7490 Vol. 10, 7490–7499, November 15, 2004 Clinical Cancer Research Featured Article Low Levels of Estrogen Receptor ␤ Protein Predict Resistance to Tamoxifen Therapy in Breast Cancer Torsten A. Hopp,1,3 Heidi L. Weiss,1,3 improved disease-free and overall survival in patients Irma S. Parra,3 Yukun Cui,1,3 treated with adjuvant tamoxifen therapy. 1,2,3 Conclusions: These findings provide evidence that C. Kent Osborne, and ␤ 1,2,3 ER- may be an independent predictor of response to ta- Suzanne A. W. Fuqua moxifen in breast cancer. Furthermore, these results suggest Departments of 1Medicine, 2Molecular and Cellular Biology, and ␤ 3 that ER- may influence tumor progression in ways differ- Breast Center, Baylor College of Medicine and the Methodist ent from those mediated by the ER-␣ isoform. Hospital, Houston, Texas INTRODUCTION ABSTRACT For more than 30 years, the classical estrogen receptor, Purpose: Breast cancer is a hormone-dependent cancer, called ER-␣, has been extensively studied as a prognostic and and the presence of estrogen receptor ␣ (ER-␣) in tumors is predictive marker in clinical breast cancer, making this nuclear used clinically to predict the likelihood of response to hor- receptor the most valuable target for the treatment of human monal therapies. The clinical value of the second recently breast cancer with selective estrogen receptor modulators or the identified ER isoform, called ER-␤, is less clear, and there is newer generation aromatase inhibitors. Patients with ER-␣– currently conflicting data concerning its potential role as a positive tumors have a significantly prolonged overall and re- prognostic or predictive factor. currence-free survival with selective estrogen receptor modula- Experimental Design: To assess whether ER-␤ expres- tors (1) and aromatase inhibitor therapy (2). Our understanding sion is associated with clinical outcome, protein levels were of ER in breast cancer became less clear with the identification measured by immunoblot analysis of a retrospective bank of of a second related ER isoform called ER-␤ (3, 4). After the tumor cell lysates from 305 axillary node-positive patients. A identification of ER-␤, researchers had to reevaluate the prior total of 119 received no adjuvant therapy, and 186 were simplistic model of estrogen action. However, we are only now treated with tamoxifen only. The median follow-up time was beginning to appreciate whether ER-␤ expression in tumors 65 months. Univariate and multivariate Cox regression exerts a clinical impact on the progression and treatment of modeling was done to assess the prognostic and predictive breast cancer. significance of ER-␤ expression. ER-␣ and ER-␤ are both ligand-induced transcription fac- Results: Expression of ER-␤ protein did not correlate tors that can modulate the expression of specific target genes. At significantly with any other clinical variables, including ER present, there is little information available concerning differ- and progesterone levels (as measured ligand binding assay), ential induction of gene expression from either ER-␣ or ER-␤. tumor size, age, or axillary nodal status. In the untreated On the structural level, both receptor isoforms encode two population, those patients whose tumors who expressed both activation functions, as well as a DNA-binding domain, which receptor isoforms exhibited the most favorable outcome as recognizes and binds to estrogen response elements within the compared with those patients who had lost ER-␣ expression. promoter of target genes (5). The two receptor isoforms also However, there was no association between ER-␤ levels share Ͼ50% similarity in their hormone-binding domains and a alone and either disease-free or overall survival in the un- 95% similarity within their DNA binding domains (4). Because treated patient population. In contrast, in both univariate there is less sequence similarity within their hormone-indepen- and multivariate analyses, high levels of ER-␤ predicted an dent activation function-1 (AF-1) domains, it has been sug- gested that the two receptors might perform distinct functions. In agreement with this possibility is a body of literature showing that ER-␣ is the dominant receptor in the mouse mammary Received 6/8/04; revised 7/28/04; accepted 8/9/04. gland essential for ductal development and hormone response Grant support: NIH/National Cancer Institute Grants RO1 CA58183 (6–8). We also know that ER-␤ binds estrogen with similar (S. Fuqua) and P01 CA30195 (K. Osborne) and Department of Defense affinity as ER-␣, but unlike ER-␣, antiestrogen-occupied ER-␤ Fellowship Grant DAMD17-02-1-0278. (Y. Cui). can activate transcription via nonclassical ER-signaling path- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ways through its binding to activator protein 1 transcription advertisement in accordance with 18 U.S.C. Section 1734 solely to factors (9). This has led some investigators to speculate that indicate this fact. ER-␤ could play a role in tamoxifen resistance through the Requests for reprints: Suzanne A. W. Fuqua, Breast