- Email: [email protected]
Immunochemical analyses of estrogen receptors in human breast tumors by a novel monoclonal estrogen receptor antibody (EVG F9)
Steroids 65 (2000) 429 – 436
Immunochemical analyses of estrogen receptors in human breast tumors by a novel monoclonal estrogen receptor antibody (EVG F9) Natalia S. Rost, Kristine Murphy, Laurie Hafer, Matthew Pavao, Abdulmaged M. Traish* Department of Biochemistry, Boston University School of Medicine, Center for Advanced Biomedical Research, Boston, MA 02118, USA Received 14 September 1999; received in revised form 21 February 2000; accepted 1 March 2000
Abstract The objective of this study was to assess the potential utility of a new site-directed, monoclonal anti-estrogen receptor antibody (EVG F9) in detection and analyses of human breast tumor estrogen receptor (ER␣), using immunoblotting and immunohistochemical assays. Using Western Blot analyses, we demonstrated that EVG F9 monoclonal antibody binds specifically to ER␣ and does not cross-react with ER. Furthermore, binding of EVG F9 to ER␣ was effectively displaced with the immunogenic peptide in Western Blots and in immunohistochemical analyses. In Western Blot analyses, EVG F9 detected ER␣ at low concentrations approaching 5 to 10 fmol/sample. Determination of ER␣ status of a series of human breast tumor samples by Western Blot analyses or immunohistochemistry using EVG F9 correlated well with ER␣ values measured by ligand binding assays. These observations suggest that EVG F9 monoclonal anti-ER␣ antibody is a valuable immunochemical tool for detection and analyses of ER␣ in human breast tumors. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Breast cancer; Estrogen receptor; Monoclonal antibodies
1. Introduction Breast cancer is the second leading cause of cancer death among women in the United States. Since the 1980s, the incidence of breast cancer has been increasing with the highest incidence occurring in postmenopausal women [1,2]. It is well documented that ovarian hormones play a major role in development and progression of breast cancer. Tumor expression of estrogen (ER) and progesterone (PR) receptors is associated with favorable response to endocrine therapy [3– 6], extended 5-year survival [7–12], and characterized by a more histologically differentiated tumor phenotype [13–16]. The expression of ER and PR varies with tumor phenotype and with age and may represent the marked heterogeneity of breast cancer [17,18]. ER␣ status is currently used as a This work was supported by a grant from the Department of the Army DAMD 1794J4468. * Corresponding author. Tel.: ⫹-617-638-4578; fax: ⫹-617-638-5412. E-mail address: [email protected] (A.M. Traish).
prognostic factor in management of human breast cancer [3–15]. Both ligand-binding studies and immunohistochemical analyses are currently utilized to determine ER and PR status in human breast tumor tissue specimens [18 –20]. The advent of mammography and other sophisticated imaging techniques has dramatically improved early detection of breast tumors. These advances have resulted in surgical removal of tumors of small size, precluding analyses of the ER␣ by ligand-binding assay, which requires a minimum of 200 to 300 mg of tissue. For this reason, alternative assay methods such as immunohistochemical detection of ER␣ and PR in fixed tissue sections or in needle aspiration biopsies are employed. Recently, we developed and characterized a site-directed monoclonal anti-ER␣ antibody, EVG F9. This antibody recognizes an epitope located in the A/B region (amino acids 140 –154) of human ER␣ [21]. The objective of this study was to assess the potential utility of EVG F9 monoclonal anti-ER␣ antibody in detecting ER␣ in human breast tumors by immunoblotting and immunohistochemical assays in human breast tumor samples.
0039-128X/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 3 9 - 1 2 8 X ( 0 0 ) 0 0 1 0 2 - 1
430
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
2. Materials and methods 2.1. Human breast tumor tissue specimens Human breast tumor tissues were obtained from patients who underwent surgical treatment of breast cancer at the Boston University Medical Center, Boston, MA, USA. Tumor specimens were transported to the laboratory on dry ice and analyzed for ER/PR concentrations by ligand-binding assay as described previously [22,23]. 2.2. Western blot analyses of human ER␣ Recombinant human ER␣ was obtained from Panavera, Madison, WI, USA. Breast tumor cytosol was prepared as described previously [22,23]. Human breast tumor cytosols and samples of purified recombinant human estrogen receptor (rhER␣), in varying concentrations, were layered onto 7% resolving sodium dodecyl sulfatepolyacrylamide gels (SDS/PAGE). The gels were electrophoresed for 15 to 17 h at 17 mAmps/500 Volts. The proteins were then electrophoretically transferred onto a nitrocellulose membrane for 3 h at 2°C at 0.36 Amps/200 Volts. The membrane was then incubated in Tris Buffered Saline containing 0.05% Tween 20 and 5% fat-free dried milk (TBST/5% milk) for 1 h to block nonspecific protein binding. This was followed by immunoblotting analysis with the monoclonal antibody EVG F9 (1:1000) for 1 h. After washing, the membranes were incubated with horseradish peroxidase-conjugated rabbit antimouse IgG secondary antibody for 1 h, and washed. ER-EVG F9 complexes were detected using enhanced chemiluminescence (ECL) (Pierce Chemical Co., Rockford, IL, USA). 2.3. Immunohistochemistry Formalin-fixed, paraffin-embedded blocks of tissue were re-cut at a thickness of 4 m, and consecutive sections were stained with Gill’s hematoxylin and eosin (H&E), standard protocol, for general light microscopic evaluation. Immunohistochemical assays with EVG F9 and H222 were performed using a modification of the antigen-retrieval technique based on microwave exposure [25]. Unless otherwise stated, all incubations were carried out at 37°C in a humidity chamber, followed by phosphate-buffer saline (PBS) plus Triton X-100 (1:500) washes, twice for 2 min each. Endogenous peroxidase activity was quenched using a 3% hydrogen peroxide/methanol solution for 30 min at 25°C. This was followed by microwave antigen retrieval and subsequent incubation in blocking reagent (1:50 Normal Horse Serum) for 20 min. EVG F9 antibody (dilution of 1:200) was added and the samples were incubated for 30 min. Parallel tissue sections were incubated with H222 monoclonal anti ER antibody (obtained in a commercial ER-ICA kit; Abbott Laboratories, North Chicago, IL, USA) according to
the manufacturer’s instructions. Nonimmune mouse IgG was used as a negative control antibody. The sections were rinsed and then incubated with the secondary antibody, horseradish peroxidase-conjugated rabbit anti-mouse IgG, for 30 min, followed by a 30 min incubation in Streptavidin solution, using DAKO’s Quick Staining LSAB Kit (DAKO Corp., Carpinteria, CA, USA) with EVG F9 and Abbott’s supplied reagents for H222 antibody. To visualize EVG F9 antibody binding, tissue sections were incubated for ten minutes in 3,3⬘-diaminobenzidine (DAB) at 25°C. To visualize staining with H222 antibody, the reagents supplied with the kit were used that produced purple-blue end product. Sections were rinsed in distilled water and counterstained for 1 min in Gill’s hematoxylin. Slides were prepared for analysis by standard light microscopy. 2.4. Semi-quantitative analysis of the immunohistochemical staining results Individual nuclei of carcinoma cells were scored on the basis of nuclear staining intensity on a 5-point scale, with a score of zero (0) indicating no staining and a score of 3 indicating very strong staining. Positive nuclear staining was observed in adjacent normal mammary luminal epithelial cells; however, tumors were considered to be positive when there was more than 20% distinct brown nuclear staining of varying intensity in malignant epithelial cells only. For each tumor specimen, the number of positive or negative malignant cells was evaluated in at least four randomly selected high-powered microscopic fields. Each stained section was assigned a histoscore (HS) calculated from the following formula: HS ⫽ ¥ (I ⫹ 1) ⫻ PI, in which I is nuclear staining intensity (0 ⫺ 3) and PI is the percentage of positively stained malignant epithelial cell nuclei [26]. Intra-observer variability was evaluated by a second examination of the tissue sections two weeks after the initial evaluation without knowledge of the initial results. An 81% agreement of HS values for ER expression using the monoclonal antibody, EVG F9, was obtained between the two separate evaluations. This value is well within the range of normal intra-observer variation. 2.5. Statistical analysis Spearman’s rank-order correlation coefficients (rs) were calculated to determine the association between ER expression as detected by the ligand-binding assay and immunohistochemistry. The strength and directionality of this association was characterized. The significance of rs was tested using critical values of Spearman’s rankorder correlation coefficients, where “n” was the number of pairs of ranks analyzed. Statistical significance was accepted if P ⬍ 0.05.
