ERß1 heterodimers can occur in vivo independently of the presence of ligand (Bai & Giguere 2003), supporting the presence in our cell line models of ERß1 predominantly in the form of ER–ERß1 heterodimers. Little, if anything, however is known about the differential function of the various homo- and heterodimers of the ER isoforms, although the models described in this study may be useful in addressing such questions.

ERß1 over-expression alone in the absence of ER-negative MDAMB231 human breast cancer cells has resulted in variable results with both growth inhibition (Lazennec et al. 2001) and growth stimulation being reported (Tonetti et al. 2003). However, when over-expression of ERß1 in an ER-positive background has been achieved, more consistent results have been obtained, and in particular over-expressed ERß1 results in growth inhibition (Paruthiyil et al. 2004, Strom et al. 2004). Our results are consistent with these latter data in that ERß1 but not ERß over-expression results in reduced growth and increased DT when cells are grown in complete medium containing FBS and oestrogen. In addition, over-expression of ERß1 but not ER reduced the magnitude of oestrogen-induced growth when cells are grown in oestrogen-depleted medium. These effects of ERß1 on the proliferation of ER-expressing breast cancer cells are dose-dependent since smaller effects were seen in the low ERß1-expressing cells compared with the high ERß1-expressing cells. These data are consistent with the hypothesis that ERß1 can negatively modulate the activity of ER when the two ERs are co-expressed. They also suggest that the downregulation of ERß1in breast tumours compared with normal breast tissue which altered the relative expression of the two ERs is functionally involved in breast tumourigenesis at least for ER+ breast cancer.

However, negative modulation of ERactivity is not a universal effect of over-expression of ERß1 since, in contrast to the effect on E2-induced proliferation, ERß1 over-expression significantly increased endogenous PR expression, previously considered a specific downstream marker of ER activity in breast cancer cells. Again, in contrast to the growth effects, the increase in PR expression is not ERß1 specific since over-expression of ER as seen in -high cells similarly increased PR expression. These results support a mechanistic connection associated with the observation in vivo where increased ERß expression determined at the protein level in human breast tumours is often correlated with PR status and expression (Jarvinen et al. 2000, Mann et al. 2001, Murphy et al. 2002). Therefore ERß1 regulation of PR expression has in vivo physiological and/or pathophysiological relevance. Interestingly, the effect of ER over-expression on PR expression occurs apparently without added ligand (E2) suggesting ligand independence, although further increases in PR expression were seen when the over-expressed ER was bound to E2 (ligand dependent). Indeed, since the majority of human breast tumors arise in older, and therefore likely postmenopausal women, one could speculate that PR expression in many breast tumors may reflect the ligand-independent activity of the combination of ER and ERß1 expressed in the tumour. Alternatively, oestrogen generated due to aromatase expression by tumor cells as well as the surrounding stroma would also contribute to the steady-state PR expression in breast tumours. However, we cannot exclude the possibility that trace amounts of oestrogens are still present in the medium and contribute to the increased PR expression. The molecular mechanism by which ERß1 may regulate PR remains to be determined, although it is due, at least in part, to increased transcription since steady-state PR RNA levels were also increased. However, since ERE half-sites adjacent to both AP-1 and Sp1 sites have been previously implicated in the mechanism by which ER regulates PR expression, it is speculated that these may also be involved in the mechanism by which ERß1 can regulated PR expression. However, if the AP-1 site is important in this regard, this would be an example of E2 activation of ERß1 being involved in increased rather than decreased transcription of an AP-1-regulated oestrogen responsive promoter. We cannot however exclude effects on PR mRNA stability at this stage.