Center, Baylor agonist activity of tamoxifen, and indeed, one study examining College of Medicine, One Baylor Plaza, BCM 600, Houston, TX 77030. ␤ Phone: (713) 798-1672; Fax: (713) 798-1673; E-mail: sfuqua@breastcenter. ER- RNA expression in a small number of tamoxifen-treated tmc.edu. patients supported this idea (10). Although the two receptors do ©2004 American Association for Cancer Research. not appear to be expressed within the same cell in normal mouse Downloaded from clincancerres.aacrjournals.org on September 23, 2021. © 2004 American Association for Cancer Research. Clinical Cancer Research 7491 mammary tissue (8), we and others have shown that they are washing three times in TBST, and then incubated for 1 hour frequently coexpressed in breast tumors (11, 12). Because the with horseradish peroxidase-labeled antimouse IgG (1:2000; two receptors can exist as heterodimers when coexpressed (13, Amersham Pharmacia Biotech, Piscataway, NJ). After extensive 14), ER-␤ might also modify the activity of ER-␣ and therefore washing in TBST, ER-␤ protein was then visualized on a the clinical outcome of patients given hormonal therapy. FluorChem digital imaging system (␣ Innotech, San Leandro, A growing number of studies have analyzed ER-␤ levels in CA) with an enhanced chemiluminescence detection system. clinical breast tumor samples (11, 12, 15–18), but results have Band intensities were measured densitometrically with the been contradictory. Therefore, currently there is no consensus AlphaEaseFC software (␣ Innotech), and then ER-␤ levels in concerning the role that ER-␤ may play as either a prognostic tumors were normalized to ER-␤ levels in the MCF-7–positive factor in patients not receiving systemic adjuvant treatment or as control lysate (10 ␮g) from the same immunoblot. a predictive factor for response to treatment. We therefore Other Biological Factors. Total ER and PR levels were performed immunoblot analysis with an ER-␤–specific antibody measured by ligand-binding assay as described elsewhere (23). in 305 patients with axillary node-positive tumors with long- Briefly, cytosolic proteins were extracted from tumor tissues term clinical follow-up. For patients who received tamoxifen that had been pulverized in liquid nitrogen. Iodine-125–labeled therapy, high ER-␤ levels predicted a favorable outcome. We estradiol and tritiated-ORG 2058 (Amersham Pharmacia Bio- conclude that as with ER-␣, ER-␤ is a biomarker of response to tech) addition allowed for the simultaneous determination of selective estrogen receptor modulators. both ER and progesterone receptor (PgR) levels in a standard multipoint dextran-coated charcoal assay. Tumors with an ER content of at least 3 fmol/mg protein and with a PgR content of MATERIALS AND METHODS at least 10 fmol/mg protein were considered positive for ER and Tumor Specimens. A cohort of frozen primary breast PgR, respectively. These levels were based on prior studies tumor specimens from 305 axillary lymph node-positive pa- calibrated to clinical outcome (23). tients was selected from the tumor bank of The Breast Center, The two separate PgR isoforms, PgR-A and PgR-B, were Baylor College of Medicine, for use in the immunoblot study. previously determined with immunoblot analysis after separa- This study was approved by the Baylor College of Medicine tion by 8% SDS-PAGE as described, with the mouse anti- Institutional Review Board. This patient cohort has previously PgR1294 (Dako, Carpinteria, CA), which specifically recog- been evaluated for expression of other clinical variables (19, nizes both PgR-A and PgR-B isoforms on immunoblots (20). 20). Proteins were extracted from the tumors as described pre- AIB1 and HER-2 levels in these same tumors were previously viously (21). Briefly, 30 mg of pulverized tumor powder were determined with immunoblot analysis after separation by 8 and solubilized in 300 ␮L of 5% SDS at 90°C for 5 minutes and 7.5% SDS-PAGE, respectively, and then staining with two centrifuged at 13,000 ϫ g for 5 minutes. Protein concentrations rabbit polyclonal antisera, anti-RAC3 antibody (19), and anti- of the supernatants were determined using the bicinchoninic HER2 (21). The MCF-7 reference cell line standard was used acid method (Pierce, Rockford, IL). Typical protein yields were for normalization of AIB1 levels, and a similarly prepared 2to5␮g/␮L. Supernatants were then stored at Ϫ70°C until use extract from T47D human breast cancer cells was used for in the immunoblot analysis. normalization of the PgR isoforms and for HER-2 levels. S- Immunoblot Analysis. We have previously shown that Phase fraction and DNA ploidy were calculated with DNA flow MCF-7 human breast cancer cells express full-length ER-␤ cytometry (24) and is reported as low, intermediate, or high. protein corresponding to the 530 amino acid isoform (22). Thus, Briefly, DNA flow cytometry was done on tumor extracts, and MCF-7 extracts (also stored at Ϫ70°C) were used as a standard the histograms were analyzed by Modfit (Verity Software on each gel to correct for gel to gel variations.