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
431
3. Results 3.1. Specificity of EVG F9 for human ER␣ Recent cloning of ER showed that the amino acid sequence of this ER isoform shares considerable homology with ER␣. The homology between ER␣ and ER is 28% in the A/B region, 96% in the DNA-binding region, 17% in the hinge region, 58% in the hormone binding region, and 18% in the C-terminal F region [28]. In light of this homology, it is likely that antibodies raised to ER␣ may cross-react with ER and vice versa. We have tested if EVG F9 monoclonal antibody cross-reacts with ER, using Western Blot analyses. Recombinant human ER␣ and recombinant rat ER were analyzed by Western Blots. As shown in Fig. 1, EVG F9 monoclonal antibody recognized recombinant human ER␣ (lane 1) but not recombinant rat ER (lane 3), suggesting that EVG F9 monoclonal antibody is specific for ER␣. Furthermore, the binding of EVG F9 to ER␣ was displaced by inclusion of the free immunogenic peptide (Fig. 1, lane 2) indicating epitope specificity. These observations suggest that EVG F9 monoclonal antibody is specific to ER␣ and is unique to a specific region of ER␣ protein. 3.2. Detection of human breast tumor estrogen receptors by anti-ER␣ site-directed monoclonal antibody: Western blot analysis To test the potential utility and sensitivity of EVG F9 monoclonal antibody in detecting human breast tumor ER␣ by Western Blot analysis, we analyzed varying concentrations of recombinant human ER␣ (Panavera, Madison, WI, USA) by electrophoresis on 7% SDS-PAGE gels. The resolved proteins were detected by Western Blot analysis using EVG F9 as the primary antibody. As shown in Fig. 2, EVG F9 detected recombinant human ER␣ in a concentration-dependent manner, with sensitivity approaching 5 to 10 fmol/sample. These data suggested that EVG F9 monoclonal antibody is a useful tool in detection of ER␣ in human breast tumor tissue extracts by immunoblotting techniques. To test this premise, we analyzed 20 human breast tumor cytosols for ER␣, using ligand-binding assays and Western Blot analyses. As shown in Fig. 3, ER␣ was detected in human breast tumor cytosolic extracts by immunoblotting. For the majority of the tumors tested, the 65 kDa band intensities reflected the receptor content determined by the biochemical assay; however, in some tumors, the value of ER␣ determined by ligand-binding assays did not correlate well with band intensity. This may be attributed to proteolysis of ER␣ in the A/B region of the receptor where the epitope of this antibody is localized. If this were the case, one would expect to detect less ER␣ by Western Blot than by ligand-binding studies. In some tumors, while the ligandbinding indicated reasonable ER␣ values, some faint or weak bands were detected by Western Blots (see Fig. 3
Fig. 1. EVG F9 monoclonal antibody binds specifically to purified recombinant human ER␣ and not ER. Aliquots of purified rh ER␣ and recombinant rat ER (generous gift of Dr C.M. Klinge, Louisville, KY, USA) were analyzed by Western blotting. Lane 1 represents ER␣ (100 fmol/ sample) immunoblotted with EVG F9 only. Lane 2 represents ER␣ (100 fmol/sample) immunoblotted with EVG F9 pre-incubated with the immunogenic peptide for 30 min before blotting. Lane 3 represents ER (1300 fmol/sample) immunoblotted with EVG F9 only. Lane 4 represents ER (1300 fmol/sample) immunoblotted with EVG F9 pre-incubated with the immunogenic peptide for 30 min before blotting.
lower panel, lanes 3, 5, and 8). Proteolysis of ER␣ in the ligand-binding region, however, would provide lower ER␣ values in ligand-binding assays but would not show decreased band intensity in Western Blot analysis. Because EVG F9 is a monoclonal antibody, proteolysis of the epitope would preclude detection of any smaller bands of proteolytically cleaved ER␣. However, proteolysis in other domains would produce smaller molecular weight bands. It is likely that the smaller bands observed in lanes 1, 2, 3, 6, 7, and 10 of the lower panel of Fig. 3 are due to ER␣ proteolysis. The high molecular weight bands observed in
Fig. 2. Detection of purified recombinant human estrogen receptor by Western Blot analysis using monoclonal antibody EVG F9. Samples of purified rh ER␣, in varying concentrations, were layered onto 7% SDS/ PAGE gels, electro-transferred onto a nitrocellulose membrane, and detected using the EVG F9 monoclonal antibody. ER concentrations are indicated below each lane.
432
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
Fig. 3. Detection of ER␣ in human breast tumor cytosol by Western Blot analysis using monoclonal antibody EVG F9. Cytosolic extracts from 20 human breast tumors were prepared, and aliquots containing ⬃200 g of total tumor protein were layered onto 7% SDS/PAGE gels, and then electro-transferred onto a nitrocellulose membrane. ER␣ was detected using EVG F9. The ER values from the ligand-binding studies for each sample are shown underneath the corresponding lanes.
Fig. 3 (lanes 2, 3, 8, and 10 of the upper panel and in lanes 8, and 10 of the lower panel) may represent receptor conjugation to ubiquitin or to other proteins in the process of turnover [29,30]. 3.3. Immunohistochemical analyses of human breast tumor estrogen receptors by ER␣ site-directed monoclonal antibody, EVG F9, and a commercially available monoclonal antibody, H222 To demonstrate that EVG F9 monoclonal antibody is a useful tool in assessing ER␣ status in human breast tumors by immunohistochemistry, we first determined that EVG F9 specifically detected ER␣ by this technique (Fig. 4A). Incubation of EVG F9 with the free immunogenic peptide for 30 min at 4°C before incubation with tissue sections eliminated the specific staining (Fig. 4B), as compared to staining of sections incubated with the antibody alone (Fig. 4A). These observations corroborated those obtained by Western Blot analyses in Fig. 1. Based on these observations, we determined ER␣ status in 25 human breast carcinoma tissue samples by ligand binding assays and immunohistochemistry, using EVG F9 and H222, a commercially available antibody from the ER-ICA kit (Abbott Laboratories, North Chicago, IL, USA). As shown in Table 1, seven out of twenty-five tumors (28%) did not express functional ER␣ (ER-negative; less than 10-fmol/mg protein), based on ligand binding assays. Four out of 25 tumors (16%) had ER␣
values ranging between 10.0 and 99.0 fmol/mg protein (mean value ⫽ 45.25 fmol/mg protein). Nine of 25 tumors (36%) had ER␣ content between 100.00 and 199.00 fmol/mg protein (mean value ⫽ 139.33 fmol/mg protein). Six of 25 (24%) had ER␣ content equal or more than 200 fmol/mg protein (mean value ⫽ 392.67 fmol/mg protein). Qualitative immunohistochemical analyses of ER␣ expression in the same tumor samples were carried out with EVG F9 (Fig. 5) and H222 antibodies (Fig. 6). Receptor expression was assessed semi-quantitatively, and a histoscore was assigned to each specimen, based on the intensity of nuclear staining and the percentage of positively stained breast carcinoma cell nuclei (Fig. 5 and 6). With EVG F9, a score of 0 indicates no detectable staining. A score of 1⫹ indicates weak staining. A score of 2⫹ indicates definitive staining of moderate intensity. A score of 3⫹ indicates strong to very strong staining. Tumors were considered ER⫹ when there was more than 20% distinct nuclear staining (1⫹ or greater) in tumor cells. Based on the intensity of the staining observed with EVG F9, we consider a score of 1⫹ or greater to indicate positive ER status. A score of 2⫹ indicates the presence of significant ER. The data revealed that ten tumors (40%) exhibited a strong staining pattern, seven (28%) exhibited moderate staining, and nine tumors (36%) exhibited a weak staining pattern. To determine the strength and directionality of association between ER␣ content in breast tumor specimens examined, Spearman’s rank-order correlational analysis was per-
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
433
Table 1 Estrogen/progesterone receptor status in 25 cases of human breast tumors determined by ligand-binding studies and immunohistochemistrya
Fig. 4. Specificity of EVG F9 monoclonal antibody in detection of ER␣ by immunohistochemical analysis. Immunohistochemical assays of human breast tumor tissue with EVG F9 monoclonal antibody were carried out as described in methods. Panel A: Tissues sections were incubated with EVG F9 only. Panel B: Tissue sections were incubated with EVG F9 together with 10 g/ml of the immunogenic peptide.