The involvement of ERß1 in tamoxifen sensitivity is controversial, and in a previous study when inducible over-expression of ERß1 was investigated after stable expression in T-47D breast cancer cells, which also express endogenous ER, only growth inhibition due to tamoxifen was observed (Strom et al. 2004). No stimulation of growth was reported. However, full dose–response analyses covering low and high concentrations of 4-OH-tamoxifen were not undertaken (Strom et al. 2004). To the best of our knowledge, our data are the first to show any effect of altered ERß1 expression on tamoxifen-regulated proliferation in breast cancer cells. In the presence of oestrogen-replete (phenol-red and FBS-containing) medium, a biphasic growth response to 4-OH-tamoxifen was observed only in tagged-ERß1 over-expressing cells. Neither the parental control nor the tagged-ER over-expressing cells showed a biphasic response. The biphasic growth response was lost when tagged-ERß1 over-expression was not induced (no doxycycline), and was not present when the tagged-ERß variant, ERß2-cx, was over-expressed under the same conditions. The growth stimulatory response seen at low doses of 4-OH-tamoxifen was only seen when tagged-ERß1 was over-expressed, its magnitude was correlated with the level of ERß1 and was completely eliminated in the absence of ERß1 induction (i.e. no doxycycline). These data suggest that tagged-ERß1 expression was functionally involved in the 4-OH-tamoxifen stimulation of proliferation seen at low doses of the anti-oestrogen. When considering the growth inhibition phase, doxycycline treatment of all cell lines except -high significantly moved the dose–response curves to the left, compared with the no doxycycline treatment, i.e. no induction of expression of the tagged-ER. This suggested that a higher ERß1 but not ER increased sensitivity to the growth-inhibitory effects of 4-OH-tamoxifen. However, removal of doxycycline also reduced the sensitivity of the MCF7 control cells to 4-OH-tamoxifen, suggesting that clonal variation may also in part contribute to, but not entirely explain, the differences seen in sensitivity to the growth-inhibitory, but not the growth-stimulatory effects of tamoxifen in these cell lines. In particular, there was no evidence that ERß1 over-expression resulted in a shift of the dose–response curve to the right of any of the control cells, supporting the conclusion that acute over-expression of ERß1 is not associated with resistance or decreased sensitivity to tamoxifen. Overall, these data support the hypothesis that increased expression of ERß1 but not over-expression of ER in the presence of endogenous expression of ER increases the sensitivity of human breast cancer cells to the growth-modulatory effects of tamoxifen.

Our results showing that ERß1 over-expressing cells display an increased sensitivity to tamoxifen are consistent with increased expression being associated with tamoxifen sensitivity and not resistance in vivo in breast cancer patients (Mann et al. 2001, Murphy et al. 2002, Esslimani-Sahla et al. 2004, Myers et al. 2004). The stimulation of proliferation due to 4-OH-tamoxifen seen at low doses is interesting and we would argue that rather than be interpreted as representing a potential form of tamoxifen resistance (Clarke et al. 2003), since it only occurs at very low concentrations (tamoxifen resistance occurs in vivo at varying periods of time following or during high doses of tamoxifen), it is likely to resemble and/or be a model of ‘tamoxifen flare’ that has been observed clinically (Plotkin et al. 1978, Mortimer et al. 2001, Biersack et al. 2004). In this condition, an apparent worsening of the disease is seen in some patients in the first few weeks of starting tamoxifen treatment; likely before sufficiently high steady-state concentrations of tamoxifen and its active metabolites have accumulated. This is often a favourable sign for responsiveness to tamoxifen therapy since, in the cases where tamoxifen therapy was continued, subsequent tumour regression often occurred (Plotkin et al. 1978, Vogel et al. 1995). Interestingly, only one other breast cancer cell culture model of ‘tamoxifen flare’ has been documented previously. In this case the cell line was T-47D and several laboratories have often described a relatively higher natural level of ERß expression in T47-D cells compared with other ER+ breast cancer cell lines. The mechanism by which ERß1 may increase the sensitivity of breast cancer cells to tamoxifen is unknown, but the higher relative binding affinity of ERß1 compared with ER for 4-OH-tamoxifen may contribute in part (Kuiper et al. 1997).

In conclusion, our data using inducible, ERß1 over-expressing MCF7 cells do not support the hypothesis that up-regulation of ERß1 in the presence of ER is a mechanism of anti-oestrogen resistance per se. However, we cannot exclude a role of over-expression of ERß in resistance in combination with other changes, such as altered cofactor expression and/or activity. In addition, our data support the hypothesis that ERß1 can be a negative modulator of ER-mediated growth response since over-expression of ERß in a dose-dependent fashion inhibits the magnitude of the E2-induced growth response in ER+ human breast cancer cells. However, over-expression of ERß1 does not inhibit or reduce all ER-regulated activities since endogenous PR expression is not decreased in these cells. Indeed, ERß1 upregulates the endogenous expression of PR in possibly both a ligand-independent and-dependent fashion. It is acknowledged that due to space limitations our interpretation of the data shown here is likely too narrow since it underestimates the complexity of the variables (only one of which is relative expression of ER isoforms) involved in the response to E2 as previously demonstrated (Hewitt et al. 2003, Stossi et al. 2004). The parallels seen with our over-expressing ERß1 MCF7 cell lines and some in vivo studies showing a relationship between PR and ERß expression as well as ERß expression and tamoxifen sensitivity of breast cancer patients suggest that these models are relevant and useful for dissecting the role of ERß1 expression in ER+ breast cancer.