formed. The model included: 1) ER␣/PR content determined by ligand binding studies and 2) ER distribution determined by immunohistochemical analyses, using EVG F9 and H222 antibodies. The Spearman rank-order correlation coefficient (r s ), the non-parametric analog of the most commonly reported measure of correlation, was determined (Table 2). There was evidence of a statistically significant strong positive correlation between the ER content determined by ligand binding assay and by the histoscore evaluation of EVG F9-stained breast tumor sections. Also, the data obtained with EVG F9 correlated well with that obtained with H222 antibody.
4. Discussion Tumor hormonal phenotype has long been utilized as a useful indicator of breast carcinoma progression. Although tumor invasion of axillary nodes predicts a high risk of metastatic disease, estrogen receptor status is an effective prognostic marker of the biological aggressiveness of the
Case #
ER (fmol/mg protein)
PR (fmol/mg protein)
ER-IHCAb (H222)
ER-IHCAb (EVG F9)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
0 0 0 0 0 0 0 23 39 55 64 100 112 132 132 144 163 171 194 196 305 327 340 356 387
0 0 40 2 4 21 9 117 0 160 6 38 113 967 141 157 10 22 493 104 114 852 195 17 176
1⫹ 0 0 0 0 2⫹ 0 1⫹ 1⫹ 0 1⫹ 1⫹ 1⫹ 2⫹ 1⫹ 3⫹ 2⫹ 2⫹ 2⫹ 2⫹ 3⫹ 1⫹ 0 2⫹ 2⫹
1⫹ 3⫹ 0 1⫹ 0 2⫹ 1⫹ 3⫹ 1⫹ 3⫹ 3⫹ 0 1⫹ 0 2⫹ 2⫹ 0 1⫹ 3⫹ 1⫹ 3⫹ 0 2⫹ 3⫹ 3⫹
a Tumor ER content was measured by ligand binding assay and immunohistochemistry with two monoclonal antibodies (H222 and EVG F9). Tumor PR content was measured by ligand binding assay. Results of the immunochemical analyses were assessed semi-quantitatively, and a histoscore was assigned to each specimen based on the intensity of staining and the percentage of positively stained cells for each antibody. b ER-IHCA indicates tumor estrogen receptor content determined by immunohistochemical analyses. Type of the anti-ER antibody used is included.
primary tumor [3,8]. Previous studies have shown that tumors, which do not express ER (ER-negative), are often poorly differentiated and more clinically aggressive [27]. Currently, several ER assay techniques are utilized for determination of ER content in human breast tumors. Due to the advanced methods of early detection of breast tumors using mammography and other imaging techniques, the small size of tumors detected and excised may preclude quantitation of ER by conventional ligand-binding assays, which requires a minimum of 200 to 300 mg of tissue. Thus, immunoblotting and immunohistochemical analyses of ER become the likely methods of choice. However, both of these methods suffer from some limitations. One of the limitations is that detection of ER is only semi-quantitative. Furthermore, these methods depend on the availability of sensitive and specific ER antibodies. In this study, we demonstrated that EVG F9 is specific for ER␣ and that this monoclonal antibody does not crossreact with recombinant rat ER. Given the high sequence
434
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
Fig. 5. Subcellular localization of estrogen receptors in human breast tumor tissue determined by immunohistochemical analysis with EVG F9. Immunohistochemical assays of human breast tumor tissue with EVG F9 monoclonal antibody were carried out as described in methods. Different degrees of nuclear staining intensity and distribution patterns were visualized by light microscopy. Panel A: strong immunostaining of human breast tumor tissue with EVG F9 (⫻10). Panel B: moderate immunostaining of human breast tumor tissue with EVG F9. (⫻10). Panel C: weak immunostaining of human breast tumor tissue with EVG F9. (⫻10). Panel D: human breast tumor tissue incubated with nonimmune mouse IgG (negative control) (⫻10).
homology between the N-terminal domains of human and rat ER, we do not anticipate that EVG F9 will react with human ER. Further, EVG F9 binds to ER␣ at a defined
epitope as indicated by complete displacement of EVG F9 binding to ER␣ by the immunogenic free peptide. We further demonstrated that EVG F9 detects denatured human
Fig. 6. Subcellular localization of estrogen receptors in human breast tumor tissue determined by immunohistochemical analysis with H222 monoclonal antibody. Tissue sections from the same blocks of human breast tumors shown in Fig. 5 were subjected to immunohistochemical analyses using monoclonal antibody H222. Different degrees of nuclear staining intensity and distribution patterns were visualized by light microscopy. Panel A: strong immunostaining of human breast tumor tissue (⫻10). Panel B: weak immunostaining of human breast tumor tissue (⫻10). Panel C: moderate immunostaining of human breast tumor tissue (⫻10). Panel D: human breast tumor tissue incubated with nonimmune mouse IgG (negative control) (⫻10).
N.S. Rost et al. / Steroids 65 (2000) 429 – 436 Table 2 Comparison of ER/PR values by ligand-binding studies and immunohistochemistrya Variable
ER
PR
ER-IHCAb (H222)
ER-IHCAb (EVG F9)
ER PR ER-IHCAb (H222) ER-IHCAb (EVG F9)
1
0.62 1
0.5 0.26 1
0.81 0.35 0.52 1
a Tumor biochemical and immunohistochemical characteristics were evaluated using Spearman’s rank-order correlational analysis. The matrix of intercorrelations represents Spearman’s rank-order correlation coefficients, (rs), for all pairs of variables. The values below the diagonal are redundant of those above the diagonal, and therefore, are not included. The values on the diagonal are all 1.00, representing the perfect correlation of each variable with itself. Bold print is used to distinguish statistically significant values (P ⬍ 0.05). b ER-IHCA indicates tumor estrogen receptor content determined by immunohistochemical analyses. Type of the anti-ER antibody used in included.
ER␣ at low concentrations, approaching 5 to 10 fmol/sample, indicating the antibody’s specificity and sensitivity. The data from Western Blot analyses reflected the values of ER␣ measured by ligand-binding assays. It should be noted that immunoblotting detects total ER␣ (i.e. functional and nonfunctional) whereas ligand-binding assays detect only the unoccupied functional ER␣. Discrepancies between the results of ligand-binding studies and Western Blots are expected in some tumor samples due to the differences in the detection methods of these two approaches. Because EVG F9 monoclonal antibody recognizes a specific epitope on the receptor, removal of this epitope due to proteolysis would result in a decrease in the amount of ER␣ detected by blotting. Even with these technical limitations, the immunoblotting assays provided an ER␣ status similar to those obtained with ligand-binding assays. Analysis of ER␣ in formalin-fixed, paraffin-embedded human breast tumor samples demonstrated that EVG F9 monoclonal antibody is specific, and produces comparable staining to that obtained with H222 antibody (r s ⫽ 0.52). The data obtained with immunohistochemistry correlated well with those determined by ligand-binding studies (EVG F9, r s ⫽ 0.81; H222, r s ⫽ 0.50). These observations suggest that EVG F9 monoclonal antibody is a useful tool for the analysis of ER␣ in tumor extracts by Western Blot and in fixed human breast tumor samples by immunohistochemistry. It is important to note that although there was strong agreement between the data obtained by ligandbinding assays and immunohistochemical assay, in some tumors the immunohistochemical assay indicated presence of ER␣, whereas ligand-binding assays did not and vice versa. These discrepancies may be related to the detection of functional and nonfunctional ER␣ by these two methods. Further, comparing the data obtained with EVG F9 and H222, there was an agreement in 14 of 25 of tumors (56%). In four tumors, EVG F9 monoclonal antibody detected low
435
levels of ER␣, whereas the H222 monoclonal antibody did not. Similarly, H222 detected ER␣ in four tumors whereas EVG F9 did not. These observations suggest that epitope preservation and retrieval may be different in some tumors and this may explain the differences in ER␣ detectability with various antibodies. Although ER␣ is undetectable by ligand-binding assays in some tumors, and these tumors are denoted as ER-negative, ER␣ is detectable by immunohistochemistry and this may reflect the greater sensitivity of ER␣ detection by EVG F9. It should be noted that ER␣ detected by immunohistochemistry correlated well with that of PR determined by ligand-binding assays. This observation lends further support to strong association between values determined by ligand-binding assays and immunohistochemistry. Furthermore, we consistently observed varying degrees of cytoplasmic staining in human breast tumors. Because the detection of ER␣ by EVG F9 using Western Blot analysis and the staining of ER␣ in immunocytochemistry are abolished by inclusion of the immunogenic peptide, we suggest that the cytoplasmic staining represents some form of ER␣ in the cytoplasm. Whether this represents a functional or nonfunctional ER␣ and the significance of this observation, however, remains to be determined. The data presented here suggest that human breast tumor ER␣ status determined by immunoblotting and immunohistochemical analyses, using EVG F9 monoclonal anti-ER␣ antibody, produced comparable results to those obtained by ligand-binding assays. These observations suggest that EVG F9 is potentially a useful immunochemical tool for assessment of ER␣ status in human breast tumor tissues, using Western Blot analysis or immunohistochemistry.
Acknowledgment We wish to acknowledge Dr Antonio de las Morenas of the Mallory Institute of Pathology for his helpful discussion of the immunohistochemistry data.
References [1] Qualters JR, Lee NC, Smith RA, et al. Breast and cervical cancer surveillance, United States, 1973–1987. MMWR CDC Surveillance Summaries 1992;41(2):1–15. [2] Miller BA, Feuer EJ, Hankey BF. Recent incidence trends for breast cancer in women and the relevance of early detection. CA-A Cancer J Clin 1993;43(1):27– 41. [3] McGuire WL. Hormone receptors: their role in predicting prognosis and responses to endocrine therapy. Semin Oncol 1978;5:428 –33. [4] Burstein NA. Overall principles of cancer management. In: Cancer manual, 6th ed. Boston: American Cancer Society, Massachusetts Div., 1982. p. 67–71. [5] Brooks SC, Saunders DE, Singhakowinta A, et al. Relation of tumor content with response of patient to endocrine therapy. Cancer 1980; 46:2775– 8.
436
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
[6] Degenshein GA, Bloom N, Tobin E. The value of progesterone receptor assays in the management of advanced breast cancer. Cancer 1980;46:2789 –93. [7] McGuire WL, Pearson OH, Segaloff A. Predicting hormone responsiveness in human breast cancer. In: McGuire WL, Carbone PP, Vollmer ED, editors. Estrogen receptors in breast cancer. New York: Raven Press, 1975. p. 17–30. [8] McGuire WL. Prognostic factor for recurrence and survival in human breast cancer. Breast Cancer Res Treat 1987;10:5–9. [9] Elledge RM, McGuire WL. Prognostic factors and therapeutic decisions in axillary node-negative breast cancer. Ann Rev Med 1993; 44:201–10. [10] Gelbfish GA, Davidson AI, Kopel S, et al. Relationship of estrogen and progesterone receptors to prognosis in breast cancer. Ann Surg 1988;207(1):75–9. [11] Williams MR, Todd JH, Ellis IO, et al. Oestrogen receptors in primary and advanced breast cancer: an 8 year review of 704 cases. Br J Cancer 1987;55:67–73. [12] Kohail HM, Elias EG, El-Nowiem SA, et al. A multifactorial analysis of steroid hormone receptors in stages I and II breast cancer. Ann Surg 1985;201(5):611–7. [13] Donegan WL. Prognostic factors, stage, and receptor status in breast cancer. Cancer 1992;70:1755– 64. [14] Millis RR. Correlation of hormone receptors with pathological features in human breast cancer. Cancer 1980;46:2869 –71. [15] McCarty KS, Barton TK, Fetter BF, et al. Correlation of estrogen and progesterone receptors with histologic differentiation in mammary carcinoma. Cancer 1980;46:2851– 8. [16] Cooper JA, Rohan TE, Cant EL, et al. Risk factors for breast cancer by oestrogen receptor status: a population-based case-control study. Br J Cancer 1989;59:119 –25. [17] Ballard–Barbash R, Griffin MR, Fisher LD, et al. Estrogen receptors in breast cancer. Am J Epidemiol 1986;124:77– 84. [18] King RJB, Redgrave S, Hayward JL, Millis RR, Rubens RD. The measurement of receptors for oestradiol and progesterone in human breast tumors. In: Steroid receptor assay in human breast tumors. King RJB, editor, Methodological and Clinical Aspects. Alpha Omega, Cardiff, 1979. p. 55–72. [19] Shousha S, Coady AT, Stamp T, James JR, Alaghband–Zadeh J. Oestrogen receptors in mucinous carcinoma of the breast: an immunohistological study using paraffin wax sections. J Clin Pathol 1989; 42:902–5.
[20] Shousha S, Stamp T, James JR, Alaghband–Zadeh J. Immunohistological study of oestrogen receptors in breast carcinomas that are biochemically receptor negative. J Clin Pathol 1990;43:239 – 42. [21] Traish AM, Pavao M. Binding of site-directed monoclonal antibodies to an epitope located in the A/B region (amino acids 140 –154) of human estrogen receptor induced conformational changes in an epitope in the DNA-binding domain. Steroids 1996;61:549 –56. [22] Traish AM, Al–Fadhli S, Klinge C, Kounine M, Quick TC. Identification of structurally altered estrogen receptors in human breast cancer by site-directed monoclonal antibodies. Steroids 1995;60:467– 74. [23] Traish AM, Newton AW, Styperek K, Beazley R, Kavanah M. Estrogen receptor functional status in human breast cancer. Diagn Mol Pathol 1995;4(3):220 – 8. [24] Joel PB, Traish AM, Lannigan DA. Estradiol-induced phosphorylation of serine 118 in the estrogen receptor is independent of p42/p44 mitogen-activated protein kinase. J Biol Chem 1998;273(21):13317– 23. [25] Oyaizu T, Seizaburo A, Takehiko H, Tsubura A. Immunohistochemical detection of estrogen and progesterone receptors performed with an antigen-retrieval technique on methacarn-fixed paraffin-embedded breast cancer tissues. J Surgical Res 1996;60:69 –73. [26] Helin ML, Helle MJ, Helin HJ, Isola J. Proliferative activity and steroid receptors determined by immunohistochemistry in adjacent frozen sections of 102 breast carcinomas. Arch Pathol Lab Med 1989;113:854 –7. [27] Anderson J, Bentzen SM, Poulsen HS. Relationship between radioligand binding assay, immunoenzyme assay and immunohistochemical assay for estrogen receptors in human breast cancer and association with tumor differentiation. Eur J Cancer Oncol 1988;24:377– 84. [28] Enmark E, Pelto–Huikko M, Grandien K, et al. Human estrogen receptor beta-gene structure, chromosomal localization, and expression pattern. J Clin Endocrinol Metab 1997;82:4258 – 65. [29] Alarid ET, Bakopoulos N, Solodin N. Proteasome-mediated proteolysis of estrogen receptor: a novel component in autologous downregulation. Mol Endocrinol 1999;13:1522–34. [30] Nawaz Z, Lonard DM, Dennis AP, Smith CL, O’Malley BW. Proteasome-dependent degradation of the human estrogen receptor. Proc Natl Acad Sci USA 1999;96:1858 – 62.