	Acknowledgements

 
We thank Christina Walker for her technical assistance. This work was supported by the Canadian Breast Cancer Research Alliance and the USA Medical Research and Materiel Command. A L holds a CancerCare Manitoba Foundation/Guardian Angel Studentship Award. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.

	References

Top Abstract Introduction Materials and methods Results Discussion References

 
Adeyinka A, Niu Y, Cherlet T, Snell L, Watson P & Murphy L 2002 Activated mitogen-activated protein kinase expression during human breast tumorigenesis and breast cancer rogression. Clinical Cancer Research 8 1747–1753.[Abstract/Free Full Text]

Anderson E, Clarke R & Howell A 1998 Estrogen responsiveness and control of normal human breast proliferation. Journal of Mammary Gland Biology and Neoplasia 3 23–35.[CrossRef][ISI][Medline]

Aronica S & Katzenellenbogen BS 1991 Progesterone receptor regulation in uterine cells: stimulation by estrogen, cyclic adenosine 3',5'-monophosphate, and insulin-like growth factor I and suppression by antiestrogens and protein kinase inhibitors. Endocrinology 128 2045–2052.[Abstract]

Bai Y & Giguere V 2003 Isoform-selective interactions between estrogen receptors and steroid receptor coactivators promoted by estradiol and erbB-2 signaling in living cells. Molecular Endocrinology 17 589–599.[Abstract/Free Full Text]

Biersack H, Bender H & Palmedo H 2004 FDG-PET in monitoring therapy of breast cancer. European Journal of Nuclear Medicine Molecular Imaging. Apr 27 [Epub ahead of print]].

Clarke R, Howell A, Potten C & Anderson E 1997 Dissociation between steroid receptor expression and cell proliferation in the human breast. Cancer Research 57 4987–4991.[Abstract/Free Full Text]

Clarke R, Leonessa F, Welch J & Skaar T 2001 Cellular and molecular pharmacology of antiestrogen action and resistance. Pharmacological Reviews 53 25–71.[Abstract/Free Full Text]

Clarke R, Liu M, Bouker K, Gu Z, Lee R, Zhu Y, Skaar T, Gomez B, O’Brien K, Wang Y et al. 2003 Antiestrogen resistance in breast cancer and the role of estrogen receptor signaling. Oncogene 22 7316–7339.[CrossRef][ISI][Medline]

Clemons M & Goss P 2001 Estrogen and the risk of breast cancer. New England Journal of Medicine 344 276–285.[Free Full Text]

Colditz G 1998 Relationship between estrogen levels, use of hormone replacement therapy, and breast cancer. Journal of the National Cancer Institute 90 814–823.[Abstract/Free Full Text]

Coutts A, Davie J, Dotzlaw H & Murphy L 1996 Estrogen regulation of nuclear matrix-intermediate filament proteins in human breast cancer cells. Journal of Cellular Biochemistry 63 174–184.[CrossRef][ISI][Medline]

Cowley S, Hoare S, Mosselman S & Parker M 1997 Estrogen receptors alpha and beta form heterodimers on DNA. Journal of Biological Chemistry 272 19858–19862.[Abstract/Free Full Text]

Dotzlaw H, Leygue E, Watson P & Murphy L 1997 Expression of estrogen receptor-beta in human breast tumors. Journal of Clinical Endocrinology and Metabolism 82 2371–2374.[Abstract/Free Full Text]

Enmark E, Pelto-Huikko M, Grandien K, Lagercrantz S, Lagercrantz J, Fried G, Nordenskjold M & Gustafsson J-A 1997 Human estrogen receptor ß gene structure, chromosomal localization and expression pattern. Journal of Clinical Endocrinology and Metabolism 82 4258–4265.[Abstract/Free Full Text]

Esslimani-Sahla M, Simony-Lafontaine J, Kramar A, Lavaill R, Mollevi C, Warner M, Gustafsson J-A & Rochefort H 2004 Estrogen receptor beta (ERbeta) level but not its ERbeta-cx variant helps to predict tamoxifen resistance in breast cancer. Clinical Cancer Research 10 5769–5776.[Abstract/Free Full Text]

Forster C, Makela S, Warri A, Kietz S, Becker D, Hultenby K, Warner M & Gustafsson J 2002 Involvement of estrogen receptor ß in terminal differentiation of mammary gland epithelium. PNAS 99 15578–15583.[Abstract/Free Full Text]