Immunochemical analyses of estrogen receptors in human breast tumors by a novel monoclonal estrogen receptor antibody (EVG F9) Natalia S. Rost, Kristine Murphy, Laurie Hafer, Matthew Pavao, Abdulmaged M. Traish* Department of Biochemistry, Boston University School of Medicine, Center for Advanced Biomedical Research, Boston, MA 02118, USA Received 14 September 1999; received in revised form 21 February 2000; accepted 1 March 2000
Abstract The objective of this study was to assess the potential utility of a new site-directed, monoclonal anti-estrogen receptor antibody (EVG F9) in detection and analyses of human breast tumor estrogen receptor (ER␣), using immunoblotting and immunohistochemical assays. Using Western Blot analyses, we demonstrated that EVG F9 monoclonal antibody binds specifically to ER␣ and does not cross-react with ER. Furthermore, binding of EVG F9 to ER␣ was effectively displaced with the immunogenic peptide in Western Blots and in immunohistochemical analyses. In Western Blot analyses, EVG F9 detected ER␣ at low concentrations approaching 5 to 10 fmol/sample. Determination of ER␣ status of a series of human breast tumor samples by Western Blot analyses or immunohistochemistry using EVG F9 correlated well with ER␣ values measured by ligand binding assays. These observations suggest that EVG F9 monoclonal anti-ER␣ antibody is a valuable immunochemical tool for detection and analyses of ER␣ in human breast tumors. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Breast cancer; Estrogen receptor; Monoclonal antibodies
1. Introduction Breast cancer is the second leading cause of cancer death among women in the United States. Since the 1980s, the incidence of breast cancer has been increasing with the highest incidence occurring in postmenopausal women [1,2]. It is well documented that ovarian hormones play a major role in development and progression of breast cancer. Tumor expression of estrogen (ER) and progesterone (PR) receptors is associated with favorable response to endocrine therapy [3– 6], extended 5-year survival [7–12], and characterized by a more histologically differentiated tumor phenotype [13–16]. The expression of ER and PR varies with tumor phenotype and with age and may represent the marked heterogeneity of breast cancer [17,18]. ER␣ status is currently used as a This work was supported by a grant from the Department of the Army DAMD 1794J4468. * Corresponding author. Tel.: ⫹-617-638-4578; fax: ⫹-617-638-5412. E-mail address: [email protected] (A.M. Traish).
prognostic factor in management of human breast cancer [3–15]. Both ligand-binding studies and immunohistochemical analyses are currently utilized to determine ER and PR status in human breast tumor tissue specimens [18 –20]. The advent of mammography and other sophisticated imaging techniques has dramatically improved early detection of breast tumors. These advances have resulted in surgical removal of tumors of small size, precluding analyses of the ER␣ by ligand-binding assay, which requires a minimum of 200 to 300 mg of tissue. For this reason, alternative assay methods such as immunohistochemical detection of ER␣ and PR in fixed tissue sections or in needle aspiration biopsies are employed. Recently, we developed and characterized a site-directed monoclonal anti-ER␣ antibody, EVG F9. This antibody recognizes an epitope located in the A/B region (amino acids 140 –154) of human ER␣ [21]. The objective of this study was to assess the potential utility of EVG F9 monoclonal anti-ER␣ antibody in detecting ER␣ in human breast tumors by immunoblotting and immunohistochemical assays in human breast tumor samples.
0039-128X/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 3 9 - 1 2 8 X ( 0 0 ) 0 0 1 0 2 - 1
430
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
2. Materials and methods 2.1. Human breast tumor tissue specimens Human breast tumor tissues were obtained from patients who underwent surgical treatment of breast cancer at the Boston University Medical Center, Boston, MA, USA. Tumor specimens were transported to the laboratory on dry ice and analyzed for ER/PR concentrations by ligand-binding assay as described previously [22,23]. 2.2. Western blot analyses of human ER␣ Recombinant human ER␣ was obtained from Panavera, Madison, WI, USA. Breast tumor cytosol was prepared as described previously [22,23]. Human breast tumor cytosols and samples of purified recombinant human estrogen receptor (rhER␣), in varying concentrations, were layered onto 7% resolving sodium dodecyl sulfatepolyacrylamide gels (SDS/PAGE). The gels were electrophoresed for 15 to 17 h at 17 mAmps/500 Volts. The proteins were then electrophoretically transferred onto a nitrocellulose membrane for 3 h at 2°C at 0.36 Amps/200 Volts. The membrane was then incubated in Tris Buffered Saline containing 0.05% Tween 20 and 5% fat-free dried milk (TBST/5% milk) for 1 h to block nonspecific protein binding. This was followed by immunoblotting analysis with the monoclonal antibody EVG F9 (1:1000) for 1 h. After washing, the membranes were incubated with horseradish peroxidase-conjugated rabbit antimouse IgG secondary antibody for 1 h, and washed. ER-EVG F9 complexes were detected using enhanced chemiluminescence (ECL) (Pierce Chemical Co., Rockford, IL, USA). 2.3. Immunohistochemistry Formalin-fixed, paraffin-embedded blocks of tissue were re-cut at a thickness of 4 m, and consecutive sections were stained with Gill’s hematoxylin and eosin (H&E), standard protocol, for general light microscopic evaluation. Immunohistochemical assays with EVG F9 and H222 were performed using a modification of the antigen-retrieval technique based on microwave exposure [25]. Unless otherwise stated, all incubations were carried out at 37°C in a humidity chamber, followed by phosphate-buffer saline (PBS) plus Triton X-100 (1:500) washes, twice for 2 min each. Endogenous peroxidase activity was quenched using a 3% hydrogen peroxide/methanol solution for 30 min at 25°C. This was followed by microwave antigen retrieval and subsequent incubation in blocking reagent (1:50 Normal Horse Serum) for 20 min. EVG F9 antibody (dilution of 1:200) was added and the samples were incubated for 30 min. Parallel tissue sections were incubated with H222 monoclonal anti ER antibody (obtained in a commercial ER-ICA kit; Abbott Laboratories, North Chicago, IL, USA) according to
the manufacturer’s instructions. Nonimmune mouse IgG was used as a negative control antibody. The sections were rinsed and then incubated with the secondary antibody, horseradish peroxidase-conjugated rabbit anti-mouse IgG, for 30 min, followed by a 30 min incubation in Streptavidin solution, using DAKO’s Quick Staining LSAB Kit (DAKO Corp., Carpinteria, CA, USA) with EVG F9 and Abbott’s supplied reagents for H222 antibody. To visualize EVG F9 antibody binding, tissue sections were incubated for ten minutes in 3,3⬘-diaminobenzidine (DAB) at 25°C. To visualize staining with H222 antibody, the reagents supplied with the kit were used that produced purple-blue end product. Sections were rinsed in distilled water and counterstained for 1 min in Gill’s hematoxylin. Slides were prepared for analysis by standard light microscopy. 2.4. Semi-quantitative analysis of the immunohistochemical staining results Individual nuclei of carcinoma cells were scored on the basis of nuclear staining intensity on a 5-point scale, with a score of zero (0) indicating no staining and a score of 3 indicating very strong staining. Positive nuclear staining was observed in adjacent normal mammary luminal epithelial cells; however, tumors were considered to be positive when there was more than 20% distinct brown nuclear staining of varying intensity in malignant epithelial cells only. For each tumor specimen, the number of positive or negative malignant cells was evaluated in at least four randomly selected high-powered microscopic fields. Each stained section was assigned a histoscore (HS) calculated from the following formula: HS ⫽ ¥ (I ⫹ 1) ⫻ PI, in which I is nuclear staining intensity (0 ⫺ 3) and PI is the percentage of positively stained malignant epithelial cell nuclei [26]. Intra-observer variability was evaluated by a second examination of the tissue sections two weeks after the initial evaluation without knowledge of the initial results. An 81% agreement of HS values for ER expression using the monoclonal antibody, EVG F9, was obtained between the two separate evaluations. This value is well within the range of normal intra-observer variation. 2.5. Statistical analysis Spearman’s rank-order correlation coefficients (rs) were calculated to determine the association between ER expression as detected by the ligand-binding assay and immunohistochemistry. The strength and directionality of this association was characterized. The significance of rs was tested using critical values of Spearman’s rankorder correlation coefficients, where “n” was the number of pairs of ranks analyzed. Statistical significance was accepted if P ⬍ 0.05.
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
431
3. Results 3.1. Specificity of EVG F9 for human ER␣ Recent cloning of ER showed that the amino acid sequence of this ER isoform shares considerable homology with ER␣. The homology between ER␣ and ER is 28% in the A/B region, 96% in the DNA-binding region, 17% in the hinge region, 58% in the hormone binding region, and 18% in the C-terminal F region [28]. In light of this homology, it is likely that antibodies raised to ER␣ may cross-react with ER and vice versa. We have tested if EVG F9 monoclonal antibody cross-reacts with ER, using Western Blot analyses. Recombinant human ER␣ and recombinant rat ER were analyzed by Western Blots. As shown in Fig. 1, EVG F9 monoclonal antibody recognized recombinant human ER␣ (lane 1) but not recombinant rat ER (lane 3), suggesting that EVG F9 monoclonal antibody is specific for ER␣. Furthermore, the binding of EVG F9 to ER␣ was displaced by inclusion of the free immunogenic peptide (Fig. 1, lane 2) indicating epitope specificity. These observations suggest that EVG F9 monoclonal antibody is specific to ER␣ and is unique to a specific region of ER␣ protein. 3.2. Detection of human breast tumor estrogen receptors by anti-ER␣ site-directed monoclonal antibody: Western blot analysis To test the potential utility and sensitivity of EVG F9 monoclonal antibody in detecting human breast tumor ER␣ by Western Blot analysis, we analyzed varying concentrations of recombinant human ER␣ (Panavera, Madison, WI, USA) by electrophoresis on 7% SDS-PAGE gels. The resolved proteins were detected by Western Blot analysis using EVG F9 as the primary antibody. As shown in Fig. 2, EVG F9 detected recombinant human ER␣ in a concentration-dependent manner, with sensitivity approaching 5 to 10 fmol/sample. These data suggested that EVG F9 monoclonal antibody is a useful tool in detection of ER␣ in human breast tumor tissue extracts by immunoblotting techniques. To test this premise, we analyzed 20 human breast tumor cytosols for ER␣, using ligand-binding assays and Western Blot analyses. As shown in Fig. 3, ER␣ was detected in human breast tumor cytosolic extracts by immunoblotting. For the majority of the tumors tested, the 65 kDa band intensities reflected the receptor content determined by the biochemical assay; however, in some tumors, the value of ER␣ determined by ligand-binding assays did not correlate well with band intensity. This may be attributed to proteolysis of ER␣ in the A/B region of the receptor where the epitope of this antibody is localized. If this were the case, one would expect to detect less ER␣ by Western Blot than by ligand-binding studies. In some tumors, while the ligandbinding indicated reasonable ER␣ values, some faint or weak bands were detected by Western Blots (see Fig. 3
Fig. 1. EVG F9 monoclonal antibody binds specifically to purified recombinant human ER␣ and not ER. Aliquots of purified rh ER␣ and recombinant rat ER (generous gift of Dr C.M. Klinge, Louisville, KY, USA) were analyzed by Western blotting. Lane 1 represents ER␣ (100 fmol/ sample) immunoblotted with EVG F9 only. Lane 2 represents ER␣ (100 fmol/sample) immunoblotted with EVG F9 pre-incubated with the immunogenic peptide for 30 min before blotting. Lane 3 represents ER (1300 fmol/sample) immunoblotted with EVG F9 only. Lane 4 represents ER (1300 fmol/sample) immunoblotted with EVG F9 pre-incubated with the immunogenic peptide for 30 min before blotting.
lower panel, lanes 3, 5, and 8). Proteolysis of ER␣ in the ligand-binding region, however, would provide lower ER␣ values in ligand-binding assays but would not show decreased band intensity in Western Blot analysis. Because EVG F9 is a monoclonal antibody, proteolysis of the epitope would preclude detection of any smaller bands of proteolytically cleaved ER␣. However, proteolysis in other domains would produce smaller molecular weight bands. It is likely that the smaller bands observed in lanes 1, 2, 3, 6, 7, and 10 of the lower panel of Fig. 3 are due to ER␣ proteolysis. The high molecular weight bands observed in
Fig. 2. Detection of purified recombinant human estrogen receptor by Western Blot analysis using monoclonal antibody EVG F9. Samples of purified rh ER␣, in varying concentrations, were layered onto 7% SDS/ PAGE gels, electro-transferred onto a nitrocellulose membrane, and detected using the EVG F9 monoclonal antibody. ER concentrations are indicated below each lane.
432
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
Fig. 3. Detection of ER␣ in human breast tumor cytosol by Western Blot analysis using monoclonal antibody EVG F9. Cytosolic extracts from 20 human breast tumors were prepared, and aliquots containing ⬃200 g of total tumor protein were layered onto 7% SDS/PAGE gels, and then electro-transferred onto a nitrocellulose membrane. ER␣ was detected using EVG F9. The ER values from the ligand-binding studies for each sample are shown underneath the corresponding lanes.
Fig. 3 (lanes 2, 3, 8, and 10 of the upper panel and in lanes 8, and 10 of the lower panel) may represent receptor conjugation to ubiquitin or to other proteins in the process of turnover [29,30]. 3.3. Immunohistochemical analyses of human breast tumor estrogen receptors by ER␣ site-directed monoclonal antibody, EVG F9, and a commercially available monoclonal antibody, H222 To demonstrate that EVG F9 monoclonal antibody is a useful tool in assessing ER␣ status in human breast tumors by immunohistochemistry, we first determined that EVG F9 specifically detected ER␣ by this technique (Fig. 4A). Incubation of EVG F9 with the free immunogenic peptide for 30 min at 4°C before incubation with tissue sections eliminated the specific staining (Fig. 4B), as compared to staining of sections incubated with the antibody alone (Fig. 4A). These observations corroborated those obtained by Western Blot analyses in Fig. 1. Based on these observations, we determined ER␣ status in 25 human breast carcinoma tissue samples by ligand binding assays and immunohistochemistry, using EVG F9 and H222, a commercially available antibody from the ER-ICA kit (Abbott Laboratories, North Chicago, IL, USA). As shown in Table 1, seven out of twenty-five tumors (28%) did not express functional ER␣ (ER-negative; less than 10-fmol/mg protein), based on ligand binding assays. Four out of 25 tumors (16%) had ER␣
values ranging between 10.0 and 99.0 fmol/mg protein (mean value ⫽ 45.25 fmol/mg protein). Nine of 25 tumors (36%) had ER␣ content between 100.00 and 199.00 fmol/mg protein (mean value ⫽ 139.33 fmol/mg protein). Six of 25 (24%) had ER␣ content equal or more than 200 fmol/mg protein (mean value ⫽ 392.67 fmol/mg protein). Qualitative immunohistochemical analyses of ER␣ expression in the same tumor samples were carried out with EVG F9 (Fig. 5) and H222 antibodies (Fig. 6). Receptor expression was assessed semi-quantitatively, and a histoscore was assigned to each specimen, based on the intensity of nuclear staining and the percentage of positively stained breast carcinoma cell nuclei (Fig. 5 and 6). With EVG F9, a score of 0 indicates no detectable staining. A score of 1⫹ indicates weak staining. A score of 2⫹ indicates definitive staining of moderate intensity. A score of 3⫹ indicates strong to very strong staining. Tumors were considered ER⫹ when there was more than 20% distinct nuclear staining (1⫹ or greater) in tumor cells. Based on the intensity of the staining observed with EVG F9, we consider a score of 1⫹ or greater to indicate positive ER status. A score of 2⫹ indicates the presence of significant ER. The data revealed that ten tumors (40%) exhibited a strong staining pattern, seven (28%) exhibited moderate staining, and nine tumors (36%) exhibited a weak staining pattern. To determine the strength and directionality of association between ER␣ content in breast tumor specimens examined, Spearman’s rank-order correlational analysis was per-
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
433
Table 1 Estrogen/progesterone receptor status in 25 cases of human breast tumors determined by ligand-binding studies and immunohistochemistrya
Fig. 4. Specificity of EVG F9 monoclonal antibody in detection of ER␣ by immunohistochemical analysis. Immunohistochemical assays of human breast tumor tissue with EVG F9 monoclonal antibody were carried out as described in methods. Panel A: Tissues sections were incubated with EVG F9 only. Panel B: Tissue sections were incubated with EVG F9 together with 10 g/ml of the immunogenic peptide.