Fuqua S, Schiff R, Parra I, Friedrichs W, Su J, McKee D, Slentz-Kesler K, Moore L, Willson T & Moore J 1999 Expression of wild-type estrogen receptor beta and variant isoforms in human breast cancer. Cancer Research 59 5425–5428.[Abstract/Free Full Text]

Fuqua S, Schiff R, Parra I, Moore J, Mohsin S, Osborne C, Clark G & Allred D 2003 Estrogen receptor beta protein in human breast cancer: correlation with clinical tumor parameters. Cancer Research 63 2434–2439.[Abstract/Free Full Text]

Gossen M, Freundlieb S, Bender G, Muller G, Hillen W & Bujard H 1995 Transcriptional activation by tetracyclines in mammalian cells. Science 268 1766–1769.[Abstract/Free Full Text]

Hall J & McDonnell D 1999. The estrogen receptor beta-isoform (ERbeta) of the human estrogen receptor modulates ERalpha transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology 140 5566–5578.[Abstract/Free Full Text]

Hewitt S, Deroo B, Hansen K, Collins J, Grissom S, Afshari C & Korach K 2003 Estrogen receptor-dependent genomic responses in the uterus mirror the biphasic physiological response to estrogen. Molecular Endocrinology 17 2070–2083.[Abstract/Free Full Text]

Htun H, Holth L, Walker D, Davie J & Hager G 1999 Direct visualization of the human estrogen receptor alpha reveals a role for ligand in the nuclear distribution of the receptor. Molecular Biology of the Cell 10 471–486.[Abstract/Free Full Text]

Jarvinen T, Pelto-Huikko M, Holli K & Isola J 2000 Estrogen receptor beta is coexpressed with ERalpha and PR and associated with nodal status, grade, and proliferation rate in breast cancer. American Journal of Pathology 156 29–35.[Abstract/Free Full Text]

Jensen E, Cheng G, Palmieri C, Saji S, Makela S, Noorden Sv, Wahlstrom T, Warner M, Coombes R & Gustafsson J-A 2001 Estrogen receptors and proliferation markers in primary and recurrent breast cancer. PNAS 98 15197–15202.[Abstract/Free Full Text]

Kuiper G, Enmark E, Pelto-Huikko M, Nilsson S & Gustafsson J-A 1996 Cloning of a novel estrogen receptor expressed in rat prostate and ovary. PNAS 93 5925–5930.[Abstract/Free Full Text]

Kuiper G, Carlsson B, Grandien K, Enmark E, Haggblad J, Nilsson S & Gustafsson J-A 1997 Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 138 863–870.[Abstract/Free Full Text]

Lazennec G, Bresson D, Lucas A, Chauveau C & Vignon F 2001 ERß inhibits proliferation and invasion of breast cancer cells. Endocrinology 142 4120–4130.[Abstract/Free Full Text]

Leygue E, Dotzlaw H, Watson P & Murphy L 1998a Altered estrogen receptor alpha and beta mRNA expression during human breast tumorigenesis. Cancer Research 58 3197–3201.[Abstract/Free Full Text]

Leygue E, Dotzlaw H, Lu B, Glor C, Watson P & Murphy L 1998b Estrogen receptor-beta: mine is longer than yours? Journal of Clinical Endocrinology and Metabolism 83 3754–3755.[Free Full Text]

Lippman M & Bolan G 1975 Oestrogen responsive human breast cancer in long term tissue culture. Nature 256 592–593.[CrossRef][Medline]

McDonnell D & Norris J 2002 Connections and regulation of the human estrogen receptor. Science 296 1642–1644.[Abstract/Free Full Text]

Mann S, Laucirica R, Carlson N, Younes P, Ali N, Younes A, Li Y & Younes M 2001 estrogen receptor beta expression in invasive breast cancer. Human Pathology 32 113–118.[CrossRef][ISI][Medline]

Miksicek R, Myal Y, Watson P, Walker C, Murphy L & Leygue E 2002 Identification of a novel mucin-like gene strongly expressed in normal and tumor human mammary epithelium. Cancer Research 62 2736–2740.[Abstract/Free Full Text]

Mortimer J, Dehdashti F, Siegel B, Trinkaus K, Katzenellenbogen J & Welch M 2001 Metabolic flare: indicator of hormone responsiveness in advanced breast cancer. Journal of Clinical Oncology 19 2797–2803.[Abstract/Free Full Text]

Murphy L & Watson P 2002 Steroid receptors in human breast tumorigenesis and breast cancer progression. Biomedical Pharmacotherapy 56 65–77.

Murphy L, Leygue E, Niu Y, Snell L, Ho S-M & Watson P 2002 Relationship of