formed. The model included: 1) ER␣/PR content determined by ligand binding studies and 2) ER distribution determined by immunohistochemical analyses, using EVG F9 and H222 antibodies. The Spearman rank-order correlation coefficient (r s ), the non-parametric analog of the most commonly reported measure of correlation, was determined (Table 2). There was evidence of a statistically significant strong positive correlation between the ER content determined by ligand binding assay and by the histoscore evaluation of EVG F9-stained breast tumor sections. Also, the data obtained with EVG F9 correlated well with that obtained with H222 antibody.
4. Discussion Tumor hormonal phenotype has long been utilized as a useful indicator of breast carcinoma progression. Although tumor invasion of axillary nodes predicts a high risk of metastatic disease, estrogen receptor status is an effective prognostic marker of the biological aggressiveness of the
Case #
ER (fmol/mg protein)
PR (fmol/mg protein)
ER-IHCAb (H222)
ER-IHCAb (EVG F9)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
0 0 0 0 0 0 0 23 39 55 64 100 112 132 132 144 163 171 194 196 305 327 340 356 387
0 0 40 2 4 21 9 117 0 160 6 38 113 967 141 157 10 22 493 104 114 852 195 17 176
1⫹ 0 0 0 0 2⫹ 0 1⫹ 1⫹ 0 1⫹ 1⫹ 1⫹ 2⫹ 1⫹ 3⫹ 2⫹ 2⫹ 2⫹ 2⫹ 3⫹ 1⫹ 0 2⫹ 2⫹
1⫹ 3⫹ 0 1⫹ 0 2⫹ 1⫹ 3⫹ 1⫹ 3⫹ 3⫹ 0 1⫹ 0 2⫹ 2⫹ 0 1⫹ 3⫹ 1⫹ 3⫹ 0 2⫹ 3⫹ 3⫹
a Tumor ER content was measured by ligand binding assay and immunohistochemistry with two monoclonal antibodies (H222 and EVG F9). Tumor PR content was measured by ligand binding assay. Results of the immunochemical analyses were assessed semi-quantitatively, and a histoscore was assigned to each specimen based on the intensity of staining and the percentage of positively stained cells for each antibody. b ER-IHCA indicates tumor estrogen receptor content determined by immunohistochemical analyses. Type of the anti-ER antibody used is included.
primary tumor [3,8]. Previous studies have shown that tumors, which do not express ER (ER-negative), are often poorly differentiated and more clinically aggressive [27]. Currently, several ER assay techniques are utilized for determination of ER content in human breast tumors. Due to the advanced methods of early detection of breast tumors using mammography and other imaging techniques, the small size of tumors detected and excised may preclude quantitation of ER by conventional ligand-binding assays, which requires a minimum of 200 to 300 mg of tissue. Thus, immunoblotting and immunohistochemical analyses of ER become the likely methods of choice. However, both of these methods suffer from some limitations. One of the limitations is that detection of ER is only semi-quantitative. Furthermore, these methods depend on the availability of sensitive and specific ER antibodies. In this study, we demonstrated that EVG F9 is specific for ER␣ and that this monoclonal antibody does not crossreact with recombinant rat ER. Given the high sequence
434
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
Fig. 5. Subcellular localization of estrogen receptors in human breast tumor tissue determined by immunohistochemical analysis with EVG F9. Immunohistochemical assays of human breast tumor tissue with EVG F9 monoclonal antibody were carried out as described in methods. Different degrees of nuclear staining intensity and distribution patterns were visualized by light microscopy. Panel A: strong immunostaining of human breast tumor tissue with EVG F9 (⫻10). Panel B: moderate immunostaining of human breast tumor tissue with EVG F9. (⫻10). Panel C: weak immunostaining of human breast tumor tissue with EVG F9. (⫻10). Panel D: human breast tumor tissue incubated with nonimmune mouse IgG (negative control) (⫻10).
homology between the N-terminal domains of human and rat ER, we do not anticipate that EVG F9 will react with human ER. Further, EVG F9 binds to ER␣ at a defined
epitope as indicated by complete displacement of EVG F9 binding to ER␣ by the immunogenic free peptide. We further demonstrated that EVG F9 detects denatured human
Fig. 6. Subcellular localization of estrogen receptors in human breast tumor tissue determined by immunohistochemical analysis with H222 monoclonal antibody. Tissue sections from the same blocks of human breast tumors shown in Fig. 5 were subjected to immunohistochemical analyses using monoclonal antibody H222. Different degrees of nuclear staining intensity and distribution patterns were visualized by light microscopy. Panel A: strong immunostaining of human breast tumor tissue (⫻10). Panel B: weak immunostaining of human breast tumor tissue (⫻10). Panel C: moderate immunostaining of human breast tumor tissue (⫻10). Panel D: human breast tumor tissue incubated with nonimmune mouse IgG (negative control) (⫻10).
N.S. Rost et al. / Steroids 65 (2000) 429 – 436 Table 2 Comparison of ER/PR values by ligand-binding studies and immunohistochemistrya Variable
ER
PR
ER-IHCAb (H222)
ER-IHCAb (EVG F9)
ER PR ER-IHCAb (H222) ER-IHCAb (EVG F9)
1
0.62 1
0.5 0.26 1
0.81 0.35 0.52 1
a Tumor biochemical and immunohistochemical characteristics were evaluated using Spearman’s rank-order correlational analysis. The matrix of intercorrelations represents Spearman’s rank-order correlation coefficients, (rs), for all pairs of variables. The values below the diagonal are redundant of those above the diagonal, and therefore, are not included. The values on the diagonal are all 1.00, representing the perfect correlation of each variable with itself. Bold print is used to distinguish statistically significant values (P ⬍ 0.05). b ER-IHCA indicates tumor estrogen receptor content determined by immunohistochemical analyses. Type of the anti-ER antibody used in included.
ER␣ at low concentrations, approaching 5 to 10 fmol/sample, indicating the antibody’s specificity and sensitivity. The data from Western Blot analyses reflected the values of ER␣ measured by ligand-binding assays. It should be noted that immunoblotting detects total ER␣ (i.e. functional and nonfunctional) whereas ligand-binding assays detect only the unoccupied functional ER␣. Discrepancies between the results of ligand-binding studies and Western Blots are expected in some tumor samples due to the differences in the detection methods of these two approaches. Because EVG F9 monoclonal antibody recognizes a specific epitope on the receptor, removal of this epitope due to proteolysis would result in a decrease in the amount of ER␣ detected by blotting. Even with these technical limitations, the immunoblotting assays provided an ER␣ status similar to those obtained with ligand-binding assays. Analysis of ER␣ in formalin-fixed, paraffin-embedded human breast tumor samples demonstrated that EVG F9 monoclonal antibody is specific, and produces comparable staining to that obtained with H222 antibody (r s ⫽ 0.52). The data obtained with immunohistochemistry correlated well with those determined by ligand-binding studies (EVG F9, r s ⫽ 0.81; H222, r s ⫽ 0.50). These observations suggest that EVG F9 monoclonal antibody is a useful tool for the analysis of ER␣ in tumor extracts by Western Blot and in fixed human breast tumor samples by immunohistochemistry. It is important to note that although there was strong agreement between the data obtained by ligandbinding assays and immunohistochemical assay, in some tumors the immunohistochemical assay indicated presence of ER␣, whereas ligand-binding assays did not and vice versa. These discrepancies may be related to the detection of functional and nonfunctional ER␣ by these two methods. Further, comparing the data obtained with EVG F9 and H222, there was an agreement in 14 of 25 of tumors (56%). In four tumors, EVG F9 monoclonal antibody detected low
435
levels of ER␣, whereas the H222 monoclonal antibody did not. Similarly, H222 detected ER␣ in four tumors whereas EVG F9 did not. These observations suggest that epitope preservation and retrieval may be different in some tumors and this may explain the differences in ER␣ detectability with various antibodies. Although ER␣ is undetectable by ligand-binding assays in some tumors, and these tumors are denoted as ER-negative, ER␣ is detectable by immunohistochemistry and this may reflect the greater sensitivity of ER␣ detection by EVG F9. It should be noted that ER␣ detected by immunohistochemistry correlated well with that of PR determined by ligand-binding assays. This observation lends further support to strong association between values determined by ligand-binding assays and immunohistochemistry. Furthermore, we consistently observed varying degrees of cytoplasmic staining in human breast tumors. Because the detection of ER␣ by EVG F9 using Western Blot analysis and the staining of ER␣ in immunocytochemistry are abolished by inclusion of the immunogenic peptide, we suggest that the cytoplasmic staining represents some form of ER␣ in the cytoplasm. Whether this represents a functional or nonfunctional ER␣ and the significance of this observation, however, remains to be determined. The data presented here suggest that human breast tumor ER␣ status determined by immunoblotting and immunohistochemical analyses, using EVG F9 monoclonal anti-ER␣ antibody, produced comparable results to those obtained by ligand-binding assays. These observations suggest that EVG F9 is potentially a useful immunochemical tool for assessment of ER␣ status in human breast tumor tissues, using Western Blot analysis or immunohistochemistry.
Acknowledgment We wish to acknowledge Dr Antonio de las Morenas of the Mallory Institute of Pathology for his helpful discussion of the immunohistochemistry data.
References [1] Qualters JR, Lee NC, Smith RA, et al. Breast and cervical cancer surveillance, United States, 1973–1987. MMWR CDC Surveillance Summaries 1992;41(2):1–15. [2] Miller BA, Feuer EJ, Hankey BF. Recent incidence trends for breast cancer in women and the relevance of early detection. CA-A Cancer J Clin 1993;43(1):27– 41. [3] McGuire WL. Hormone receptors: their role in predicting prognosis and responses to endocrine therapy. Semin Oncol 1978;5:428 –33. [4] Burstein NA. Overall principles of cancer management. In: Cancer manual, 6th ed. Boston: American Cancer Society, Massachusetts Div., 1982. p. 67–71. [5] Brooks SC, Saunders DE, Singhakowinta A, et al. Relation of tumor content with response of patient to endocrine therapy. Cancer 1980; 46:2775– 8.
436
N.S. Rost et al. / Steroids 65 (2000) 429 – 436
[6] Degenshein GA, Bloom N, Tobin E. The value of progesterone receptor assays in the management of advanced breast cancer. Cancer 1980;46:2789 –93. [7] McGuire WL, Pearson OH, Segaloff A. Predicting hormone responsiveness in human breast cancer. In: McGuire WL, Carbone PP, Vollmer ED, editors. Estrogen receptors in breast cancer. New York: Raven Press, 1975. p. 17–30. [8] McGuire WL. Prognostic factor for recurrence and survival in human breast cancer. Breast Cancer Res Treat 1987;10:5–9. [9] Elledge RM, McGuire WL. Prognostic factors and therapeutic decisions in axillary node-negative breast cancer. Ann Rev Med 1993; 44:201–10. [10] Gelbfish GA, Davidson AI, Kopel S, et al. Relationship of estrogen and progesterone receptors to prognosis in breast cancer. Ann Surg 1988;207(1):75–9. [11] Williams MR, Todd JH, Ellis IO, et al. Oestrogen receptors in primary and advanced breast cancer: an 8 year review of 704 cases. Br J Cancer 1987;55:67–73. [12] Kohail HM, Elias EG, El-Nowiem SA, et al. A multifactorial analysis of steroid hormone receptors in stages I and II breast cancer. Ann Surg 1985;201(5):611–7. [13] Donegan WL. Prognostic factors, stage, and receptor status in breast cancer. Cancer 1992;70:1755– 64. [14] Millis RR. Correlation of hormone receptors with pathological features in human breast cancer. Cancer 1980;46:2869 –71. [15] McCarty KS, Barton TK, Fetter BF, et al. Correlation of estrogen and progesterone receptors with histologic differentiation in mammary carcinoma. Cancer 1980;46:2851– 8. [16] Cooper JA, Rohan TE, Cant EL, et al. Risk factors for breast cancer by oestrogen receptor status: a population-based case-control study. Br J Cancer 1989;59:119 –25. [17] Ballard–Barbash R, Griffin MR, Fisher LD, et al. Estrogen receptors in breast cancer. Am J Epidemiol 1986;124:77– 84. [18] King RJB, Redgrave S, Hayward JL, Millis RR, Rubens RD. The measurement of receptors for oestradiol and progesterone in human breast tumors. In: Steroid receptor assay in human breast tumors. King RJB, editor, Methodological and Clinical Aspects. Alpha Omega, Cardiff, 1979. p. 55–72. [19] Shousha S, Coady AT, Stamp T, James JR, Alaghband–Zadeh J. Oestrogen receptors in mucinous carcinoma of the breast: an immunohistological study using paraffin wax sections. J Clin Pathol 1989; 42:902–5.
[20] Shousha S, Stamp T, James JR, Alaghband–Zadeh J. Immunohistological study of oestrogen receptors in breast carcinomas that are biochemically receptor negative. J Clin Pathol 1990;43:239 – 42. [21] Traish AM, Pavao M. Binding of site-directed monoclonal antibodies to an epitope located in the A/B region (amino acids 140 –154) of human estrogen receptor induced conformational changes in an epitope in the DNA-binding domain. Steroids 1996;61:549 –56. [22] Traish AM, Al–Fadhli S, Klinge C, Kounine M, Quick TC. Identification of structurally altered estrogen receptors in human breast cancer by site-directed monoclonal antibodies. Steroids 1995;60:467– 74. [23] Traish AM, Newton AW, Styperek K, Beazley R, Kavanah M. Estrogen receptor functional status in human breast cancer. Diagn Mol Pathol 1995;4(3):220 – 8. [24] Joel PB, Traish AM, Lannigan DA. Estradiol-induced phosphorylation of serine 118 in the estrogen receptor is independent of p42/p44 mitogen-activated protein kinase. J Biol Chem 1998;273(21):13317– 23. [25] Oyaizu T, Seizaburo A, Takehiko H, Tsubura A. Immunohistochemical detection of estrogen and progesterone receptors performed with an antigen-retrieval technique on methacarn-fixed paraffin-embedded breast cancer tissues. J Surgical Res 1996;60:69 –73. [26] Helin ML, Helle MJ, Helin HJ, Isola J. Proliferative activity and steroid receptors determined by immunohistochemistry in adjacent frozen sections of 102 breast carcinomas. Arch Pathol Lab Med 1989;113:854 –7. [27] Anderson J, Bentzen SM, Poulsen HS. Relationship between radioligand binding assay, immunoenzyme assay and immunohistochemical assay for estrogen receptors in human breast cancer and association with tumor differentiation. Eur J Cancer Oncol 1988;24:377– 84. [28] Enmark E, Pelto–Huikko M, Grandien K, et al. Human estrogen receptor beta-gene structure, chromosomal localization, and expression pattern. J Clin Endocrinol Metab 1997;82:4258 – 65. [29] Alarid ET, Bakopoulos N, Solodin N. Proteasome-mediated proteolysis of estrogen receptor: a novel component in autologous downregulation. Mol Endocrinol 1999;13:1522–34. [30] Nawaz Z, Lonard DM, Dennis AP, Smith CL, O’Malley BW. Proteasome-dependent degradation of the human estrogen receptor. Proc Natl Acad Sci USA 1999;96:1858 – 62.